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

[GardenersHelper]

infact you single handed prove that you don't need expensive kit which is what a lot of newbies think.

Thanks Bryn.

if I used that kit I doubt very much I would get even half as good images as you do Nick.

But surely that can't be right. If you used that kit, on the subjects I deal with (and you used the same sort of post processing), I'd expect you to get similar results. For example, In terms of image capture all I did with the springtail and the fly was to point the camera at them and shoot. I was concentrating almost exclusively on just getting them into focus, especially with the springtail, and also (for the springtail) not having the subject too large in the frame because I knew I was going to crop for dof.

More generally though, the whole question of "kit or photographer" is a tricky one I think, and I do think that kit tends to be over-rated. I don't remember if it was here or on another site, so forgive me if this is repetition, but one of the things I sometimes do when this question arises, on line or in the tiny photo groups I run, is to show people this set of images of flowers, and this set of images of invertebrates. I ask them to tell me (without looking at the EXIF data if they are on line) which images they think used which camera, or at least which type of camera, or a the very least a small sensor or a larger sensor. For the invertebrate images (and I tell them this) there are two images each from three small sensor bridge cameras of varying age and number of pixels, a micro four thirds camera and a dSLR. For the flower images there are two images from each of these and also two from a point and shoot (ultrazoom travel) camera.

Basically, at the size they are processed at (1100 pixels high) and with them having had similar post processing, no one can tell which is which. (Well not yet anyway.) One person (on line) did say he knew which was which, but when I said I was very interested to know what the give-away features were that had enabled him to tell them apart (and I was genuinely interested to know), he didn't reply.

On the other hand, as you know well, some equipment issues can have a huge impact, most particularly of course given our interests, how flash is used (if/when it is used). Sensor size makes a difference too - if you want to let people look at big versions of your images without you getting embarrassed, then you do need to use a large sensor. I can get away with a smaller sensor only because I don't let people pixel peep my images.

So I sit on the fence on this one. There are many factors that go into making a nice looking photo (assuming for the moment that nice looking is what we're interested in, however we variously interpret that, rather than say interest, shock value or documentary value). And I do think kit is one of them. But only one factor, in a complicated mix of other considerations (which is pretty much what this thread is about of course ).

 

[TimmyG]

And your images are far cleaner than mine. I saw that today when I looked at the very big version of the Young Woodlouse image.

Just remember if checking my images for noise, it's best to check the shadowy areas on the subject. I apply not-inconsiderable noise reduction on the background (usually if it's needed or not). Also, it's best not to compare to stacks, as noise can often be multiplied as the images get merged. I'll try to shoot some single images for comparison when I can get some time to experiment.

Thanks for the diffuser reminder. It's a similar design to my old ice-cream box diffuser but with a bit more advanced diffusion layers. I'm still trying to get a concave diffuser setup I am happy with, but I may be better going back to this style.

 

[GardenersHelper]

Just remember if checking my images for noise, it's best to check the shadowy areas on the subject. I apply not-inconsiderable noise reduction on the background (usually if it's needed or not).

Ah, understood. So we are taking the same approach in the sense zapping backgrounds. I hadn't realised you were doing that too. The difference is that most of the time at the moment - all my recent image posts for example - I haven't done any luminance NR at all. But when I do use luminance NR, it is only on the background (except for very, very rarely on the subject, and then as little as I feel I can get away with).

I'm working through some images from today (I went out to a local wood - lots of leaf litter - you know what I was looking for ), and for some reason the camera was on ISO 800 at the start of the session. Now, when I used the G3 before (for a two year period, before getting the FZ200 and then the 70D), I got reasonably comfortable with ISO 800. But I was shocked today at how horrible it looked to me. I don't recall, I've been through so many permutations and combinations, and I have a memory like the proverbial sieve, but I think I must have been routinely zapping the backgrounds. This has reminded me of how much better I thought the 70D was than the G3 at higher ISOs when I first got the 70D. Given that I've got used to using up to ISO 3200 with the 70D in some circumstances (e.g. early morning natural light snail in motion images, albeit with background zapping), I think that squelches the idea of a simple, single-camera solution with the G3. I find the 70D a terrific natural light camera. I don't think I'll be forsaking it for the G3 any time soon. (But a newer MFT .... stop it, stop it! enough )

Also, it's best not to compare to stacks, as noise can often be multiplied as the images get merged. I'll try to shoot some single images for comparison when I can get some time to experiment.

Ah, good point. I forgot about that.

Thanks for the diffuser reminder. It's a similar design to my old ice-cream box diffuser but with a bit more advanced diffusion layers. I'm still trying to get a concave diffuser setup I am happy with, but I may be better going back to this style.

The second layer, and/or the material from the purchased diffuser (don't know which it was, or perhaps both) seemed to make a big difference compared to what I was using before. Mind you, the light-emitting area is somewhat larger too, so that may be contributing to the effect.

 

[TimmyG]

So we are taking the same approach in the sense zapping backgrounds. I hadn't realised you were doing that too. The difference is that most of the time at the moment - all my recent image posts for example - I haven't done any luminance NR at all. But when I do use luminance NR, it is only on the background (except for very, very rarely on the subject, and then as little as I feel I can get away with).

An alternate method of "zapping" backgrounds I use is a slight gaussian blur across the frame, and then mask out the subject (which I do with NR also). This should probably be referred to a "nuking" the background, I think the effect is much stronger than traditional noise reduction.]

The second layer, and/or the material from the purchased diffuser (don't know which it was, or perhaps both) seemed to make a big difference compared to what I was using before. Mind you, the light-emitting area is somewhat larger too, so that may be contributing to the effect.

Yeah I think one of the main problems I am having is i am not getting the apparent size large enough. I think I may have the flash a bit too close to the subject and I need to move it back a bit, allowing more space for the light to "expand" into. I'm hoping it's just the position and diffusion of the flash that is the problem as I want to avoid going back to the big flash if possible (or having to buy the next size up)!

