Bridge, MFT, achromats, dSLR, primes - a journey of exploration

[LCPete]

Hi Nick sorry about yesterday was worn out busy day at work
I work 4 on4 off which is great for the time off but 12 hours shifts can be hard going!
Anyway I hadnt thought about the effective aperture, to be honest I don't go into the technical side of photography that much or even experiment very often I like to keep things simple and just concentrate on the subject
I wonder if the effective apeture effect at the minimum focus distance is why my Marumi shots didn't come out that sharp as I was using a smaller apeture than I thought but to be honest I set the apeture to 8 so the effective apeture would have been about 16 So should still be ok
I do remember that my friend who uses the MPe 65 for close shots said that he never uses an apeture smaller than 5.6 as you get diffraction softening at high magnification
I do still wonder if there's something that I'm not doing right with the Marumi achromat as the online reviews for it are very good , I will try again with it in the spring when the flies and hover flies etc are out, maybe photographing coins isn't that relevant to shooting insects


Looking at your shots starting at the sigma 105 at f22 and 11 the f 11 ones are sharper probably F11 is close the the sweet spot for that lens when you go smaller than that you start to see softening in theory anyway!

With the next 4 with the different cameras
they all look good any of them on its own would be ok

On the last 4 the two 70D setups
I agree the difference between the Effective F 45 and 22 is the same for both the f22 ones are sharper
I do prefer the look of the Sigma shots I think that the colour is more accurate, looks better
I know shooting in raw and setting up the colour profiles should make it possible to get the correct colour I think that the sigma lens in your shots just seems to gives better colours
I first thought that the sigma shots were sharper but now think that it's a contrast difference

As for my original idea about if it was possible to do higher magnification shots without flash I will just have to try it and see in the spring
Will probably have to use a higher ISO tho
Bye for now

 

[GardenersHelper]

Hi Nick sorry about yesterday was worn out busy day at work
I work 4 on4 off which is great for the time off but 12 hours shifts can be hard going!

Just one 12 hour shift would be bad enough, but four on consecutive days? That must be really hard.

Anyway I hadnt thought about the effective aperture, to be honest I don't go into the technical side of photography that much or even experiment very often I like to keep things simple and just concentrate on the subject
I wonder if the effective apeture effect at the minimum focus distance is why my Marumi shots didn't come out that sharp as I was using a smaller apeture than I thought but to be honest I set the apeture to 8 so the effective apeture would have been about 16 So should still be ok

Yes, it should have been ok even at f/16.

However, I think you are comparing tubes and the Marumi both with the 100L. In that case the effective aperture would have been smaller than f/8 for the tubes, but it would still have been f/8 for the Marumi. The smaller effective aperture with the tubes would have made the focused areas less sharp, but the less focused areas more sharp, and overall that could make the tubes version look sharper. That was what confused me in that earlier test when I thought that using the Raynox was degrading the sharpness. It was more that the smaller effective aperture with the Sigma was making the dof bigger and making the image look sharper overall.

I do remember that my friend who uses the MPe 65 for close shots said that he never uses an apeture smaller than 5.6 as you get diffraction softening at high magnification

Absolutely. The effective aperture is (1 + magnification), so for example if you go up right up to 5:1 magnification with the MPE, f/5.6 become f/(5.6 x 6), which is roughly f/32. At 3:1 magnification it would be f/(5.6 x 4), which is roughly f/22. At 2:1 it would be f/(5.6 x 3), or around f/16.

I do still wonder if there's something that I'm not doing right with the Marumi achromat as the online reviews for it are very good , I will try again with it in the spring when the flies and hover flies etc are out, maybe photographing coins isn't that relevant to shooting insects

Apart from going outside the acceptable range of working distance, I don't think there is anything to get wrong. The way I'm judging that these days is to use the manual focus ring to check that I can make it go out of focus "on both sides of the subject". That is, the camera doesn't move - I rotate the focus ring to get the subject in focus, then check I can make it go out of focus by turning the focus ring both ways, to focus nearer and to focus further away. If I can, that's fine, I'm at a suitable distance and will be able to get good focus. If I can't, then I may be at the edge of the range where I can get good focus, or worse I may be beyond it, in which case I can adjust the manual focus to get the sharpest image I can, but it still won't be as sharp as it could be if I was in the right range of working distance. So, if I can't make it go out of focus by focusing nearer, then I move away from the subject until I can. And vice versa for focusing further away.

Looking at your shots starting at the sigma 105 at f22 and 11 the f 11 ones are sharper probably F11 is close the the sweet spot for that lens when you go smaller than that you start to see softening in theory anyway!

It's an f/2.8 lens, so I imagine the sharpness sweet spot is larger than f/11, in the f/4 to f/5.6 range perhaps. (Just looked it up. According to this, it is sharpest at f/5.6 in the centre and f/8 at the edges.)

With the next 4 with the different cameras
they all look good any of them on its own would be ok

I think so. And that is a pretty interesting conclusion in itself. I think it tends to support what Bryn @Tintin124 keeps telling newcomers to close-up/macro, that you don't have to use expensive kit to get nice results. (I do think you need expensive kit to get the very best results.)

On the last 4 the two 70D setups
I agree the difference between the Effective F 45 and 22 is the same for both the f22 ones are sharper
I do prefer the look of the Sigma shots I think that the colour is more accurate, looks better

Yes, the colour business is a real puzzle. I need to look at that some more. I suppose each lens has different coatings which are affecting the colours, even when using the same camera as in this case as between the Sigma and the 55-250. As between the different cameras, presumably sensor differences complicate the issue even more.

I know shooting in raw and setting up the colour profiles should make it possible to get the correct colour

Should. But doesn't seem to, not for me at least. I'm pretty irked about that.

I think that the sigma lens in your shots just seems to gives better colours

I think it's important not to read too much into this example. I played with the white balance, and used different settings for each rig, including different settings for the Sigma and for the 55-250. So the colours we see may be more to do with how I happened to slide the sliders as I was getting more and more annoyed, and where I happened to leave them when I gave up, rather than the true differences between the rigs.

However, what is certainly the case, and what I find surprising, is just how different the colour rendition is between the rigs. I'm not so surprised about the differences between cameras (although I'm pretty miffed that camera calibration doesn't equalise them), but I am surprised at how different the colours are with two different lenses on the same camera.

I first thought that the sigma shots were sharper but now think that it's a contrast difference
As for my original idea about if it was possible to do higher magnification shots without flash I will just have to try it and see in the spring
Will probably have to use a higher ISO tho

Yes, suck it and see. Natural light works for me at base ISO on calm days with a stationary subject, but then again I'm using a tripod, and using it hands-off with a remote release for those shots, so I can use very slow exposures. But hand held, beyond 1:1, I think you'd need (well, I'm sure I need) strong ambient light for that to work hand held, quite possibly even when using as high an ISO as you are comfortable with. ISO 800 perhaps? Don't know. It depends on the circumstances as to how well it would work. Worth a try though - natural light gives (for my taste) a much nicer look to images, but I suspect you won't get as much fine detail, even if the shutter speeds are high. Given my tastes, I'm happy to live with that. But YMMV of course.

Thanks for taking the time on this Pete. It's really helpful to have your input.

 

[GardenersHelper]

I do prefer the look of the Sigma shots I think that the colour is more accurate, looks better

I think it's important not to read too much into this example. I played with the white balance, and used different settings for each rig, including different settings for the Sigma and for the 55-250. So the colours we see may be more to do with how I happened to slide the sliders as I was getting more and more annoyed, and where I happened to leave them when I gave up, rather than the true differences between the rigs.

I've had another go at getting the colours the same.

I used the four f/45-equivalent images. To recap, the Sigma was shot at 1:1. It was shot at f/22, which at 1:1 is f/45 equivalent.


0667 01 2015_01_26 70D + Sigma 105 F22 IMG_2682 Uncropped 1100h by gardenersassistant, on Flickr

The others were shot with a 1:2 framing so that cropping them to 1:1 equivalent would make them f/45 equivalent (roughly). Here for example is the framing for the 55-250 shot. This was shot at f/22. As this used the Raynox, the effective aperture was unchanged at f/22. However, cropping it to match the Sigma shot takes it to f/45 equivalent.


0667 04 2015_01_26 70D + 55-250 at 100mm + Raynox 150 F22 IMG_2690 Uncropped 1100h by gardenersassistant, on Flickr

I started again from the RAW files for these four images. I cropped the Sigma image around the top, left and right of the plant, retaining the 3:2 aspect ratio. I then cropped the others the same way. This means of course that the other three had much bigger crops and so have far fewer pixels. The smallest is the FZ200 image, which was only 1300 pixels high when cropped, so when I exported them all to JPEG I resized them all to 1300 pixels high. This made them as large as possible to help the comparisons, without wanting to go beyond a 100% crop for any of them. That said, the FZ200 version is a 100% crop, and the G3 and non-Sigma 70D versions are not far off 100%. I think that's worth thinking about when considering the image quality of the three achromat versions against an uncropped prime macro lens.

Before exporting them to JPEG I tried to get the colours, the overall brightness and the balance of light and dark as similar as I could between the four of them. This took ages. As well as altering the white balance (Temp and Tint), I made some adjustments to Exposure, Contrast and Highlights. Apart from that they have all had the same, mild, import adjustments (my standard settings for Clarity, Vibrance, Sharpening and colour noise reduction), and the same output sharpening (Lightroom's "Sharpen for Screen", Standard setting). I didn't use camera profiles this time, instead using "Adobe Standard" for them all.

I find the differences in white balance astonishing. Here are how Temp and Tint ended up.

• Sigma: Temp 12750, Tint -8
• 55-250: Temp 8300, Tint +10
• G3: Temp 5321, Tint +14
• FZ200: Temp 4900, Tint +39

I have added the 1300 pixel high versions to the previously linked album, along with the large versions of the following graphics.

Here are the new comparisons.


0667 22 2015_01_26 comparison of F45 equivalent shots with equalised crop, colour and brightness - top left by gardenersassistant, on Flickr


0667 23 2015_01_26 comparison of F45 equivalent shots with equalised crop, colour and brightness - top right by gardenersassistant, on Flickr


0667 24 2015_01_26 comparison of F45 equivalent shots with equalised crop, colour and brightness - bottom left by gardenersassistant, on Flickr


0667 25 2015_01_26 comparison of F45 equivalent shots with equalised crop, colour and brightness - bottom right by gardenersassistant, on Flickr

I did some more experiments today, comparison shots between the Sigma and 55-250, doing comparisons using the bare lenses and also using them with the Canon 500D, Raynox 150 and Raynox 250. These were done to explore the colour difference issue. I want to know if it is the lenses causing the difference, and/or the use of an achromat. I haven't looked at them yet, so it will be a while before I can draw (or not be able to draw!) any conclusions.

 

[LCPete]

Just one 12 hour shift would be bad enough, but four on consecutive days? That must be really hard.



Yes, it should have been ok even at f/16.