 

[GardenersHelper]

An alternate method of "zapping" backgrounds I use is a slight gaussian blur across the frame, and then mask out the subject (which I do with NR also). This should probably be referred to a "nuking" the background,

That's similar to how I do it, in Lightroom, except that I apply strong noise reduction across the frame, and then paint the subject with maximum negative noise reduction to take it back to its original state (or less than maximum on the rare occasions when I do want to give the subject some noise reduction).

I think the effect is much stronger than traditional noise reduction.

I have found the Lighroom NR (using noise reduction up to 100, and Detail 0) strong enough to kill all the noise I've thrown at it so far. For my purposes the advantage over Gaussian blur is that I can do everything in Lightroom, and I'm finding that makes handling the quite large sets that I deal with significantly quicker and easier than using CS for things that Lightroom doesn't do, as I was doing until fairly recently.

Yeah I think one of the main problems I am having is i am not getting the apparent size large enough. I think I may have the flash a bit too close to the subject and I need to move it back a bit, allowing more space for the light to "expand" into. I'm hoping it's just the position and diffusion of the flash that is the problem as I want to avoid going back to the big flash if possible (or having to by the next size up)!

Hmm.... Whatever it is I hope you find it soon and that the solution is quick and easy.

 

[GardenersHelper]

This turned into the story of a journey with four cameras, a small sensor bridge camera (Panasonic FZ200), a micro four thirds interchangeable lens camera (Panasonic G3), a dSLR (Canon 70D) and a point and shoot (“travel”) camera (Panasonic TZ60). My aim was to decide which camera(s) would be best to use to produce the sort of photographs that appeal most to my tastes – pretty pictures of flowers etc and invertebrates. For me, identifying species, photographing species I haven't seen before, recording what is around, where, and when, and revealing exquisite tiny details of my subjects are all secondary. We each have our own “take” on photography, and a desire to capture/create pretty pictures is primary for me. (“Pretty”, of course, means what looks good to my eyes, which may or not be appealing to other people. We all have our own tastes.)

I decided early on that dSLR prime macro lenses, despite their undeniably superior optics, were not for me. I like, and I'm comfortable with, using zoom lenses, usually with achromats on them, and sometimes without. When I started recording this journey I was comfortable with three achromats, the Canon 500D, the Raynox 150 and the Raynox 250. Since then, I have seen and enjoyed many higher magnification photos which other people here produce, mainly with dSLRs, together with special lenses such as the Canon MPE-65 and other approaches such as extension tubes, tele-converters and reversing rings, and of course many individualistic approaches to deploying flash.

Appreciating these higher magnification photos has drawn me slowly into higher magnification work myself, using a more powerful achromat, the Raynox MSN-202, and increasingly into using flash, which is pretty much indispensable for higher magnification work, and through that route coming to use flash increasingly for more moderate magnification work, where previously I had a much heavier emphasis on using natural light. I have also started doing some night time flash work.

On this journey I have also become more comfortable with working hand-held, although a tripod is still an essential part of my equipment, and I have become more comfortable with using manual focus, although I still use autofocus a great deal. I have come to understand more about what can achieved with stacking, even though I don't use it much, and have learnt about some of the more arcane technicalities such as effective apertures, even though some of it doesn't seem to apply to the achromats I use. I now routinely make use of camera profiles and white balance reference shots, although I still use my eyes to adjust white balance manually when reference shots produce results that don't look nice/right. And I've changed the software and techniques I use for post processing.

So the journey has been very productive, even if it didn't lead to what I think I started out looking for (even if I didn't admit it); a single “best” hardware solution for my preferred subject matter. It hasn't worked out that way. On my journey I think I have embraced a wider range of subjects, and a broader set of techniques and technologies, to be mixed and matched as appropriate, with a considerable amount of experimentation needed to work out what matches successfully with what in practice, and an element of art/craft/judgement in deciding what to put into the mixture for particular subjects, ambient lighting, weather, output requirements and (if I'm honest), my mood.

And I think that is pretty much where I have arrived in relation to those four cameras. I did think I might end up using just one of them, probably the 70D as it should be able to produce the highest quality output, or possibly two of them. However, my latest thinking is that I may well use all four, which one I use depending on … subject, ambient lighting, weather, output requirements and mood.

I have very recently done some real world (ie not an artificially constructed test scene) carefully controlled comparison shots using the 70D, G3 and FZ200. No achromats, just the cameras and for the G3 and 70D the zoom lens that I normally use on them. This made it very clear that the 70D can produce superior image quality, in terms of detail, dynamic range and colour rendition. If you have been following this journey you may find this puzzling. For example, I suggested in this post earlier in the thread that the like for like shots I had done seemed to show no significant difference between the 70D and the lowly FZ200 (although Tim @TimmyG did see more detail in the 70D shots). I think the difference is that those earlier comparison shots were captured using the smallest available aperture on each camera/lens, and the resultant (large amount of) loss of sharpness/detail from diffraction tends to “equalise” the cameras in terms of detail capture. The more recent comparison shots were not close-ups, and used large apertures.

The other thing that the recent comparison shots showed was that the 45-175mm lens I use on the G3 has a sharpness problem – in some of the test shots the FZ200 was actually sharper/more detailed than the G3 shot of the same scene, shot within several seconds of the G3 shot.

Many people have said that at certain shutter speeds (around 1/60 sec to 1/125 or so as far as I recall) the lens often does not produce sharp images. When I first bought the lens I had seen this issue discussed and I tested it as best I could to see if it seemed flawed. At that time it did not appear to have the symptoms that were being talked about. It does now.

This may seem very odd. After all I used the lens for the best part of two years and got some nice results with it, both with natural light and flash, and I have recently started using it again, for flash work, and getting some good results. However, with flash, whether recently or earlier, this issue would not arise, because if flash is the dominant light source then the effective shutter speed is the length of the flash pulse, and that is too short for a blurry image to develop. In addition to this, I have always had very high failure rates for close-ups (irrespective of which camera I was using), and if this blurriness problem did arise sometimes with natural light shots it may have become submerged by the much wider issues of lack of sharpness in many of my close-up shots.

The upshot of these tests is that I am now more than ever convinced that the 70D is the best camera to use for natural light shots, especially for flowers but also for some (typically but not only) early morning shots where medium to large sized invertebrates are sitting around motionless for extended periods.

I'm also convinced that I should avoid using the G3 for natural light shots.

This leaves the question of what equipment to use for flash shots of invertebrates (I hardly ever use flash for flowers). Despite repeated attempts I have failed to get the 70D to work properly with flash for close-ups (it works fine for “normal” flash, bouncing flash off a ceiling to take an indoors photo of people for example). And in any case, the 55-250mm lens that I use on the 70D extends a long way and this makes it less and less suitable for use with achromats as the zoom (and hence the magnification) increases. The choice for flash shots of invertebrates is therefore between the G3 and the FZ200.

I have been finding the G3 good to use with flash, and a good, sharp image from the G3 can be cropped significantly and still be useful for my purposes. This makes the G3 especially good (and better than the FZ200) for higher magnification work (with the Raynox MSN-202), where I'm finding it very useful to be able to use deep crops as an aid to improving dof at higher magnification.

This leaves the issue of what equipment to use for flash captures of invertebrates which are larger and for which the less powerful Raynox 150 or 250 is suitable. In this case cropping is not an issue. In addition, I almost always capture invertebrate images using the smallest available aperture, which tends to equalise the details captured with each camera. Which to use therefore comes down to operational characteristics. The FZ200 focuses better (quicker), but the G3 allows a greater range of magnifications for each achromat. Either of these could be important, depending on the circumstances. There is no clear “winner” here for me. What I may well do is take both cameras and leave the MSN-202 more or less permanently attached to the G3, and use the FZ200 for the Raynox 150 and 250. This would allow me to switch between very small subjects and larger subjects simply by transferring the flash unit to the other camera, which is really quick to do.

And the TZ60? I might take this along as my “butterfly camera”, for use with butterflies, dragonflies and damselflies. The light is generally good when those insects are around, so the noisiness of the sensor and the small maximum aperture are not a problem, and unlike many point and shoot cameras it has a viewfinder, which will make it more practical to use in sunshine. Of course, if I have the 70D with me and it isn't set up for some other type of shot on the tripod, then the 70D would be better for butterflies etc. Another possible use of the TZ60 is wider angle environmental/contextual shots of the site I am working on, something of which I should probably do more.

So, when the new season starts in the Spring, I may well take four cameras with me to mix and match depending on the circumstances.

Having got to this point, I think it may be time for me to walk away from this thread now, and revert to normal posting.

 

[TimmyG]

Great summary Nick, and thanks for taking us along with you on your journey. I've learnt lots from it, and funnily enough I believe I am coming to similar conclusions as you, albeit from completely the other angle.

Recently I have been trying to reduce the size of my setup, mainly by swapping out my dSLR for a mirrorless body and reducing the size of the flash. This was primarily to reduce the weight I was carrying when travelling (as I wouldn't just be taking away my macro setup). Although I feel I have achieved a suitable travel set, I had hoped to reduce all my gear and use one setup for all (regardless of if I was travelling or not). Ultimately I think the compromises I am facing with the smaller gear are outweighed by the ease of use and therefore higher hit rate/better result with my original setup. I think I might settle on having a larger "home based" kit, and the smaller one when I am travelling. At least I know I have the option of doing high magnification work when I am abroad, even if I may have to work a bit harder to get the shots I want.

The thing I am finding is that even at home, I don't necessarily want to use either kit all the time. Sometimes I feel like using the smaller setup and can't be bothered carting the larger gear around, other times I want to optimise my image quality as much as possible. Furthermore seeing the results you are achieving with achromats has made me want to explore this option further, as I may be able to reduce my travel kit further and finally wean myself off of the MP-E.

We'll see. All I know at the moment is that it looks like my kit is expanding rather than shrinking and maybe this is as it should be. Having the variety available gives me more options, and perhaps keeps things more interesting. I am able to use the right tool for the right job as it were, and I enjoy the "tinkering" side of photography. This is a hobby after all, so as long as I am enjoying myself, why not?

 

[Tintin124]

I glad you have come to some conclusions Nick and this thread has been great reading and very helpful.

Unlike @TimmyG my set of "stuff" started off small got large now I'm looking to reduce again but I can't do it as I like to play and the 202 goes so nice on the 50mm 1.8 and 250 did go on the 55-250 but as that lens got replaced by the bigma it now additional to the 90mm tammy.

Now I think I want the MP-E mainly to give it a go. Might hate it but might love it too though kit lens has done me proud and it's constantly improving infront of my eyes.

Would love to try these lenses and reversed lens on a Sony mirrorless especially the new one with ibis and it's superior sensor.

 

[TimmyG]

Would love to try these lenses and reversed lens on a Sony mirrorless especially the new one with ibis and it's superior sensor.

IBIS looks very interesting. Possible game changer??

 

[GardenersHelper]

Great summary Nick, and thanks for taking us along with you on your journey. I've learnt lots from it,

That's good. I've certainly learnt a huge amount. And I think that knowing I was going to write it up and needed to be able to justify what I wrote made me a bit more thorough than I might otherwise have been.

and funnily enough I believe I am coming to similar conclusions as you, albeit from completely the other angle.