However, I think you are comparing tubes and the Marumi both with the 100L. In that case the effective aperture would have been smaller than f/8 for the tubes, but it would still have been f/8 for the Marumi. The smaller effective aperture with the tubes would have made the focused areas less sharp, but the less focused areas more sharp, and overall that could make the tubes version look sharper. That was what confused me in that earlier test when I thought that using the Raynox was degrading the sharpness. It was more that the smaller effective aperture with the Sigma was making the dof bigger and making the image look sharper overall.



Absolutely. The effective aperture is (1 + magnification), so for example if you go up right up to 5:1 magnification with the MPE, f/5.6 become f/(5.6 x 6), which is roughly f/32. At 3:1 magnification it would be f/(5.6 x 4), which is roughly f/22. At 2:1 it would be f/(5.6 x 3), or around f/16.



Apart from going outside the acceptable range of working distance, I don't think there is anything to get wrong. The way I'm judging that these days is to use the manual focus ring to check that I can make it go out of focus "on both sides of the subject". That is, the camera doesn't move - I rotate the focus ring to get the subject in focus, then check I can make it go out of focus by turning the focus ring both ways, to focus nearer and to focus further away. If I can, that's fine, I'm at a suitable distance and will be able to get good focus. If I can't, then I may be at the edge of the range where I can get good focus, or worse I may be beyond it, in which case I can adjust the manual focus to get the sharpest image I can, but it still won't be as sharp as it could be if I was in the right range of working distance. So, if I can't make it go out of focus by focusing nearer, then I move away from the subject until I can. And vice versa for focusing further away.



It's an f/2.8 lens, so I imagine the sharpness sweet spot is larger than f/11, in the f/4 to f/5.6 range perhaps. (Just looked it up. According to this, it is sharpest at f/5.6 in the centre and f/8 at the edges.)



I think so. And that is a pretty interesting conclusion in itself. I think it tends to support what Bryn @Tintin124 keeps telling newcomers to close-up/macro, that you don't have to use expensive kit to get nice results. (I do think you need expensive kit to get the very best results.)



Yes, the colour business is a real puzzle. I need to look at that some more. I suppose each lens has different coatings which are affecting the colours, even when using the same camera as in this case as between the Sigma and the 55-250. As between the different cameras, presumably sensor differences complicate the issue even more.



Should. But doesn't seem to, not for me at least. I'm pretty irked about that.



I think it's important not to read too much into this example. I played with the white balance, and used different settings for each rig, including different settings for the Sigma and for the 55-250. So the colours we see may be more to do with how I happened to slide the sliders as I was getting more and more annoyed, and where I happened to leave them when I gave up, rather than the true differences between the rigs.

However, what is certainly the case, and what I find surprising, is just how different the colour rendition is between the rigs. I'm not so surprised about the differences between cameras (although I'm pretty miffed that camera calibration doesn't equalise them), but I am surprised at how different the colours are with two different lenses on the same camera.



Yes, suck it and see. Natural light works for me at base ISO on calm days with a stationary subject, but then again I'm using a tripod, and using it hands-off with a remote release for those shots, so I can use very slow exposures. But hand held, beyond 1:1, I think you'd need (well, I'm sure I need) strong ambient light for that to work hand held, quite possibly even when using as high an ISO as you are comfortable with. ISO 800 perhaps? Don't know. It depends on the circumstances as to how well it would work. Worth a try though - natural light gives (for my taste) a much nicer look to images, but I suspect you won't get as much fine detail, even if the shutter speeds are high. Given my tastes, I'm happy to live with that. But YMMV of course.

Thanks for taking the time on this Pete. It's really helpful to have your input.

Hi Nick yes 12 hour shifts are hard going I do 2 12 hour days then 2 12 hour nights but do get 4 days off which is great in the summer
our factory is on shutdown for 12 days at the end of may which is brilliant I will get lots of time off at peak season at my places

yes you are right about the best F stop for sharpness I've used my Sigma 105 at 3.5 many times and its amazingly sharp at that aperture

yes agree again that it doesn't matter what kit you use its whatever is easier to get on with
and Bryn is right that for people trying out macro its perfectly possible to get decent results with cheap kit
but it is nice to use a " proper" macro lens
it makes it easier for me that the autofocus on my canon macro works well a macro distances
when things go well the results are wonderful too

Yes with going for higher magnification than 1:1 in natural light without flash
I think that I may struggle a bit
with natural light but as you say I will suck it and see
I will just have to get a bit lucky and find a cooperative subject on a bright day
the insects are often approachable on rainy days but I will have to see if I can get a high enough shutter speed at higher magnification at lower light levels
am always glad to try to help and it's interesting too and hopefully we are both getting somewhere

I had planned to spend a few days in the Chester zoo butterfly house but on the last two occasions there were very few butterflies which is disappointing
never mind theres plenty of larger subjects at the zoo including tiger cubs
it's nice in a way to do something different in the winter as I wont get jaded with doing macro which could happen if I did it all year round

 

[LCPete]

I really don't understand why the white balance settings in raw conversion that you had to use between the cameras was so different
just differences with different formats/gear I guess
at least you have shown that it is possible to match the colour between them
also shows that it definitely pays to always shoot in raw as you can get so much adjustment afterwards
on the previous set of tests that you did the Sigma 105 did seem to have the best colours and contrast out of the camera if you know what I mean . At least with the sigma you wont have to spend too much time in lightroom making adjustments

 

[GardenersHelper]

I really don't understand why the white balance settings in raw conversion that you had to use between the cameras was so different

Nor me.

just differences with different formats/gear I guess

Maybe there is some operator error in here somewhere. Especially the fact that I can't get camera profiles to produce consistent results between rigs makes me wonder if I'm doing something wrong.

at least you have shown that it is possible to match the colour between them

Possible yes. But I wouldn't want to try doing that too often. Although perhaps I'd get better at it, and faster, with practice!

also shows that it definitely pays to always shoot in raw as you can get so much adjustment afterwards

Agree 100%

on the previous set of tests that you did the Sigma 105 did seem to have the best colours and contrast out of the camera if you know what I mean . At least with the sigma you wont have to spend too much time in lightroom making adjustments

I suspect so. My thinking is similar with the 70D versus the other cameras - I find it much easier to get the 70D images to look nice, without really doing too much to them usually. With the G3, and even more so with the FZ200, it is harder work.

 

[LCPete]

[QUOTE="GardenersHelper, post: 6694422, member:



I suspect so. My thinking is similar with the 70D versus the other cameras - I find it much easier to get the 70D images to look nice, without really doing too much to them usually. With the G3, and even more so with the FZ200, it is harder work.[/QUOTE]

Maybe thats a good reason on it's own to use the 70D and Sigma 105 combination
it will save a lot of time afterwards on the computer
I'm not at all familiar with non SLR digital cameras but suspect that Canon and Nikon at least have got the out of the camera colour settings pretty spot on by now
I know that I normally only have to make slight adjustments to white balance and macro shots seem to need no additional changes at all I just convert to tiff sharpen in photoshop then save as a jpeg

 

[GardenersHelper]

This example hints at something that sounds rather unlikely, namely that I might be able to get results with the 55-250 STM on the 70D that are as good as I can get with the Sigma 105 Macro, both when using the 55-250 by itself and when using it with an achromat for magnifications around 1:1. It would be surprising if a zoom lens could match a prime, and even more surprising if it could do it with an inexpensive achromat on it.

That was only one example, not nearly enough to base any firm conclusions on. I decided to do some more comparisons today. I was primarily interested in natural light, low magnification shots, thinking about what kit to use for larger invertebrates when they turn up and for flowers. With it being windy and wet I decided to work indoors. My wife provided some subjects for me with some fine detail, varied textures and subtle colours.

I captured the shots in the kitchen again, because of its solid floor, and used a hands-off tripod technique, with a wired remote shutter release. I used live view and magnified each shot to 5x on the LCD so I could make sure it had stopped moving before taking the shot. I used autofocus, and using it with 5x magnification meant I could place the centre of focus quite precisely and consistently.

I captured images of six scenes. For two of the scenes I used three setups, all on the 70D: the Sigma 105, the 55-250 STM by itself, and the 55-250 with a mild +2 diopters Canon 500D close-up lens. For the other four scenes I used just the Sigma and the bare 55-250.

I captured each scene from f/5.6 (the maximum aperture that the 55-250 can use) to f/22 (the minimum aperture that the Sigma 105 can use). I captured just one shot at each aperture.

The light varied a lot during the session, varying from low sun coming strongly through thin foliage right through to dark overcast as sleet/snow started falling, at which point it was gone 5pm and the light level was dropping anyway. All the images were captured at ISO 100, and the light level became so low that the final, f/22 capture had a 25 second exposure.

I imported the raw files into Lightroom using my default input adjustments.

After looking through the images I decided to work in detail on one aperture for each scene. This resulted in one comparison each for f/5.6, 8, 11 and 16, and two for f/22.

Because of changes in the lighting while I was capturing scenes (and possibly for other reasons, as discussed, inconclusively, previously), some of the images came out looking rather different from one another in terms of colour. I adjusted the white balance (Temp and Tint) to try to get the images for each scene to look similar. I also did some cropping/rotation to try to get the framing as similar as I could so as to aid comparisons, and for the same reason I adjusted Exposure for several images to try to get each pair/triplet to a similar brightness.

This image (larger version over at Flickr) gives an overview of the scenes as adjusted. From left to right, and top to bottom, they are

• f/5.6 - Sigma 105, bare 55-250 and 55-250 with 500D
• f/8 – Sigma 105 and bare 55-250
• f/11 - Sigma 105, bare 55-250 and 55-250 with 500D
• f/16 - Sigma 105 and bare 55-250
• f/22 - Sigma 105 and bare 55-250
• f/22 - Sigma 105 and bare 55-250

0670 15 2015_01_29 Overview of selected comparisons by gardenersassistant, on Flickr

Here, in the same order, is what they looked like before I made the adjustments.


0670 16 2015_01_29 Overview of selected comparisons prior to equalising adjustments by gardenersassistant, on Flickr

As the conclusion I have drawn from these comparisons may be a little controversial, I have posted JPEGs here at Flickr which are at the full size remaining after the post processing. I will happily make raw files available on request, and I will put unadjusted full size JPEGs of all the scenes at all the apertures I used if anyone would like to make sure that I haven't “cherry picked” the ones that best fit my prejudices. I think I picked them pretty randomly, apart from wanting one of each aperture, but the unconscious mind is a mysterious and devious operator sometimes.


Conclusions

Having pixel-peeped these images, I can see that there are some differences in the amount of detail in some parts of the images. However, each of the three setups appears sharpest/most detailed in some areas. I have not been able to identify any systematic variation which indicates that one of the rigs is better overall than the others, either in terms of sharpness or in the rendition of colours. I find this very surprising indeed and find it hard to accept. But unless there is some unconscious prejudice going on here (and I don't rule that out), that is the evidence that my eyes see. Because I can scarcely believe my own eyes here, I will be very interested indeed if anyone else is interested enough to take some time to take a close look at some of these.