Recently I have been trying to reduce the size of my setup, mainly by swapping out my dSLR for a mirrorless body and reducing the size of the flash. This was primarily to reduce the weight I was carrying when travelling (as I wouldn't just be taking away my macro setup). Although I feel I have achieved a suitable travel set, I had hoped to reduce all my gear and use one setup for all (regardless of if I was travelling or not). Ultimately I think the compromises I am facing with the smaller gear are outweighed by the ease of use and therefore higher hit rate/better result with my original setup. I think I might settle on having a larger "home based" kit, and the smaller one when I am travelling. At least I know I have the option of doing high magnification work when I am abroad, even if I may have to work a bit harder to get the shots I want.

A lot of this thinking does seem very familiar to me.

The thing I am finding is that even at home, I don't necessarily want to use either kit all the time. Sometimes I feel like using the smaller setup and can't be bothered carting the larger gear around, other times I want to optimise my image quality as much as possible.

I'm very much with you here. Sometimes, especially with flowers, I find it nice to just take out a little camera, perhaps the point and shoot, relax into its simplicity and go with the flow, concentrating on and experimenting with light and composition.

... and I enjoy the "tinkering" side of photography. This is a hobby after all, so as long as I am enjoying myself, why not?

Why not indeed. I did wonder after I'd posted my summary whether in fact there was a second primary element for me apart from pretty pictures - a fascination with experimentation/tinkering with kit, capture options, compositions and processing techniques, just for the sake of it, out of curiosity, for pure enjoyment.

 

[GardenersHelper]

I glad you have come to some conclusions Nick

Although, as you know by now with me, they might change later. (Or indeed sooner )

and this thread has been great reading and very helpful.

Good. I'm glad of that.

Unlike @TimmyG my set of "stuff" started off small got large now I'm looking to reduce again but I can't do it as I like to play and the 202 goes so nice on the 50mm 1.8 and 250 did go on the 55-250 but as that lens got replaced by the bigma it now additional to the 90mm tammy.

Now I think I want the MP-E mainly to give it a go. Might hate it but might love it too though kit lens has done me proud and it's constantly improving infront of my eyes.

Would love to try these lenses and reversed lens on a Sony mirrorless especially the new one with ibis and it's superior sensor.

I did wonder if you would be able to resist the lure of the MPE-65.

 

[GardenersHelper]

IBIS looks very interesting. Possible game changer??

I've read that image stabilisation becomes less effective as magnification increases. Presumably though, high magnification work will use flash, so the usefulness of image stabilisation wouldn't be with image capture, but would be in stabilizing what you see in the viewfinder, assuming an EVF linked to the sensor, rather than an OVF using optical passthrough, which presumably wouldn't be stabilised.

And if image stabilisation doesn't do much for higher magnification image capture, I wonder how much use it would be for stabilisation to aid finding the subject and composition? Hmmm, I feel an experiment coming on, perhaps.

 

[GardenersHelper]

And if image stabilisation doesn't do much for higher magnification image capture, I wonder how much use it would be for stabilisation to aid finding the subject and composition? Hmmm, I feel an experiment coming on, perhaps.

Here we go. Two videos here at Flickr, one with Image stabilisation on and the other with image stabilisation off. I found it instructive to run them side by side. If you start them at the same time you'll need to pause the second one when it goes to a higher magnification to allow the first video to catch up.

 

[Tintin124]

Here we go. Two videos here at Flickr, one with Image stabilisation on and the other with image stabilisation off. I found it instructive to run them side by side. If you start them at the same time you'll need to pause the second one when it goes to a higher magnification to allow the first video to catch up.

Is that in body image stablisation (IBIS) though or lens? In body might have more benefits when macro shooting than lens.

 

[GardenersHelper]

Is that in body image stablisation (IBIS) though or lens? In body might have more benefits when macro shooting than lens.

Lens.

I'm wondering why IBIS might be better. It would be very interesting if it was. I don't have any IBIS kit to try it out on unfortunately.

 

[Tintin124]

Lens.

I'm wondering why IBIS might be better. It would be very interesting if it was. I don't have any IBIS kit to try it out on unfortunately.

You are floating the sensor so in theory every lens has stablisation. Guess you understood that so onto macro... Now handshake is commonly present on the actual body itself as that's where your hands are 90% of time. So you are nipping at source also reducing the movement that occurs (think of a ruler if you hold on one end the movement is excentuated to the other end by original movement times x amount ), apparently ibis in the a7II is 4 stops of exposure which could mean using slower shutter speeds to capture what you want.

Now above is my logic and conjecture so could be totally wrong.

 

[GardenersHelper]

You are floating the sensor so in theory every lens has stablisation. Guess you understood that so onto macro... Now handshake is commonly present on the actual body itself as that's where your hands are 90% of time. So you are nipping at source also reducing the movement that occurs (think of a ruler if you hold on one end the movement is excentuated to the other end by original movement times x amount ), apparently ibis in the a7II is 4 stops of exposure which could mean using slower shutter speeds to capture what you want.

Now above is my logic and conjecture so could be totally wrong.

You may be right. I don't know.

I can't find anything this specific about the effectiveness of IBIS for macro work, but for the in-lens IS of the 100L macro Canon claim "a 4 stop advantage at longer focus distances, dropping to 3 stops at 0.5x magnification and 2 stops at full 1x magnification".

If IBIS has a similar drop-off in effectiveness as magnification increases up to 1:1, then I would expect the effectiveness to continue falling off as magnification increases beyond 1:1, the area in which I think you work a lot.

I'd be delighted to be shown that IBIS is good for higher magnifications (and if so I might want to move in that direction too), but it's the sort of thing you can only prove or disprove with practical experiments I think.

 

[Tintin124]

You may be right. I don't know.

I can't find anything this specific about the effectiveness of IBIS for macro work, but for the in-lens IS of the 100L macro Canon claim "a 4 stop advantage at longer focus distances, dropping to 3 stops at 0.5x magnification and 2 stops at full 1x magnification".

If IBIS has a similar drop-off in effectiveness as magnification increases up to 1:1, then I would expect the effectiveness to continue falling off as magnification increases beyond 1:1, the area in which I think you work a lot.

I'd be delighted to be shown that IBIS is good for higher magnifications (and if so I might want to move in that direction too), but it's the sort of thing you can only prove or disprove with practical experiments I think.

Indeed just need to convince someone to buy one for me to try.

When it comes out may hire one with a canon adapter and try it out.

 

[Tintin124]

You may be right. I don't know.

I can't find anything this specific about the effectiveness of IBIS for macro work, but for the in-lens IS of the 100L macro Canon claim "a 4 stop advantage at longer focus distances, dropping to 3 stops at 0.5x magnification and 2 stops at full 1x magnification".

If IBIS has a similar drop-off in effectiveness as magnification increases up to 1:1, then I would expect the effectiveness to continue falling off as magnification increases beyond 1:1, the area in which I think you work a lot.

I'd be delighted to be shown that IBIS is good for higher magnifications (and if so I might want to move in that direction too), but it's the sort of thing you can only prove or disprove with practical experiments I think.

Additionally you will be going to FF with that body so that's other considerations too. Though saying that it is regarded as the best sensor currently avaliable.

 

[GardenersHelper]

Additionally you will be going to FF with that body so that's other considerations too. Though saying that it is regarded as the best sensor currently avaliable.

Yes, FF. I've been thinking about that. You lose magnification because of it being a larger sensor, but gain in croppability. I wonder where the balance lies with that?

(Well, I suppose that strictly speaking you don't lose magnification. For example the MPE-65 is 5:1 on Full Frame just like it is on APS-C. But the scene size covered at 5:1 is larger with full frame (about 7mm wide) than with APS-C (about 4.5mm wide). That's the sense in which I mean you "lose magnification" by using Full Frame rather than APS-C.)

Looking at the Sony site, one thing I've noticed is that although it has 5-axis stabilisation (which is obviously a good thing, assuming it does in fact help with higher magnification work), but there is one axis that is not corrected for. It corrects for three rotations (pitch, yaw, roll) and two translations (vertical and horizontal movement). What it doesn't correct for is the third translation (forward/backwards), which is extremely important for higher magnification work because of the effect on focus placement (or mis-placement!).

 

[TimmyG]

Looking at the Sony site, one thing I've noticed is that although it has 5-axis stabilisation (which is obviously a good thing, assuming it does in fact help with higher magnification work), but there is one axis that is not corrected for. It corrects for three rotations (pitch, yaw, roll) and two translations (vertical and horizontal movement). What it doesn't correct for is the third translation (forward/backwards), which is extremely important for higher magnification work because of the effect on focus placement (or mis-placement!).

I guess for macro work you don't want it to correct for forward/backward movement. Imagine trying to micro-adjust the focus point using the "rocking" technique and the camera is trying to correct you! That would be very annoying...

 

[GardenersHelper]

I guess for macro work you don't want it to correct for forward/backward movement. Imagine trying to micro-adjust the focus point using the "rocking" technique and the camera is trying to correct you! That would be very annoying...

Ah. Very true. Hadn't thought of that!

 

[GardenersHelper]

Yes, FF. I've been thinking about that. You lose magnification because of it being a larger sensor, but gain in croppability. I wonder where the balance lies with that?

@Tintin124
Bryn, regarding cropping to regain lost "magnification", you might want to look at this post over at dpr. I have no idea if what it says is true, but you might want to dig a bit on this if you thinking seriously about a Sony FF (or FF of any other brand for that matter).

 

[GardenersHelper]

Having got to this point, I think it may be time for me to walk away from this thread now, and revert to normal posting.

Hmmm.... Only fair I think to come back to this thread now, to explain how it is that I have just purchased a Sigma 105mm macro HSM OS.

Early in the year I bought a Canon 100L macro along with an MPE-65 and MT-24ex flash. I didn't get on with the MPE-65 and decided that since I wouldn't be using the MPE-65 for magnifications beyond 1:1 it was difficult to see what the role would be for the 100L. So I sent all three items back.

I kept the camera, a Canon 70D with a 55-250 STM lens on which I mounted my achromats. That worked ok for natural light but I couldn't get flash to work well, so I ended up using the Panasonic cameras (G3 and FZ200) for flash work.

So what has changed?

Three things. A discussion about macro lenses versus achromats on another site, some information I got here from Alf @alfbranch about using achromats with prime macro lenses, and some problems I realised I had with natural light shots.

First the natural light problems. I have had a nasty cold++ which has kept me indoors for a couple of weeks or more and I have spent some time working on my rather large backlog of images. While doing this I realised that while the 70D with the 55-250 and achromats could produce very nice shots, a lot of the time it didn't because of not focusing well. One example that really made me sit up and think was some images of a smallish cricket. I had a sequence of them captured with the FZ200, pleasing shots of the cricket on a flower, all nicely in focus. Then when working on some images from a few days prior to that, I found a sequence of the same sort of cricket on a similar flower captured with the 70D. None of them was properly in focus – they were all unusable. These issues reminded me of something I had put to the back of my mind, that I had previously wondered if I had intermittent focusing issues with the 70D, but had never really pinned it down.

This set me to wondering whether it might be better to use the G3 for natural light shots. I knew it can produce nice natural light images, albeit possibly not quite as nice as the 70D at its best, but the G3 was pretty reliable about getting things in focus. But as I looked through G3 natural light images in my backlog, and then did some comparison tests between the G3 and the 70D, I realised that the G3 also had problems. The focusing was ok, but a lot of the images were too soft/mushy/lacking in detail to use, and some had “ghost” outlines. There has been a lot of talk about issues with the 45-175 lens that I use on the G3, about it having shutter shock or some such issues at some shutter speeds around 1/60 sec or so. I tested for these when I first got the camera, and didn't find them. But now, I think perhaps I do have this issue.