Perhaps these subjects were not valid in relation to the subjects I want to use this kit for. I don't know. Perhaps there is something else I've missed/done wrong – that wouldn't be the first time! But if this conclusion were to be valid for my prospective subjects then I think I might have little reason to use the Sigma for low magnification natural light shots, because

• The Sigma has less reach than the 55-250, so I can't reach some subjects with it that I can with the 55-250.
• The Sigma can be awkward to use compared to the 55-250 with/without achromats when composing shots using a tripod, which is how I usually work.
• The Sigma is heavier than the 55-250 and this adds a bit of difficulty when framing shots because the extra weight makes the tripod arm and ball head move further when settling into position.
• I rarely use apertures larger than f/5.6, which is something the Sigma offers but the 55-250 doesn't.

That said, my wife (who is a pretty acute observer, both of real life flora, their colours and their textures, and of images) thinks some of the flower images I have captured with the Sigma have something extra, something “special” about them, but that is difficult to pin down. However, looking at the larger size version of the overview image above she said she couldn't really see much difference but that she did marginally prefer non-Sigma versions in a couple of cases. So, no firm conclusions from either of us yet.

There is also the issue of whether the Sigma images are easier to handle in post processing. I did think so, but after today's test I'm not so sure.

These tests used a hands-off tripod technique. Working hand held might be a different matter, if the image stabilisation on the Sigma is significantly better for the type of scenes I have in mind than the 55-250. I did some preliminary tests of that today and these did not indicate any signficant difference between the two lenses in respect of image stabilisation effectiveness. These were quick, unrealistic, ultra high ISO “ruler shots” at various shutter speeds. I'll do some more realistic tests of this out of doors on better subjects.

 

[LCPete]

I see what you mean I can't see any meaningful differences between them but I am viewing on a small screen (am at work! )
I think that the marble is a difficult subject to use tho as the details are inside the glass
the duck shots all look good though
maybe as you say you need "real" subjects to really test each setup properly
I think that insects have a different sort of finer detail than the test subjects you've used
don't know what the best things to use tho I couldn't think of something that has the same sort of fine details that insects have
it does look like there isn't a great deal of difference between the setups tho and it may be a case of which is the easiest to use
I'm not familiar with the new Sigma 105 with OS (mine is an old version) but my Canon 100L has IS that works at macro distance (1:1) it gives two stops of help at 1:1 if the Sigma is the same thats a good reason to use the macro lens for close work
I still believe that the macro lens will givethe best results for insects close up as that what's its designed for but thats just my feeling I don't actually know if that's true

 

[GardenersHelper]

I'm not familiar with the new Sigma 105 with OS (mine is an old version) but my Canon 100L has IS that works at macro distance (1:1) it gives two stops of help at 1:1 if the Sigma is the same thats a good reason to use the macro lens for close work

I read somewhere, can't recall where though, that the 100L image stabilisation is a bit special. My guess is that the effectiveness of the IS in the Sigma, like in possibly all other macro lenses other than the 100L, approaches zero as the magnification approaches 1:1. However, I just found this, which talks about "very usable, crisp images from shutter speeds as low as 1/4 of a second". I don't know what magnification he was referring to though. This review says "OS allowed reasonably-sharp-image hit rates with the lens handheld down to 1/10 sec. About 73% of my images were usably sharp and about 50% were very sharp at this shutter speed. This equates to about 3 1/3 stops of assistance for me." Again, no mention of the magnification. This says "At a focal distance of about 33 cm, the effectiveness is approximately 1 stop." 33cm from subject to sensor is almost 1:1 (1:1.2). The 1/focal length guideline would make this about 1/50 sec at 1:1.

In my little test today there was nothing sharp at 1/20 sec or slower at 1:1.

I'm not too bothered either way though as I normally use flash and/or a tripod (hands-on or hands-off).

I still believe that the macro lens will givethe best results for insects close up as that what's its designed for but thats just my feeling I don't actually know if that's true

I'd like to think that is the case, but given these latest results I can't really see why it would be (but that doesn't mean it isn't!) I'll just have to try it when the opportunity arises, although as I discovered with my short time with the 100L it's really difficult to get a good comparison out in the wild. A still early morning would be the obvious test-bed - dew-covered flies etc.

 

[GardenersHelper]

I read somewhere, can't recall where though, that the 100L image stabilisation is a bit special. My guess is that the effectiveness of the IS in the Sigma, like in possibly all other macro lenses other than the 100L, approaches zero as the magnification approaches 1:1. However, I just found this, which talks about "very usable, crisp images from shutter speeds as low as 1/4 of a second". I don't know what magnification he was referring to though. This review says "OS allowed reasonably-sharp-image hit rates with the lens handheld down to 1/10 sec. About 73% of my images were usably sharp and about 50% were very sharp at this shutter speed. This equates to about 3 1/3 stops of assistance for me." Again, no mention of the magnification. This says "At a focal distance of about 33 cm, the effectiveness is approximately 1 stop." 33cm from subject to sensor is almost 1:1 (1:1.2). The 1/focal length guideline would make this about 1/50 sec at 1:1.

In my little test today there was nothing sharp at 1/20 sec or slower at 1:1.

Yesterday, using the 70D, I did a more detailed test of the impact of image stabilisation on hand-held shots. I photographed a door mat a bit more than 1100 times, at four magnifications (1:1, 1:1.5, 1:2 and 1:5) using the Sigma 105 and the 55-250 (with Raynox 150 for 1:1, 1:1.5 and 1:2 and Canon 500D for 1:5).


0671 01 2015_01_30 Test scenes by gardenersassistant, on Flickr

I used a variety of apertures and ISOs to get a spread of shutter speeds. I didn't capture shots for every combination of lens, shutter speed and IS on/off. In the following table cells are blank where there were no shots for that combination.

The cells for 1/8 sec to 1/125 sec include shutter speeds 1/3 stop more and 1/3 stop less than the shown shutter speed.

I went through all the shots and picked out those I thought were sharp, or "fairly sharp". There was no clear line between the two categories, and there are likely to have been inconsistencies in my marking. Also, there is the question of "How sharp is sharp enough?". Personal preferences, output medium and size, eyesight etc come into this. For example, I'm not sure that I would want to use shots that were only as sharp as some that I marked as "fairly sharp".

Also of course, there was variability in the shooting, depending on how much my hand was shaking for each shot. In addition, some of the cells are based on very few shots. All this variability coupled with uneven data spread means that it is important not to put too much emphasis on comparing any single pairs of cells. However, even where there is a lot of variability in a dataset it can be possible to see consistent patterns emerge, and I think that is the case here.

Here are the overall results.


0671 02 2015_01_30 Percentage of sharp or fairly sharp images by gardenersassistant, on Flickr

It is easier to see what is going on by breaking this down a bit.

The next table compares the results using and not using image stabilisation with the Sigma 105.


0671 03 2015_01_30 Impact of stabilisation for Sigma 105 by gardenersassistant, on Flickr

Image stabilisation has a positive effect, but as magnification increases the effect decreases. In very broad terms (a qualification which applies to all that follows) it looks like there might be something like 2 to 3 stops benefit at 1:5, 1 to 2 stops at 1:2, and possibly a small benefit at 1:1.5 and 1:1.

Now the results using and not using image stabilisation with the 55-250 STM.


0671 04 2015_01_30 Impact of stabilisation for 55-250 STM by gardenersassistant, on Flickr

The overall pattern looks similar to that with the Sigma 105.

Next a comparison of the two lenses with both having IS on.


0671 05 2015_01_30 Comparison between lenses with IS on by gardenersassistant, on Flickr

This makes the Sigma 105 stabilisation look a bit better at lower magnifications, with no obvious advantage at 1:1.5 and 1:1. This is consistent with the little table at the bottom right of the overall results table above.

Finally, how the lenses compare without IS. This might seems like a pointless comparison, but the Sigma 105 is a lot heavier than the 55-250 and I wondered if this might have any impact.


0671 06 2015_01_30 Comparison between lenses with IS off by gardenersassistant, on Flickr

The results are variable. There might just be a small overall advantage for the Sigma 105 I suppose, but if there is it looks very small.

 

[TimmyG]

Interesting results. Maybe this helps explain why there is no MP-E mark II with IS.

 

[GardenersHelper]

Interesting results. Maybe this helps explain why there is no MP-E mark II with IS.

Good point. I hadn't thought of that, but now you mention it, it does look like it would be pretty pointless.

I seem to remember Canon claim 2 stops for their hybrid IS on the 100L at 1:1, but even so it looks like that would disappear to little or nothing as the magnification increases. And in any case I imagine almost everyone using the MPE65 uses flash, which makes image stabilisation irrelevant for image capture. And most probably use the viewfinder rather than live view for image acquisition/composition, so even a bit of steadying of the live view wouldn't help for the most part. No, not much mileage in it for Canon.

 

[GardenersHelper]

More on white balance variations.

In this post, the question arose of why images of the same scene should look so different in colour even though camera profiles and shot-specific white card white balance settings had been used. Was it because of differences in ambient lighting, or was it something else such as differences between the lenses or the sensors?

I have had another look at the images form this post, in which I had noticed something curious but forgot to mention it.

Here are two pairs of images. In all four cases the white balance is As Shot. On the top we can see one of each pair which looks looks desaturated or something similar, most obviously in the background, and on the bottom we can see the other ones of the pairs which look more colourful, again most obviously in the background. The thing is that in one case it is the Sigma shot which looks desaturated and in the other case it is the 55-250 that looks desaturated. So it can't be the lenses rendering the colours differently. And both were shot with the 70D, so it isn't a sensor difference.


0670 25 2015_01_29 As shot pair comparison by gardenersassistant, on Flickr

The annotations on the side show that for each of the pairs the camera registered a similar white balance for each of the shots in the pair. But the colours look different.

On the other hand, here are the same images, in the same order, after I had attempted to make them look similar. As you can see, I had moved the white balances further apart in order to make the images look more similar, especially for the right hand pair which have a particularly large difference in Temp.


0670 26 2015_01_29 Equalised appearance pair comparison by gardenersassistant, on Flickr

As I mentioned in the original post about the images, the ambient light was changing back and forth rapidly, and changing a great deal. My guess is that this is where the differences in colours come from.

What still puzzles me, and perhaps someone who understands these things could explain it, is this: If the ambient light was different enough to cause the images to look so different, why was the As Shot white balance also not different? I was expecting to see different white balances for the different looking As Shot pairs and more similar white balances in the similar looking adjusted pairs. But what I actually got was the other way round. There's obviously something I don't understand about white balance here.

 

[TimmyG]

Not sure i fully understand all the factors here or how you are using profiles. When you say camera profiles, are they in camera or applied in post? What was the white balance setting for the "as shot?" It wasn't set to auto by any chance?

 

[GardenersHelper]

Not sure i fully understand all the factors here or how you are using profiles. When you say camera profiles, are they in camera or applied in post?

Yes, applied in post. This post and the following posts was my initial summary of how I went about it. I then was told about a better way of handling it, which I tried, liked, adopted and still use - it is described in this post and the subsequent posts.

What was the white balance setting for the "as shot?" It wasn't set to auto by any chance?