So, I wasn't fully happy with the 70D + 55-250 or the G3 + 45-175 for natural light shots, and the FZ200 didn't seem like the answer either, partly because of noise issues (noticeable sometimes even at base ISO when I go in for deep shadow recovery, which I do quite a lot because of the extent to which I underexpose botanical images in order to avoid colours “bleaching”).

I spent some time thinking about and researching other options, including

• Full frame, Sony with 5-axis IBIS, with a Sony or third party 100mm or so macro lens
• Alternative micro four thirds hardware, Olympus OMD EM-1, also with 5-axis IBIS, with Olympus 60mm macro
• A proper macro lens on the G3, such as the Panasonic Leica 45mm macro, or possibly the Olympus 60mm or a third party macro lens using an adaptor.

There were reasons against all of these.

Which brings us to Alf's input.

I was used to using achromats on a zoom lens, in which case you keep the camera stationary and zoom in and out to change the magnification. What I (wrongly) concluded from this was that for any particular focal length you get a particular magnification, which would mean that if you put an achromat on a prime macro lens you would be limited to one particular magnification and so this wouldn't be much use to me in trying to go beyond 1:1 because I would be limited to just four or five magnifications, one for each of my achromats, and that would be too inflexible for my needs.

I also misled myself with an experiment. Thinking about getting a 100mm or so macro lens, I wanted to see what the effect of adding an achromat to it would be, so I set my zoom lens to 100mm and added achromats, and it didn't seem that they were going to have a very useful effect in terms of the amount of magnification they would provide.

Enter Alf, who tells me about him and others using achromats on prime macro lenses, and telling me the magnification I could expect with various achromats on a Sigma 105 like Alf uses. These were really useful magnifications. Obviously there was something wrong in the way I was thinking about this.

Some quick tests made clear something I “knew” already (was vaguely aware of) but hadn't thought through. With an achromat on the camera, if you keep the focal length constant there is a range of distances from the subject at which you can get a sharp image. And as you change the distance the magnification changes, along with the framing/scene width that is covered.

The, in retrospect, obvious point seemed to be that macro lenses have a much larger range over which they can focus than non-macro lenses like I have been using. They will all focus at infinity but whereas normal lenses of the sort I'm using stop at about a metre or so, primes go on down to a few inches. Put an achromat on a prime macro lens and change the distance to the subject, and because of the large range over which they will focus you get a correspondingly large range of magnifications, including much bigger magnifications than I could get with the same focal length on a non-macro lens. That is my theory at least. It remains to be tested when I get the new lens.

And lastly, the discussion at another site. Someone there insisted that I would get better detail, clarity, colour rendition, bokeh etc from a proper macro lens. I explained, and he accepted, my argument about invertebrates, where I typically use such a small aperture that it probably doesn't really matter what optics I use because diffraction is a great equaliser (one of the things I discovered comparing like for like 100L and achromat images at f/22-f/32). However, I'm if anything even more interested in flowers and other botanical subjects than invertebrates, and for botanical subjects I do use a whole range of apertures, and only rarely the very smallest apertures. And for botanical subjects the rendition of textures, colours and backgrounds is, for me, very important. I had to admit that a prime macro lens might well be better for my botanical subjects.

I then read about the Canon 7D mark II, a terrific, top of the line bang up to date APS-C camera. It's particular strengths aren't that important to me, but I read one thing that really made me think. Allegedly, the sensor in the 70D is pretty much as good as the sensor in the 7D mark II. I remembered that the whole point about getting the 70D was that it has a larger sensor than I have previously used, with the potential for good image quality that that implies. When I added that to what Alf had told me about the flexibility that achromats can add to a prime macro lens, and the discussions about botanical imaging, and when I thought of the image quality people here and elsewhere get with APS-C cameras with 90/100/105mm macro lenses, it struck me that it surely had to be worthwhile having another go with a macro lens on the 70D. If it worked nicely for natural light botanical imaging that would be good. And of course it was possible that I could also get it to work for invertebrates, including possibly with flash. But if that didn't work out I could live with it, as the G3 works nicely for invertebrates when using flash, right down to springtail size.

It seemed to me that it had to be worth trying again with a macro prime on the 70D, and the finance manager agreed.

90/100/105mm seemed about right – too little working distance for comfort with shorter focal lengths, and too much weight with longer focal lengths. I did some research on the 90mm Tamron, the Canon 100, which I have used briefly, and the Sigma 105. I decided to go for the Sigma 105, the OS HSM version that has internal focusing (just under £400 compared to £665 for the Canon 100L). There was some difference of opinion over on the Canon forum at dpreview as to how well autofocusing would work in live view with non Canon lenses, but the general feedback was positive (I couldn't get a specific answer about the Sigma 105, although the Sigma 150 amongst others was said to work fine). In any case my main concern was with botanical work, for which I have increasingly been using manual focus.

So, I would use the Sigma for botanical work, all of which is natural light, except for very occasional fill flash, for which I have before now simply used the pop up flash. Worst case, if necessary, I would use only manual focus. If the Sigma proved to be ok for natural light invertebrate shots that would be nice, but not essential, and similarly for flash and higher magnification invertebrate work.

I ordered the Sigma this afternoon. It has been dispatched and should arrive tomorrow.

I did wonder if I should hold off posting anything until I've tested it out to see whether I've made (another) potentially expensive and given past history possibly stupid mistake, about which perhaps I might want to keep quiet. However it seems to me that the story of this journey has included ups and downs before now, and if I've been stupid, or even if I haven't but it doesn't work out, well, so be it. That will be a valid part of the story.

Oh, on the possible stupidity front, one of the half dozen or so reviews of the Sigma 105 that I read mentioned sharpness problems around 1/60 sec. And I read that before I ordered it. Best to get that out into the open now rather than later I suppose.

 

[Tintin124]

And Nick we start again... look forward to seeing the results. Only question is how often do you go above 1:1 with flowers though? In response to putting Achromats on the macro lens?

 

[GardenersHelper]

And Nick we start again...

look forward to seeing the results.

So am I. Well, a bit nervous too actually. "Have I been stupid..." etc.

Only question is how often do you go above 1:1 with flowers though? In response to putting Achromats on the macro lens?

Not very often. (Most of my flower shots are done either with no achromat or with the 500D, which only goes to about 1:1.5.) That is why I'm relatively relaxed about how well the Sigma will work for me with achromats, given my emphasis for the Sigma on flowers.

On the subject of how well or otherwise the Sigma might work with achromats, I had another thought this morning. I remembered that the first (non-STM) 55-250 that I had didn't play nicely with achromats as far as autofocus goes (not at all with the 250 and upwards, and very poorly with the 150). It was only with the STM version that achromats became a realistic option.

So, even if autofocus turns out to work ok with the bare Sigma (which it might not anyway), it might not work with some or all of the achromats. If it does turn out that way, I'm still not sure about the significance of it, as I'm using manual/rocking focus more these days (especially since trying the very small stuff like springtails, for which autofocus doesn't work).

Perhaps this will turn into a longer term thing of concentrating more on, and with practice getting better with, manual focus. Using manual focus with live view seems quite practical with the 70D, its LCD being very good, and the 5x and 10x magnification options (which I haven't used much yet) appearing to be quite well implemented.

 

[GardenersHelper]

I have been getting to grips with the Sigma 105 macro. It is the version with image stabilisation and internal focusing (so the lens does not extend/contract when focusing). The aperture is from f/2.8 to f/22.

It is an EF (full frame) lens and is quite big and heavy (slightly bigger and heavier than the Canon 100L macro for example). However, I find it quite comfortable on the 70D, and the weight is manageable for hand holding, even with the Metz 58 A-2 flash attached.

The magnification and focus distance are shown in a windows on top of the lens. Unfortunately when the flash is mounted the diffuser blocks my view of the window. I don't think this will matter much (but perhaps I shall regret saying that!).

The maximum magnification is 1:1 with a scene width (on Canon APS-C) of 22.5mm, which is achieved at a working distance of about 140mm.

The lens has three switches:

• A 3-way focus limiter, with settings for

The full range from a working distance of about 140mm to infinity.
From a working distance of about 280mm (for a magnification of about 1:2.5 and scene width about 55mm) to infinity.
From a working distance of about 140mm to about 280mm (1:1 to 1:2.5).​

• An autofocus/manual focus switch

• A 3-way image stabilisation switch: off, on, on for panning (vertical stabilisation only)

Apertures

f/2.8 is a larger aperture than I have ever used before (apart from my very brief time with the Canon 100L, when I didn't really explore large apertures). It will be interesting to see if I find any use for the narrow dof it produces.

Given that I often use very small apertures, f/22 on APS-C appears slightly restrictive at first sight. After all, the 55-250 goes to f/32. However, while the 55-250, used with achromats, does not suffer from a reduction in effective aperture as magnification increases, the Sigma 105 does, like any other macro lens. This means that at 1:2, a nominal f/22 produces an effective aperture of f/33, and at 1:1 an effective aperture of f/44. Given that the 55-250 only goes to f/32 at the longest focal lengths, while being restricted to f/22 at the shortest focal lengths, it looks like the minimum effective aperture of the Sigma is in fact less, for closeups, than the minimum (nominal = effective) aperture of the 55-250 with achromats.

I suspect this is reflected in the examples below. The snail may be the most telling of them. For snails I would typically use f/22-f/32, ,the smallest possible aperture, depending on focal length. I used f/14 for the snail example below. I think I will be experimenting with a more flexible approach to apertures for invertebrates. As to flowers, the first example used f/22. Given that this was a small flower (and hence a relatively large magnification) I suspect the effective aperture may have been small enough to give me greater dof than I could have achieved with an achromat on the 55-250, or at least as much.


Autofocus

The Sigma 105 autofocuses well on the 70D. Focusing is fast using the optical viewfinder, which uses phase detect autofocusing (with 19 focus points on the 70D). In live view the focusing is not as fast, but looks like it will be fast enough for my purposes. Live view uses contrast detect autofocusing, but with some lenses the 70D can use on-sensor phase detection in combination with contrast detection, which is faster than just using contrast detection. I think this only works with Canon lenses, and only some of them. I think the 55-250 STM is one of them, and this (without and with achromats) focuses significantly faster in live view than the Sigma 105, which I believe only uses contrast detection.

The Sigma 105 appears to work well with achromats. Autofocusing (both live view and viewfinder) works with my achromats about as well as with my other cameras, becoming harder to use and less useful as the magnification increases, and not much use at all beyond 2:1 or so.


Magnification, scene widths and achromats

The first of the two tables below shows what magnifications are available using various achromats on my Panasonic G3, and the associated scene widths and working distances. The top part of the lower table shows similar information for the Sigma 105 mounted on the Canon 70D without an achromat, and the bottom part shows the information when achromats are mounted on the Sigma 105. (The measured numbers, especially the small ones, are somewhat approximate, and for ease of presentation the calculations have been rounded, in not necessarily consistent ways.)


Magnification, scene sizes and working distances for achromats with G3+45-175 and 70D+Sigma105 by gardenersassistant, on Flickr

The situation is quite complicated, and the following graph might, or might not, make some of this easier to digest. (Each of the straight lines should probably be a curve, but I only measured the minimum and maximum values so the graph is rather crude.)


Scene width and working distance graph for achromats with G3+45-175 and 70D+Sigma105 by gardenersassistant, on Flickr

I think the main message for me is that the working distance for a particular magnification depends on which achromat (if any) is used, but below 1:5 or so, for any particular magnification it is always possible to use a larger working distance with the G3 than the 70D if you choose the right achromat. How significant the smaller working distances with the 70D might be in terms of scaring off jumpy invertebrates remains to be seen.