Yes. Auto White Balance.

 

[TimmyG]

If the ambient light was different enough to cause the images to look so different, why was the As Shot white balance also not different? I was expecting to see different white balances for the different looking As Shot pairs and more similar white balances in the similar looking adjusted pairs. But what I actually got was the other way round.

So the way I see it, the camera will always try to achieve a standardised colour for a scene, regardless of lighting conditions (in a similar way Auto ISO/Aperture/Shutter will always try to expose a scene to middle grey, even if the majority of the image is bright white or pitch black). So regardless if you have two shots set to auto, yet changing colour conditions between, the camera will alter the RGB values between shots to achieve the colour it "thinks" is correct. That is why you are seeing similar results from the "as shot" images, the camera is compensating for the changing light leaving you with similar looking images (of course it will never actually be able to give consistent results, and one of the reasons you shouldn't use auto for stitching images together etc.).

If you go for on of the other WB options (daylight, shadow, tungsten etc.) you are "fixing" your RGB values and you will be able to better see the affects the changing light conditions are having.

 

[GardenersHelper]

So the way I see it, the camera will always try to achieve a standardised colour for a scene, regardless of lighting conditions (in a similar way Auto ISO/Aperture/Shutter will always try to expose a scene to middle grey, even if the majority of the image is bright white or pitch black). So regardless if you have two shots set to auto, yet changing colour conditions between, the camera will alter the RGB values between shots to achieve the colour it "thinks" is correct. That is why you are seeing similar results from the "as shot" images, the camera is compensating for the changing light leaving you with similar looking images (of course it will never actually be able to give consistent results, and one of the reasons you shouldn't use auto for stitching images together etc.).

Thanks Tim. For some reason I'm having trouble getting my head around this. I'm not seeing similar looking images, I'm seeing different looking images, and they look different even though they have the (more or less) same white balance. The right hand example in the two pairs above is the best example of this - the white balance is almost exactly the same in the two images, but they look very different. Intuitively, if they look different, I'd expect them to have different white balances.

But, following your line of thought ... if the light had changed but the camera had used the same white balance, then it doesn't seem unreasonable that they would look different. So I change the white balance to make them look similar...... Yes, that does make sense.

If you go for on of the other WB options (daylight, shadow, tungsten etc.) you are "fixing" your RGB values and you will be able to better see the affects the changing light conditions are having.

Yes, that would simplify things. Good idea. I've set the 70d to Cloudy. Thanks again Tim.

 

[TimmyG]

Yes, that would simplify things. Good idea. I've set the 70d to Cloudy. Thanks again Tim.

It's worth checking your metering mode also. If you have it on spot metering I imaging moving the "spot" a few pixels either side could have significant impacts on the White Balance when set to Auto. For example in your images above, the metering spot could be a single pixel on the red vase. How red should that vase be? The camera has no idea so will make a best guess. If that spot moves slightly into one in more light (or indeed the light changes) then it could have dramatic effect on the rest of the image. Of course, this issue goes away if you specify a fix White Balance "temperature" with one of the other white balance settings.

 

[GardenersHelper]

It's worth checking your metering mode also. If you have it on spot metering I imaging moving the "spot" a few pixels either side could have significant impacts on the White Balance when set to Auto. For example in your images above, the metering spot could be a single pixel on the red vase. How red should that vase be? The camera has no idea so will make a best guess. If that spot moves slightly into one in more light (or indeed the light changes) then it could have dramatic effect on the rest of the image. Of course, this issue goes away if you specify a fix White Balance "temperature" with one of the other white balance settings.

I don't know how large an area the camera looks at to make its white balance calculation, and if the area it uses would match the spot metering area. That said, I certainly had a huge change in colours from a tiny shift in position on one of the few occasions that I used spot metering. I think the nearby spots were of greatly different brightness and I suspect that metering from one of them caused radical overexposure of one (I assume the red) channel, which turned the red poppy petals yellow.

But I use evaluative metering almost all the time, which presumably wouldn't be sensitive to small shifts in position.

 

[GardenersHelper]

I have been experimenting with stacking for botanical subjects. This has revealed a rather interesting difference between the Sigma 105 and the 55-250 STM.

I read a while ago that one of the advantages of stacking for flower shots was that by stacking large aperture shots you could (depending on the spatial characteristics of the scene) get a nicely out of focus background while still getting all of the subject in focus, in a way that was impossible with a single shot. Here is an example from my experimenting a few days ago.

Here is a single-image version of a scene captured with the 55-250 on the 70D. I wanted to get the flower and the buds in focus, but despite using f/32 for maximum depth of focus, the front-to-back distance of the flower and buds (coupled with filling so much of the frame with the subject) meant that not everything I wanted was in focus.


0672 02 2015_01_30 IMG_4570 LR by gardenersassistant, on Flickr

Here is a 10-image stack of the scene. The individual images used f/10. This was one of several experiments using different apertures for the stack shots. In general I will probably be using larger apertures, but even at f/10 the flower looks sharper and more detailed, and so do the buds, while the background is much more out of focus.


0672 03 2015_01_30 IMG_4583-92 ZS PMax LR by gardenersassistant, on Flickr

However, I noticed while doing these experiments that the image was changing size as I moved the focus rail from back to front to get different parts of the scene in focus. Here is an example I constructed indoors to explore this effect.


0672 04 2015_01_30 Subject size change comparison by gardenersassistant, on Flickr

This change in size doesn't matter to Zerene Stacker, which handles it fine, but it turns out that it does matter to me. I often compose my botanical stuff right to the edge of the screen and the changing size was a bit offputting, complicating matters. I discovered, as illustrated above, that when I moved the focus point with the focus ring rather than the focus rail the size did not change significantly. I found this more natural and easy to work with. And this is where the difference between the Sigma 105 and the 55-250 showed up.

The Sigma 105 focus ring turns about 120 degrees from closest focus to infinity. It is not stiff, but it is quite firm. In contrast, the 55-250 STM turns about 270 degrees, and can be moved with a much lighter touch. I assume the Sigma 105 has mechanical focusing while the 55-250 is focus-by-wire. I can understand now why many people prefer mechanical focusing. It is very quick to use. Sigma 105 manual focusing has a very positive feel to it. However, the lighter touch and greater rotation of the 55-250 focus ring means that you can make much finer adjustments. This means that you can get a sequence of images for stacking with very small differences in focus positioning. No matter how hard or often I tried, I simply found it impossible to get such small differences with the Sigma 105. Here is the scene where this really came home to me.

Here is a version captured with the 55-250. It used 10 images at f/5.6.


0672 05 2015_01_30 55-250 F5.6 IMG_4688-97 ZS retouched LR 1100h by gardenersassistant, on Flickr

Here is a version captured with the Sigma 105. It used 3 images at f5.6. I did manage to get one version with the Sigma 105 with 4 images, but in several attempts I could only manage 3 like this.


0672 06 2015_01_30 Sigma 105 F5.6 IMG_4675-7 ZS retouched LR 1100h by gardenersassistant, on Flickr

When I looked closely I could see "gaps" in this version.


0672 07 2015_01_30 Focus inconsistency through the stack using the Sigma 105 focus wheel by gardenersassistant, on Flickr

Looking at the left hand image, which is the Sigma 105 version, going from back to front of the scene/top to bottom of the image, there is an in-focus area of moss in the background at the top, then the back dead leaf is out of focus, then further forward is moss in focus beside the nearer dead leaf which is partly in focus, and then more out of focus at the front/bottom.

Looking closely, the 55-250 version I'm not sure that is is perfect - is the middle dead leaf slightly soft? I'm not sure, but the image looks usable to me. I suspect I could use smaller steps between images with the 55-250. I'll be testing that in future stacking attempts.

I also found it more difficult and time-consuming to use Zerene's "retouching" facility to correct errors with the Sigma 105 version such as arose in this scene when the breeze had moved a flower or stem between shots, giving a "ghost image" effect around the edges.

This all tends to reinforce my feeling that the 70D + 55-250 is going to be my "go to" rig for botanical work this year. So where does that leave the Sigma 105, especially as I've discovered that the 55-250 seems to be pretty much as sharp as the Sigma 105, and I'm using the G3 for flash work?

Ah ha, well, there's quite a story to come on that front. Potentially very positive for the Sigma 105, with some positive implications for the 55-250 too. It looks as though it may be the G3 that is at risk now of being put to one side. More in a while - it's proving rather complicated and I'm having to do a lot of detailed work and tests, and organising my thoughts and writing it all up may take a while. So I may go quiet for bit as far as the forum is concerned.

 

[LCPete]

I have been experimenting with stacking for botanical subjects. This has revealed a rather interesting difference between the Sigma 105 and the 55-250 STM.

I read a while ago that one of the advantages of stacking for flower shots was that by stacking large aperture shots you could (depending on the spatial characteristics of the scene) get a nicely out of focus background while still getting all of the subject in focus, in a way that was impossible with a single shot. Here is an example from my experimenting a few days ago.

Here is a single-image version of a scene captured with the 55-250 on the 70D. I wanted to get the flower and the buds in focus, but despite using f/32 for maximum depth of focus, the front-to-back distance of the flower and buds (coupled with filling so much of the frame with the subject) meant that not everything I wanted was in focus.


0672 02 2015_01_30 IMG_4570 LR by gardenersassistant, on Flickr

Here is a 10-image stack of the scene. The individual images used f/10. This was one of several experiments using different apertures for the stack shots. In general I will probably be using larger apertures, but even at f/10 the flower looks sharper and more detailed, and so do the buds, while the background is much more out of focus.


0672 03 2015_01_30 IMG_4583-92 ZS PMax LR by gardenersassistant, on Flickr

However, I noticed while doing these experiments that the image was changing size as I moved the focus rail from back to front to get different parts of the scene in focus. Here is an example I constructed indoors to explore this effect.


0672 04 2015_01_30 Subject size change comparison by gardenersassistant, on Flickr

This change in size doesn't matter to Zerene Stacker, which handles it fine, but it turns out that it does matter to me. I often compose my botanical stuff right to the edge of the screen and the changing size was a bit offputting, complicating matters. I discovered, as illustrated above, that when I moved the focus point with the focus ring rather than the focus rail the size did not change significantly. I found this more natural and easy to work with. And this is where the difference between the Sigma 105 and the 55-250 showed up.

The Sigma 105 focus ring turns about 120 degrees from closest focus to infinity. It is not stiff, but it is quite firm. In contrast, the 55-250 STM turns about 270 degrees, and can be moved with a much lighter touch. I assume the Sigma 105 has mechanical focusing while the 55-250 is focus-by-wire. I can understand now why many people prefer mechanical focusing. It is very quick to use. Sigma 105 manual focusing has a very positive feel to it. However, the lighter touch and greater rotation of the 55-250 focus ring means that you can make much finer adjustments. This means that you can get a sequence of images for stacking with very small differences in focus positioning. No matter how hard or often I tried, I simply found it impossible to get such small differences with the Sigma 105. Here is the scene where this really came home to me.

Here is a version captured with the 55-250. It used 10 images at f/5.6.


0672 05 2015_01_30 55-250 F5.6 IMG_4688-97 ZS retouched LR 1100h by gardenersassistant, on Flickr

Here is a version captured with the Sigma 105. It used 3 images at f5.6. I did manage to get one version with the Sigma 105 with 4 images, but in several attempts I could only manage 3 like this.


0672 06 2015_01_30 Sigma 105 F5.6 IMG_4675-7 ZS retouched LR 1100h by gardenersassistant, on Flickr

When I looked closely I could see "gaps" in this version.


0672 07 2015_01_30 Focus inconsistency through the stack using the Sigma 105 focus wheel by gardenersassistant, on Flickr

Looking at the left hand image, which is the Sigma 105 version, going from back to front of the scene/top to bottom of the image, there is an in-focus area of moss in the background at the top, then the back dead leaf is out of focus, then further forward is moss in focus beside the nearer dead leaf which is partly in focus, and then more out of focus at the front/bottom.

Looking closely, the 55-250 version I'm not sure that is is perfect - is the middle dead leaf slightly soft? I'm not sure, but the image looks usable to me. I suspect I could use smaller steps between images with the 55-250. I'll be testing that in future stacking attempts.

I also found it more difficult and time-consuming to use Zerene's "retouching" facility to correct errors with the Sigma 105 version such as arose in this scene when the breeze had moved a flower or stem between shots, giving a "ghost image" effect around the edges.

This all tends to reinforce my feeling that the 70D + 55-250 is going to be my "go to" rig for botanical work this year. So where does that leave the Sigma 105, especially as I've discovered that the 55-250 seems to be pretty much as sharp as the Sigma 105, and I'm using the G3 for flash work?

Ah ha, well, there's quite a story to come on that front. Potentially very positive for the Sigma 105, with some positive implications for the 55-250 too. It looks as though it may be the G3 that is at risk now of being put to one side. More in a while - it's proving rather complicated and I'm having to do a lot of detailed work and tests, and organising my thoughts and writing it all up may take a while. So I may go quiet for bit as far as the forum is concerned.

Interesting as always

I focus by moving the focus ring so haven't had the problem of the subject changing size when using a focus rail
on my Sigma 105 I find the focus ring ok to use but maybe mine has loosened with age!
When I use my Canon 100L I almost always use auto focus even when stacking I just move the focus point around or move left or right slightly
the 7D is really good in that respect its quick and easy to move the focus point around the subject with the joystick control

 

[GardenersHelper]

Interesting as always

I focus by moving the focus ring so haven't had the problem of the subject changing size when using a focus rail
on my Sigma 105 I find the focus ring ok to use but maybe mine has loosened with age!
When I use my Canon 100L I almost always use auto focus even when stacking I just move the focus point around or move left or right slightly
the 7D is really good in that respect its quick and easy to move the focus point around the subject with the joystick control

Thanks Pete. Interesting that you use autofocus for stacks. I have never thought of trying that.

I'm sorry for not responding sooner but I've been really tied up with the next part of the saga (extension tubes and teleconverters). I'm writing up the results as fast as I can, but I keep finding new angles to investigate and it's all really time-consuming. It will be (even by my standards) a rather large (set of) posts when it arrives. Hopefully some time in the next week or so. Don't know if anyone will want to plough through it all, but writing it is certainly helping me sort my thoughts out. And the experiments are proving really informative (especially as I'm starting from the point of never having used either extension tubes or teleconverters before).

 

[LCPete]

Thanks Pete. Interesting that you use autofocus for stacks. I have never thought of trying that.

I'm sorry for not responding sooner but I've been really tied up with the next part of the saga (extension tubes and teleconverters). I'm writing up the results as fast as I can, but I keep finding new angles to investigate and it's all really time-consuming. It will be (even by my standards) a rather large (set of) posts when it arrives. Hopefully some time in the next week or so. Don't know if anyone will want to plough through it all, but writing it is certainly helping me sort my thoughts out. And the experiments are proving really informative (especially as I'm starting from the point of never having used either extension tubes or teleconverters before).

No worries Nick
I use teleconverters quite often with telephoto lenses normally with the 300 F2.8 when I need a bit more reach
I find that with the Canon 1.4 TC with that lens I cant see any difference in image quality with or without the teleconverter
I haven't used them with macro lenses very much though

 

[GardenersHelper]

A new stage in my journey. Or perhaps a little detour. Three weeks ago I took delivery of some extension tubes and teleconverters. The following posts tell the story of my experiments with them to find out how they work, to compare them with my other kit and to decide what place if any they will have in my kitbag.

Here are links to following posts:

The new equipment
Extension tubes
Effective aperture, loss of light, dof, diffraction and autofocus issues
Teleconverters
Combining teleconverters and extension tubes
Teleconverters and extension tubes with the Sigma 105
Using flash with the Sigma 105
Teleconverters and extension tubes with the 55-250 STM
Grouping the alternatives
Mid to low magnification alternatives
High magnification alternatives
Low magnification alternatives
Image quality
Crop for magnification
Crop for magnification examples
Indoor teleconverter comparisons
Sigma 105 vs 55-250, extension tubes vs achromats
Another look at teleconverter image quality
Sigma 105 vs 55-250, edge sharpness and crop quality
Image quality and kit
Usefulness, usability, and the implications for my toolkit


CLARIFICATION AND APOLOGY ABOUT THE USE OF "3.4x"

In these posts I have used the term "3.4x" as shorthand to refer to the use of a 1.4x magnification teleconverter stacked on a 2x magnification teleconverter. This was a very bad choice. When a 1.4x teleconverter is added on to a 2x teleconverter the total magnification is 2.8x (2 * 1.4 = 2.8). I meant "1.4x and 2x stacked", but "3.4x" obviously invites the standard interpretation of 3.4x magnification.

I have considered changing this but the term is heavily embedded not only in the text of the posts, which wouldn't be too bad to fix, but in the illustrations and in amongst over 1,000 file names each of which is stored/backed up in several places, offline and online, the changing of which presents significant logistical overheads and risks. I think it is best therefore simply to hang my head in shame, ask your forgiveness, and ask you to remember that by "3.4x" I don't mean 3.4x magnification, I mean both teleconverters stacked. Mea culpa. (And thanks to @1Balance at dpreview who pointed this out to me)



Notes

All measurements of scene sizes and working distances are approximate, especially the smallest and largest of them.

All calculations of magnifications and ratios are rounded to the nearest convenient number.

There are lots of numbers in these posts. There are going to be some wrong ones in amongst them. I found and corrected some errors as I went back to check things while writing all this stuff, but I am sure I will have missed some.

I measured working distances and magnifications by capturing images of a ruler. Magnifications are expressed in terms of the 70D's APS-C sensor width of 22.5mm. So a ruler measurement (scene width) of around 22.5mm is counted as 1:1, a scene width of around 11mm as 2:1, and a scene width of around 45mm as 1:2. This is slightly inconsistent for the Panasonic G3 and FZ200 cameras as they have a 4:3 aspect ratio rather than the 70D's 3:2, but I think it is near enough for my purposes.

There are some comparisons of the same scene taken with different setups, including both the 70D and the G3. Having captured an image with one of the setups, I then tried to match the centre, top and the bottom of the scene with the other setups. This means that the G3 images are like the 70D images but with the sides chopped off. (In fact, as you can see from the examples, I could not get the alignments perfect a lot of the time, but I think the images are close enough to illustrate dof and image quality differences etc.)

 

[GardenersHelper]

The new equipment

These are the recent purchases:

• a set of Kenko Teleplus DG AF extension tubes
• a Kenko Teleplus DGX 1.4X Pro-300 teleconverter
• a Kenko Teleplus DGX 2X Pro-300 teleconverter

These are all Canon fit versions, for use with my 70D.

They are potential replacements for some or all of my achromats (Canon 500D, Raynox 150, 250 and MSN-202), or might be used in combination with the achromats on the 70D. Unlike achromats, which are attached to the end of the camera lens, extension tubes and teleconverters fit between the camera and the camera lens.


0681 01 Extension tubes and teleconverters 1 P1890859 LR by gardenersassistant, on Flickr


0681 02 Extension tubes and teleconverters 2 by gardenersassistant, on Flickr

The extension tubes and teleconverters are larger than achromats. At the front here, left to right, are the Canon 500D, the Raynox 150 (the Raynox 250 is the same size) and the Raynox MSN-202.


0681 03 Comparative size of extension tubes and teleconverters P1890861 LR by gardenersassistant, on Flickr



Back to table of contents

 

[GardenersHelper]

Extension tubes

These extension tubes can be used with Canon-fit full frame lenses (“EF” lenses) and crop lenses (“EF-S”) lenses, as indicated by the red (EF) and white (EF-S) alignment markers on the tubes.

The tubes have electrical connections which means that autoexposure and autofocus work as normal. Flash autoexposure also works normally.

Unlike achromats and teleconverters, extension tubes are hollow, with no glass in them to degrade the image. The tubes are fitted between the camera and the lens, and they hold the lens further away from the sensor. This allows the lens to focus on subjects nearer to it, and this provides magnification, but you lose infinity focus.

The amount of magnification depends on how far the lens is held away from the sensor. The three tubes can be used in any combination to provide extensions of 12mm, 20mm, 32mm, 36mm, 48mm and 68mm. The longer the extension, the more the magnification. This graphic shows how large the image is when using a lens with 55mm focal length and focusing as near to the subject as possible using no tubes and using three lengths of tube.


0681 04 Extension tube working distances and magnifications at 55mm focal length by gardenersassistant, on Flickr

This setup provides a lot of magnification, but it also reduces the working distance drastically, from about 640mm when not using any of the tubes to about 25mm when using all the tubes. This is so near that it may be difficult to get enough light on to the subject, especially if using flash. And if the subject is an insect, spider etc, you may frighten it off if you get too close to it.

As well as depending on the length of tube you use, the amount of magnification and the changes in working distance also depend on the focal length of the lens. The smaller the focal length of the lens, the greater the magnification. In this example, with 55mm focal length, the minimum scene width with no tubes used was about 280mm. The minimum scene width with all the tubes used was about 17mm. So the maximum magnification you can get from the tubes is about 16x compared to using the lens by itself.

In contrast, when used with my Sigma 105 macro lens, the maximum magnification using all the tubes is a little under twice what you can get with the lens by itself. Along with this smaller increase in magnification goes a smaller decrease in working distance, going from about 140mm with the bare lens to about 105mm when using all three tubes.


0681 05 68mm extension tubes on 18-55 at 55mm and Sigma 105 by gardenersassistant, on Flickr

The following table illustrates scene widths that can be captured when using the extension tubes on the Sigma 105, along with the associated magnifications and working distances.


0681 06 Magnification and working distances of extension tubes on Sigma 105 by gardenersassistant, on Flickr

When using extension tubes you lose infinity focus, and so for each length of extension there is a range of magnifications/working distances over which you can get a sharp image. As the amount of extension goes up, the range of magnifications/working distances you can use gets smaller.

The Sigma 105 by itself will focus at infinity and so rather than giving minimum and maximum values, the table gives several examples, from 1:4 magnification down to 1:1.



Back to table of contents

 

[GardenersHelper]

Effective aperture, loss of light, dof, diffraction and autofocus issues

Moving the lens further away from the sensor makes the effective aperture smaller than the aperture you set on the camera (the “nominal” aperture). The extension tubes “lose light” so you have to use a slower shutter speed, a higher ISO or increase the aperture set on the camera.

I have not been able to find a realistic (for me at least) way of calculating how much the effective aperture changes for particular combinations of lens and extension. So I tried to measure the loss of light by photographing a sheet of plain paper with various amounts of extension. I did this for the 55-250 at 55mm and at 250mm, and for the Sigma 105. As the finest-grain measurement of shutter speed is only 1/3 stop I didn't get “smooth” results, but the picture is pretty clear in broad terms. Here are the measurements I got.


0681 07 Measured light loss with extension tubes by gardenersassistant, on Flickr

As with other measurements shown later, I didn't use the 32mm (12mm + 20mm) extension as it is so similar to the 36mm extension. For the other five extensions that I used it looks like roughly a 1/3 stop decrease in effective aperture for each increase in the amount of extension, up to a total of about 1 2/3 stops when using all three tubes.

Decreasing the effective aperture has another impact – it increases depth of field. Dof increases by about 40% for a 1 stop decrease in aperture and doubles with a 2 stop decrease in aperture. When using all three tubes on either of these lenses there is an almost two stop reduction in effective aperture, and so an almost doubling of dof. Especially when starting from a smallish aperture, when you add tubes and the effective aperture decreases there is a corresponding increase in image degradation from diffraction, which causes loss of sharpness/detail/microcontrast.

When using all three tubes, increasing the aperture by 1 2/3 stops will make the shutter speed revert to what it would have been without the tubes, and make the dof and diffraction losses revert to where they would have been without the tubes. So, in terms of shutter speed, ISO, dof and diffraction losses, using tubes appears to be neutral if you increase the aperture appropriately, depending on the amount of extension being used.

This is fine for close-ups/macros, assuming you are using a middling to small aperture. If you are using a large aperture then you may not have enough “slack” in the aperture to be able to increase it enough to balance out the effective aperture affects. For example, if you were using f/16 for a shot and then put all three extension tubes on the camera you would need to increase the aperture by 1 2/3 stops to f/9. That is ok. But if you were using an f/2.8 lens at f/4 and then put all three tubes on, increasing the aperture by 1 2/3 stops would mean going to f/2.2, which isn't possible with an f/2.8 lens.

A reduction in effective aperture won't be neutral for dSLR users using the optical viewfinder. As the effective aperture decreases the viewfinder gets dimmer, which can make life difficult, both in terms of locating the subject and getting it in focus manually. Also, autofocus may not work. For example, if you have a lens like the 55-250 STM which is f/4 to f/5.6, using all the tubes would take it to almost f/8 to f/11, at which point autofocus may not work with the viewfinder. It doesn't on my 55-250 STM.



Back to table of contents

 

[GardenersHelper]

Teleconverters

Teleconverters do have glass in them and so, like achromats, they degrade image quality. I'll look at the question of image quality in later posts, but for now I'll just say that based on the (real world, and quite carefully executed) comparisons I've done so far, it looks to me like image quality can be acceptable sometimes, and sometimes not, depending on circumstances, preferences, output media etc.

Teleconverters increase the focal length of a lens. These teleconverters can connect to full size (EF) lenses, but not crop (EF-S) lenses. (But that doesn't mean they can't be used with crop lenses. See next post.)

A 1.4x teleconverter increases the focal length by 40%, a 2x teleconverter doubles the focal length. Like extension tubes, teleconverters decrease the effective aperture. A 1.4x teleconverter reduces the effective aperture by 1 stop, a 2x by 2 stops. So, with a 1.4x teleconverter my Sigma 105mm, which goes from f/2.8 to f/22, becomes 147mm and goes from f/4 to f/32. With a 2x teleconverter it becomes 210mm, f/5.6 to f/45.

As with extension tubes, teleconverters “lose light” so you have to use a slower shutter speed, a higher ISO or increase the aperture. Along with the reduction in effective aperture you get an increase in dof and diffraction losses, and possible viewfinder and viewfinder autofocus issues. As with extension tubes, increasing the nominal aperture by one or two stops should neutralise the shutter speed, dof and diffraction effects, but it won't help with any viewfinder issues.

Where teleconverters differ from extension tubes is that they retain infinity focus, which may not matter much to those primarily interested in macro/closeups, but they also do not increase the minimum working distance much or at all, which I believe is their primary attraction for macro/closeups. What this means is that you can get greater magnification with the same working distance, or the same magnification from further away.

For example, my Sigma 105 gives a maximum magnification of 1:1 (a scene width of 22.5mm) at a working distance of about 130 mm. Using the Sigma 105 with the 1.4x teleconverter gives me a maximum magnification of about 1.6:1 (scene width about 14mm) at a working distance of about 135mm, and using it with a 2x teleconverter gives maximum magnification of 2:1 (scene width 11mm), also at a working distance of about 135mm.

Alternatively, using the 1.4x teleconverter with the Sigma 105 I can get 1:1 magnification at a working distance of about 170mm compared to about 130mm with the Sigma 105 by itself. With the 2x teleconverter with the Sigma 105 I can get 1:1 magnification at a working distance of about 225mm.

Like the extension tubes, these teleconverters have electrical connections which mean that autoexposure and autofocus work as normal. These connections are a bit “smarter” than those on the tubes. Unlike with the extension tubes, the teleconverters report the effective aperture to the camera, so for example when using minimum aperture on the Sigma 105 with the 2x teleconverter what I see on the LCD and in the viewfinder is f/45 rather than f/22, and these revised figures turn up in the Exif data too. The Exif data also contains the effective focal length too, which would be 210mm when using the 2x teleconverter rather than 105mm.



Back to table of contents

 

[GardenersHelper]

Combining teleconverters and extension tubes

Important note: These are direct quotations from the documentation that came with the teleconverters:

“... Do not attach two or more TELEPLUS simultaneously....”
“... Other converters or extension tubes cannot be combined...”

You have been warned!

The teleconverters fit together mechanically, and fit together with the extension tubes, and they all work together electrically. Be aware though that doing this may (I assume, given what Kenko say about it) cause damage. That wouldn't entirely surprise me, because with a heavy lens on the end of several linked tubes/converters, the stresses on the joints may be considerable, and even more so perhaps the stress on the thread in the camera where the tripod is attached. In my case, because of the strange tripod I use, there might also be sufficient rotational force on the tripod arm to make it give way and rotate, which would probably make the rig topple over.

For testing purposes I have briefly used the Sigma 105, both converters and all three tubes at the same time, with the camera mounted on the tripod, but I felt quite nervous about it.



Back to table of contents

 

[GardenersHelper]

Teleconverters and extension tubes with the Sigma 105

There are a lot of possible ways to combine the teleconverters and extension tubes, depending on which ones you use and which order you attach them in. You can use any combination of them in any order with the Sigma 105. I don't know if this is true for all EF lenses with these particular extension tubes and converters, and/or for other makes of tubes and converters. For example, Canon extension tubes only work with some Canon lenses, and Canon teleconverters only work with some lenses.

Here are several of the possible combinations with the Sigma 105 on the 70D, compared with the achromat setups I use on the G3 to obtain similar magnifications.


0681 08 Working distances, scene widths and magnifications 70D and G3 by gardenersassistant, on Flickr

As you can see, the 70D setups are much larger, and as you can imagine, they are much heavier in the hand and exert much greater strains on the tripod arm and joints. This makes them physically more taxing to use.

The G3 setups can all use zoom in/out from a fixed position (using a quick, gentle touch from the index finger on the zoom ring) to help locate subjects and frame compositions. This is not possible with any 70D setup using a prime lens like the Sigma 105, and this adds to the effort and awkwardness of using tubes or teleconverters (and the Sigma 105 by itself for that matter).

On the other hand the 70D setups have significantly larger (and more practical to work with) working distances at higher magnifications than the G3 setups. In addition, the 70D setups can deal with a much wider range of magnifications than the equivalent G3 setup. This would reduce the number of equipment changes needed to deal with subjects of different sizes.

In attempting to get a handle on the options I did a lot of measurements. Here is what I measured for the Sigma 105. It is rather indigestible in this form, and subsequent posts present a simplified view of just the combinations that I thought might be most useful for me at the moment. However, there may be bits of this data that are of interest to one or two people so I thought I would include it.


0681 09 Scene sizes, working distances and magnifications of combinations using the Sigma 105 by gardenersassistant, on Flickr

As you can see from this table, you can add achromats to these combinations of extension tubes and teleconverters.

You can also see, from the bottom part of the table, that when using a combination of tube(s) and converter(s), what magnifications and working distances you get depends on what order you mount the tube(s) and converter(s).



Back to table of contents

 

[GardenersHelper]

Using flash with the Sigma 105

My current flash arrangement works with all but one of the combinations of extension tube(s), teleconverter(s) and achromat(s) that I have tried. Only with the Raynox 150 and 250 stacked on the end of all the extension tubes and both of the teleconverters does the flash light get blocked from the subject. In all other combinations I tried flash light reached the subject ok.

However, even though I could get the light on to the subject, quite often with the Sigma setups I could not get enough light on to the subject. Although the minimum working distances are fine with all the setups, with some of the setups you have to move quite a long way from the subject to decrease the magnification through the entire ranges shown in the above table. This means that using my current flash setup I can only use the near end of some of the ranges. The G3 does not have this problem as its working distances never exceed 500mm or so, and generally for flash work its working distances are under 220mm as most of my flash work is done with the Raynox 150, 250 and MSN-202 rather than the longer working distance Canon 500D.

The G3 does share another flash issue with the 70D setups, which I think arises from the small apertures I use a lot. Although my flash units are quite powerful, they can't always provide enough illumination for me to work at base ISO. Sometimes, during some recent testing, I had to go to ISO 1600 (and possibly higher, I don't recall) to get enough light on to the subject. There are two issues with this. One is the increase in noise and loss of detail/clarity/subtle colour rendition at higher ISOs. The other is the somewhat hit and miss nature of exposures and the need to adjust ISO to get to a suitable level of illumination. Especially with jumpy subjects this is not good.

The illumination issues were sufficiently troubling for me to get out the FZ200 again, used like the G3 with achromats. A short session with the FZ200 reminded me of how well it works with flash, including working at base ISO at all magnifications I can use with my achromats (roughly 1:9 down to 7:1).



Back to table of contents

 

[GardenersHelper]

Teleconverters and extension tubes with the 55-250 STM

As mentioned above, you can't attach a crop factor EF-S lens to one of these teleconverters. However, you can attach an extension tube to an EF-S lens and then attach a teleconverter to the extension tube. This gives almost as many combinations of tubes and converters as with EF lenses. For example, here is the 55-250 STM with the 12mm extension tube attached to it, and the 2x teleconverter attached to the extension tube.


0681 10 70D + 55-250 + 2x + 12mm P1890949 by gardenersassistant, on Flickr

The 55-250 is much lighter than the Sigma 105, and so using combinations of tubes and teleconverters places less strain on the tube/converter/camera connections, the camera's tripod attachment thread and the various joints of the tripod. And of course the setups are lighter for hand-held use too.

Another advantage of the 55-250 is that you can zoom with it. This provides more flexibility for composition than the Sigma, for example allowing some adjustments of magnification/framing without moving the camera.

A big disadvantage of the 55-250 is that it cannot focus nearly as closely as the Sigma. This rules out many of the higher magnification options available with the Sigma 105.

However, there are a couple of lower magnification issues for which the 55-250 + teleconverter may be very useful. One is the “butterfly problem”. None of my camera/lens combinations to date have let me fill enough of the frame for my taste with a small butterfly from a large enough distance. For example, with the 55-250 by itself on the 7D the minimum scene width is about 75mm at a working distance of about 625mm. The working distance is ok (ish, although I'd prefer a bit more), but the scene size is too large. With the 500D on the G3 + 45-175 I can get down to a scene width of about 35mm at a working distance of about 210mm, and with the 500D on the 70D + 55-250 I can get down to about 33mm at about 270mm. The scene widths are ok, but the working distance is too small. It is similar with the FZ200.

With the 2x teleconverter on the 55-250 (with the 12mm extension tube, needed to connect the lens to the converter), I can get a scene width of about 35mm at a working distance of about 655mm. That could be very useful.

The other lower magnification issue is insufficient reach. The minimum scene width of the 55-250 by itself is about 75mm at a working distance of about 625mm. I do encounter situations in the garden where there is a flower which I can't get that close to, and I can't get the framing I want. With the 2x teleconverter on the 55-250 I can get a scene width of 75mm from about 1600mm. If that is too far, I can reduce the working distance to anything down to about 425mm by reducing the focal length down as far as 100mm, still retaining a scene width of 75mm.

Here is a practical example. At the top is the nearest I could get to this flower with the bare 55-250. Underneath is what I could get from the same position with the 2x teleconverter on the 55-250.


0681 11 2x teleconverter reach example by gardenersassistant, on Flickr

The “butterfly problem” illustrates the teleconverter benefit of getting larger magnification at a given distance. The flower example illustrates the benefit of getting the same magnification at a larger distance. If the quality from the converters proved to be acceptable these two benefits would be worth having (for me).

A disadvantage of the 55-250 is that it is f/4-f/5.6 rather than f/2.8 for the Sigma 105. That hasn't mattered much to me to date as I have been using the 55-250 mainly for flowers and I find f/4-f/5.6 enough for that. However, with the reduction in effective aperture caused by tubes and converters this could be problematic, especially as I have recently been getting more interested in stacking botanical shots, and in order to get good sharpness in stacks I don't want the aperture to drift far above f/5.6.

As with the Sigma 105, the following tables provide some measurements I did while working through the options. As with the Sigma 105 measurements, a selection of these are used in the following sections when comparing the alternatives that seem most appropriate to my needs.


0681 12 55-250 combinations working distance and magnifications by gardenersassistant, on Flickr


0681 13 55-250 with 1.4x teleconverter by gardenersassistant, on Flickr



Back to table of contents

 

[GardenersHelper]

Grouping the alternatives

There are a lot of options. I need to decide if any of them is going to help me create better photos. Perhaps none of them will. At the other extreme, perhaps extension tubes and teleconverters will let me move to a simpler, single-camera solution. But as the tables of data grew larger it was looking more and more complicated to sort out.

I pored over the figures in the tables and as well as doing more measurements I went out into the garden a few times to try various combinations. There were two things I was primarily interested in:

Could I get better image quality by using extension tubes and/or teleconverters?
What were they like to use – Practical? Enjoyable? Creative?

As the previous tables illustrate, there are a lot of alternatives to consider. To make the task manageable I decided to look at the alternatives from the point of view of my current approach, using the G3 with achromats as my baseline against which to compare the usefulness or otherwise of extension tubes and teleconverters on the 70D. As far as the measurements are concerned, I decided to break down the comparisons into three groups:

• Low to mid magnification. The baseline here is the G3 with its 45-175 lens and the Raynox 150, Raynox 250 and Canon 500D. This is the kit I have been using most for invertebrates lately. This typically involves magnifications on either side of 1:1 in APS-C terms. At these magnifications I may or may not use flash, depending on the circumstances.
  
  
• High magnification. The baseline here is the G3 + 45-175 with the Raynox 150 and 250 stacked, and with the Raynox MSN-202. This is the kit I have been using recently for small subjects such as springtails, barkflies and fruit flies. Magnification goes down to 5:1 or so, and this work always involves the use of flash.
  
  
• Low magnification. The baseline here is the G3 with the bare 45-175 and with the Canon 500D on the 45-175. This is a setup kit I have used a lot for botanical subjects. Magnifications are typically much less than 1:1 and I almost always work with natural light, with only very occasional use of flash for fill. It is also the kit I use for butterflies, but the working distance is a bit too small when using the Canon 500D on the G3 and the magnification is a bit too small when using the bare lens. The situation is similar with the FZ200 and the 55-250 on the 70D.

Back to table of contents

 

[GardenersHelper]

Mid to low magnification alternatives

Looking at the numbers it seemed to me that for the 70D my best options might be to use the Sigma 105 with the 1.4 teleconverter and either 20mm or 36mm extension tube. With the 20mm extension this would cover the whole range covered by the Raynox 250, Raynox 150 and Canon 500D on the G3. This looked very appealing. With the 36mm extension the magnification would range from about half the maximum scene width of the Canon 500D on the G3 to a bit more magnification than the Raynox 250 on the G3; another very appealing combination.

In the this table the colour coding links setups covering the same scene width to make it easier to compare the working distances of the various setups for that particular scene width. The colour coded scene widths are the minimum and maximum for the Raynox 150, Raynox 250 and Canon 500D on the G3.


0681 14 Low to mid magnification alternatives by gardenersassistant, on Flickr

The ranges of magnification covered by the FZ200 and G3 with the three achromats is broadly similar, with the FZ200 tending towards smaller scene widths for each of the achromats. With both cameras the range of magnifications covered by each achromat is quite small, and so switching between subjects/scenes of different size often involves swapping between achromats. This is what made the large range of scene widths offered by a single setup with a teleconverter and extension tube so attractive to me.



Back to table of contents

 

[GardenersHelper]

High magnification alternatives

The extension tubes and teleconverters seemed less appealing for high magnification. Getting down to 5:1 would mean either using the 2x teleconverter or even the 2x and 1.4x stacked, or alternatively using an achromat in addition to one of the teleconverters (and some extension too). I was not at all keen on using both teleconverters stacked, as my initial image quality tests suggested that this produced rather soft images, and I was not convinced that 2x would be good enough either. Whatever combination I used for high magnification, the setup was going to be rather lengthy and unwieldy.

On the other hand, the working distances were much better for setups using extension tubes and teleconverters without the addition of an achromat. The working distances with the G3 and the FZ200 are very short at high magnification.


0681 15 High magnification alternatives by gardenersassistant, on Flickr



Back to table of contents

 

[GardenersHelper]

Low magnification alternatives

The Sigma 105 can provide suitable magnification for butterflies and similar sized subjects, but the working distance is too small for jumpy subjects. This made the use of one of the teleconverters on the 55-250 look very attractive, as it seemed to provide both sufficient magnification and sufficient working distance.

I decided to concentrate on testing the 2x teleconverter with the 55-250, despite my unresolved questions about image quality with the 2x. This was because I was thinking that it would simplify changing lenses, tubes and converters if I kept the 1.4x converter attached to the 36mm extension tube for use with the Sigma 105 and kept the 2x converter attached to the 12mm extension tube for use with the 55-250.

If image quality turned out to be inadequate with the 2x converter on the 55-250 then the magnification and working distance of the 1.4x would probably be sufficient.


0681 16 Low magnification alternatives by gardenersassistant, on Flickr



Back to table of contents

 

[GardenersHelper]

Image quality

To get some data about the effect of using teleconverters on image quality I did some comparisons, in as controlled and consistent manner as I could manage. I used outdoor scenes of varying size, and captured each with shots at various apertures for each of several setups. I have only done 10 of these so far because it is a surprisingly difficult and time-consuming exercise.

Here are the 10 scenes I used. I have spent some time poring over full size versions of these. There are 1100 pixel high versions of all of the (237) test captures here at Flickr. If anyone would like to see full size versions of some of them (or any of the other test images mentioned later) I'll be happy to make them available.


0681 17 10 scenes used to compare image quality using teleconverters by gardenersassistant, on Flickr

There was a slight, intermittent breeze for some of the scenes, but I tried to pick scenes that were not affected much by any breeze, and when there was any air movement I tried to pick moments when it had abated. As far as I can tell the results were not affected significantly by breeze issues.

I used a tripod and remote release to capture the test shots, and attempted to keep the centre of focus in the same place for all the shots of each scene. However, when changing setups I had to move the camera, not just to a different distance from the subject but often to a different height above ground (not helped by the fact that the ground was sloping).

While moving the camera between setups I sometimes didn't get the distance to the subject right and so the magnification between setups was slightly different. I sometimes got some sideways movement and/or rotational movements that changed the angle on the scene. In addition, and especially with the higher magnification comparisons, I did not always get the centre of focus in precisely the same position for each setup. I had lots of attempts with each scene to get the different setups aligned the same but eventually gave up and went for a “good enough” approach.

This all meant that particular parts of the scene were in/out of focus to different extents for reasons that had nothing to do with the optical qualities of the setups. These issues make the interpretation of the comparison images quite complicated, and the results can be very misleading if you concentrate on just one area of a particular scene. Here is an example. It compares 100% crops from four versions of a scene, all captured with the 70D and Sigma 105 Macro:

• Top left: the bare 105,
• Top right: 105 + 1.4x extension tube
• Bottom left: 105 + 2x extension tube
• Bottom right 105 + 1.4x and 2x extension tubes stacked

0681 18 Alignment issues by gardenersassistant, on Flickr

All four captures are f/11. This is the (correct) effective aperture as reported in the Exif data (correct because each teleconverter alters the Exif data to take account of the fact it is being used).

This was the first comparison I captured and examined. My immediate impression, flicking quickly between the full size images, was that the version using both teleconverters (3.4x) was extremely soft compared to the others, which was no surprise. But I then noticed that the 2x version seemed better than the 1.x version, which surely couldn't be right. This led me to look more closely and to conclude (as can be seen from the 100% crops above) that the 1.4x version and the 3.4x versions were focused nearer to the camera than the bare and 2x versions.

So I then compared the bare version to the 2x version, and the bare version seemed to have better sharpness, detail and microcontrast wherever I looked, both in front and behind the area of best focus, as illustrated by the following 100% crop comparison. This was as expected.


0681 19 100pc comparison bare 105 vs 105 + 2x by gardenersassistant, on Flickr

Similarly, when I compared the 1.4x version to the 3.4x version, the 1.4x version seemed to have better sharpness, detail and microcontrast wherever I looked, both in front and behind the area of best focus. This too was as expected.


0681 20 100pc comparison 105 + 1.4x vs 105 + 3.4x by gardenersassistant, on Flickr

Comparing the same area as between the 1.4x and 2x versions did not produce such a clear conclusion. They look rather similar, although the 1.4x version looks slightly better to me. But how should I take into account the fact that the 1.4x is focused nearer to the camera, and the effects of the different vertical angle of the two shots? I couldn't pull those factors apart, and I would not want to draw any conclusion from this particular comparison.


0681 21 100pc comparison 105 + 1.4x vs 105 + 2x by gardenersassistant, on Flickr

This sort of issue plagued all the comparisons. Overall, it seems that, as you would expect, bare lenses give better sharpness/detail/microcontrast at a particular aperture than with a teleconverter added to the lens, and that the results get worse as the power of the teleconversion increases.

But what practical difference would it make? I couldn't really tell. My impression was that the difference between a bare lens and adding the 1.4x teleconverter would probably be hardly or not at all noticeable when looking at images of the size I produce (mainly on screen 1100 pixels high and occasionally printed, usually A4). So I would probably be comfortable using the 1.4x teleconverter. On the other hand, the combination of 1.4x and 2x seemed capable of producing results that might well appear soft at my output sizes. Suspecting this was the case, I would be uncomfortable stacking the teleconverters. And the 2x? Somewhere in between. Would I want to use it? Don't know.

A lot of work. Not much by way of answers.



Back to table of contents

 

[GardenersHelper]

Crop for Magnification

I then thought of a different angle on the image quality problem. I wondered whether it would be possible to do just as well by cropping an image captured with a bare lens compared to using a teleconverter.

I used an indoor scene to test this. The scene wasn't very small, but it had the great advantage that nothing was going to move in the breeze, and the ground was solid, dry (did I mention the mud outdoors?) and flat.

I captured the scene with both teleconverters stacked on the Sigma 105, and then with the camera in the same position captured (progressively larger) scenes using the 2x, then the 1.4x and finally the bare Sigma 105. Unfortunately the teleconverters pushed the rather heavy Sigma 105 away from the point of attachment of the camera to the tripod, increasing the turning moment and causing the rig to “droop” on the tripod. As I moved through the sequence and the 105 came nearer to the camera, the turning moment decreased and the camera rotated up a bit, enough to require adjusting the ball head, which got me back to alignment issues.)

For each setup, I captured the scene using all the available apertures:

• f/2.8 to f/22 for the bare Sigma 105
• f/4 to f/32 with the 1.4x teleconverter
• f/5.6 to f/45 with the 2x
• f/8 to f/64 with the 1.4x and 2x stacked (3.4x)

I then did a very similar exercise with the 55-250 at 250mm focal length. I had to use the 12mm extension tube to connect the teleconverters to the lens and as a result I had to move closer to the scene to be able to gain focus with the teleconverters. I had wanted to do a Sigma 105 to 55-250 comparison, but in practice it seemed too complicated to work out how to do it. In this case the apertures were:

• f/5.6 to f/32 for the bare 55-250
• f/8 to f/45 with the 1.4x teleconverter
• f/11 to f/64 with the 2x
• f/16 to f/64 with 3.4x (It seems that I omitted to do f/90. Subsequent investigation showed that the 55-250 is a bit quirky when it comes to setting the minimum aperture at large focal lengths. This is a characteristic of the lens and has nothing to do with using extension tubes or teleconverters.)

In Lightroom I did a number of crops.

• The bare lens versions cropped to show the same subject area as the 1.4x version
• The bare lens versions cropped to show the same subject area as the 2x versions
• The bare lens versions cropped to show the same subject area as the 3.4x versions
• The 1.4x versions cropped to show the same subject area as the 2x versions
• The 1.4x versions cropped to show the same subject area as the 3.4x versions
• The 2x versions cropped to show the same subject area as the 3.4x versions

I also had uncropped versions of the bare lens, 1.4x, 2x and 3.4x captures.

I then down sized the images to the height of the most severe crops so I could compare images of the same size. Specifically:

• The uncropped 1.4x images have been downsized to the pixel height of the bare lens images as cropped to match the 1.4x image scene coverage.
• The uncropped 2x images and the cropped 1.4x images have been downsized to the pixel height of the bare lens images as cropped to match the 2x image scene coverage.
• The uncropped 3.4x images and the cropped 2x and 1.4x images have been downsized to the pixel height of the bare lens images as cropped to match the 3.4x image scene coverage.

This resulted in some quite small images, particularly the images that were cropped to match the coverage of the 3.4x capture. In this case the images were only 1139 pixels high. The images cropped to 2x capture coverage were 1635 pixels high. The images cropped to 1.4x capture coverage were 2323 pixels high.

These crops can all be seen here at Flickr.

Here are examples of these crops as applied to the Sigma 105 images.


0681 22 Magnification crops by gardenersassistant, on Flickr

I could then make comparisons like this.


0681 23 Magnification crops at a common aperture by gardenersassistant, on Flickr

At the bottom right this shows the whole image captured with both teleconverters stacked on the Sigma 105. At the bottom left it show the same coverage produced by cropping the image captured with the 2x. At the top right is the cropped 1.4x version, and at the top left the cropped bare lens version. All four captures used f/11.

Looking more closely at these, we can see that the cropped versions are noisier, with the noise increasing along with the amount of cropping. (The originals were captured at ISO 400.)


0681 24 Magnification crops at a common aperture - different dof by gardenersassistant, on Flickr

The images were all focused on the front of the box that the duck is sitting on. The ruler was quite a long way in front of that. We can see here that although each of the images was captured at f/11, the cropped versions have greater dof – the greater the crop then the greater the dof. (This is what I have described previously as “crop for dof”.)

If we want to do a like for like comparison we need to take the dof difference into account. The following composite shows versions which are each one stop of aperture larger than the next. The apertures are:

• f/11 for the bare Sigma 105
• f/16 for the 1.4x version
• f/22 for the 2x version
• f/32 for the 3.4x version

0681 25 Magnification crops at equivalent apertures - common dof by gardenersassistant, on Flickr

Now the images have the same dof. We can see that the 3.4x version is better than the cropped bare lens and 1.4x versions. There does not seem to be much in it between the 3.4x version and the cropped 2x version.

Here is another way of looking at cropping for dof/magnification.


0681 25a Decreasing aperture vs cropping to increase depth of field by gardenersassistant, on Flickr

On the left are four images captured using the 3.4x teleconverter, at apertures of f/8, f/11, f/16 and f/22. On the right hand hand side of the images there are background and foreground elements that come more into focus as the aperture decreases and the dof increases.

On the right are four images captured from the same position, all using the same aperture of f/8. The top one is the same as the image on the top left, captured using the 3.4x teleconverter. The next one down was captured using the 2x teleconverter, and then cropped to match the 3.4x image. Next down an image captured with the 1.4 x teleconverter is given a larger crop so as to match the 3.4x image. And at the bottom right an image captured with the bare Sigma 105 is given an even larger crop to match the 3.4x image.

The dof changes in exactly the same way by increasing the crop as it does by decreasing the aperture.

One other thing to note is that shutter speed stays constant for the crops, but increases for the aperture changes. This is not necessarily the advantage for the cropping approach that it might at first appear to be. For hand held shooting, as the magnification increases the effect of handshake is magnified, which would tend to cancel out the advantage of using a faster shutter speed. On the other hand, a faster shutter speed could sometimes help reduce the impact of subject movement.



Back to table of contents

 

[GardenersHelper]

Crop for magnification examples

Here is a comparison of part of the images captured by the 55-250 at f/5.6 (bare lens), f/8 (1.4x), f/11 (2x) and f/16 (3.4x). The 3.4x version is the best of these.


0681 26 Magnification crops at equivalent apertures 1.1 - different noise, detail, microcontrast by gardenersassistant, on Flickr

Here is the same comparison, but with all the apertures decreased by one stop (f/8, f/11, f/16, f/22). I think the 3.4x version is very slightly better than the cropped 2x version, but not much. Both of these are much better than the bigger crops.


0681 27 Magnification crops at equivalent apertures 1.2 - different noise, detail, microcontrast by gardenersassistant, on Flickr

Here is the same comparison shifted by another stop (f/11, f/16, f/22, f/32). This time the cropped 2x version is the best one. Presumably what is happening is that each time the apertures are decreased the loss of detail/sharpness/microcontrast caused by diffraction is increasing disproportionately at the smallest aperture.


0681 28 Magnification crops at equivalent apertures 1.3 - different noise, detail, microcontrast by gardenersassistant, on Flickr

We can see a similar pattern with the Sigma 105 example, although I think it is confused a bit in this case by the 2x version apparently being focused more towards the back of the scene. All the same, in this first comparison with apertures of f/2.8, f/4, f/5.6 and f/8, the 3.4x is clearly better than the bare lens and 1.4x crops. Because of the back-focusing of the 2x version the comparison with the 3.4x version is difficult to call.


0681 29 Magnification crops at equivalent apertures 2.1 - different noise, detail, microcontrast by gardenersassistant, on Flickr

However, shifted by two stops to f/5.6, f/8, f/11 and f/16 the 2x version is now definitely better than the 3.4x version – it has better sharpness/detail/microcontrast nearest and furthest from the camera, and everywhere in between.


0681 30 Magnification crops at equivalent apertures 2.3 - different noise, detail, microcontrast by gardenersassistant, on Flickr

The same pattern is visible with the crops from bare lens and 1.4x to 2x, and from bare lens to 1.4x.

So it looks as though, for “like for like” images with the same dof, a “1-stop crop” (cropping the bare lens to 1.4x coverage, 1.4x to 2x or 2x to 3.4x) may be better or worse than the uncropped teleconverter version, depending on the apertures involved. From this exercise it also looked as though teleconverter versions are always better than 2-stop or 3-stop crops, but some more detailed experiments showed this is not always the case. This is discussed further in this section.

Setting apertures and dof aside (i.e. concentrating on the most in-focus areas), is the best of the 1-stop crops better in terms of sharpness, clarity and microcontrast than the best of the uncropped teleconverter versions?

For these examples, for the 55-250, the best version appears to be an uncropped teleconverter version.

For the Sigma 105 the best of the 1-stop crops looks to be neck and neck with the uncropped teleconverter version, with a bit more detail balanced out by a bit more noise. With a lower ISO than the ISO 400 used here I think the best version might be a 1-stop crop. Conversely, with a higher ISO I suspect an uncropped teleconverter version would be best. This issue is examined in more detail in this section.

Overall, it does look as though these teleconverters have potential for providing increased reach for botanical subjects that I can't get near enough to at the moment. As to whether the quality proves adequate will only become clearer through practical use. Over the past few days I have found both the 1.4x and 2x teleconverters useful to get at a number of awkwardly positioned subjects, but I haven't examined the results yet.



Back to table of contents