The other major message is that it will be possible to use the Sigma 105 to get to 4:1, which is about as far as I want to get anyway. I suspect I will go looking for springtails etc to test out some higher magnifications. I did try some hand held shots of fruit flies in my initial night time test (see below) but none of them was even remotely usable. Unlike springtails at ground level, for which hand held is practical because I can brace the camera on the ground or on my legs (I'm kneeling down for these shots), the fruit flies were on a bush and I had no way of bracing myself sufficiently to make focus acquisition practicable.


Image stabilisation

Sigma claim four stops of image stabilisation at non-macro distances. It's difficult to measure, but from the limited tests I did it looked more like three stops or so to me. As the magnification increases image stabilisation becomes less useful. The most thorough test I did was at about 1:2, and although the results were very variable from shot to shot (with and without image stabilisation), overall it looked as though stabilisation was still giving a clear benefit at that magnification, although I couldn't say how great in terms of stops.(My guess would be perhaps one stop or so at 1:2 and quite possibly more or less none at 1:1.)

As well as a hand-held test at about 1:2, I did a test at the same magnification of the same subject using my hands-on tripod technique, with and without image stabilisation. Using the tripod without image stabilisation there was a significant improvement over hand held without stabilisation. Turning the image stabilisation on provided a bit more of an improvement. Here too I couldn't say how great in terms of stops. So, unlike with a "pure", hands-off tripod approach, for which image stabilisation should be turned off, it seems that keeping it on is a good idea for hands-on tripod use (which is the way I have been working for some time. Don't know why I never did a comparison test before.)


Flash

I've had trouble getting flash to work with my achromats on the 70D. The Sigma 105 arrived late in the day and in my enthusiasm I wanted to try some “real world” shots with it rather than just test shots, and I didn't want to wait, so I went out late in the evening with the night time flash arrangement. Rather to my surprise, it worked fairly well.

There are obviously not a lot of flowers around at the moment, and most of the invertebrates I tried to photograph were slugs. The equipment worked fine (e.g. focusing was good), but I really don't like the flash reflections I get with slugs so I didn't keep any of the slug photos. Here are four of the eight photos that I did keep.


0639 1 2014_12_15 IMG_7309 LR by gardenersassistant, on Flickr


0639 3 2014_12_15 IMG_7350 LR by gardenersassistant, on Flickr


0639 4 2014_12_15 IMG_7370 LR by gardenersassistant, on Flickr


0639 6 2014_12_15 IMG_7455 LR by gardenersassistant, on Flickr

These make me feel that the Sigma 105 is working ok with the 70D. As to whether it will produce better results than using the 55-250 is not obvious to me. As usual, it is complicated, involving a mixture of image quality and usability issues such as speed and accuracy of working, whether subjects are frightened away etc.

In the next instalment ….

The following morning I went out and worked hand-held for a little while photographing what flowers I could find. The light level was rather poor, and I think I would have done better to have used the tripod, so I don't know if there will be much to show from that session. My next task is to go through the photos from that session. More when I've done that. I don't know how long that might be – pre-Christmas and gardening duties are taking a fair bit of time at the moment.

 

[LCPete]

I haven't looked in this section for ages so I've lost track, will try to keep up with this section as it looks like people are still shooting macro
glad that you're getting on well with the sigma 105 it's an excellent lens
I'm only viewing on a small screen but the results look great

 

[GardenersHelper]

I haven't looked in this section for ages so I've lost track, will try to keep up with this section as it looks like people are still shooting macro

Indeed they are. Lots of great shots of small subjects amongst some other interesting stuff.

glad that you're getting on well with the sigma 105 it's an excellent lens
I'm only viewing on a small screen but the results look great

Thanks Pete. But as far as the Sigma goes it's an ongoing, and as usual somewhat twisty/turny, story, as you'll see from the following posts.

 

[GardenersHelper]

In the next instalment ….

The following morning I went out and worked hand-held for a little while photographing what flowers I could find. The light level was rather poor, and I think I would have done better to have used the tripod, so I don't know if there will be much to show from that session. My next task is to go through the photos from that session. More when I've done that. I don't know how long that might be – pre-Christmas and gardening duties are taking a fair bit of time at the moment.

Change of plan.

Having discovered that the 70D+Sigma 105 works with all of my achromats, and with flash, I decided to try it out with the MSN-202 for small subjects like springtails. I went to a local wood where I had previously found springtails. There were plenty of subjects, mainly globular springtails and also some of a more cylindrical, slightly larger type of beastie. (Another springtail? I don't know.)


0640 4 2014_12_18 IMG_8162 LR by gardenersassistant, on Flickr

I worked hand held, using manual focus. I used an aperture of f/16. My thinking was that the magnification I would be working at would take the effective aperture towards and possibly beyond f/22, which is what I use with the G3 and MSN-202. I captured about 180 shots of these subjects.

I thought it went quite well. It's true that the gear was heavier than I'm used to using, and I did have some difficulties locating the subjects (that is, having found a – possible – subject, I had difficulty getting the camera pointing in exactly the right direction and positioned at the right distance from the subject so as to be able to make it out on the LCD). However, having lined the camera up I thought the captures seemed to go quite well from the brief glance I had of each shot on the LCD.

I was wrong. And very disappointed. On my first quick run through on the PC I thought there were no usable images. With a second, much slower and more careful trawl came up with two. This was one of them, the other was very similar. (Looking at it here it seems to have the Flickr-induced over-sharpening issue. Oh well. As usual there are larger versions of all the images over at Flickr.)


0640 1 2014_12_18 IMG_8268 LR-2 by gardenersassistant, on Flickr

I didn't get anything usable for the other, larger subject. The image above was the nearest I got – and that was very much better than most of the others. (The above image is a (post processed) 100% crop.)

There were lots of images which were grossly out of focus. Of the rest, almost all had dof which seemed to be either too narrow, or/and misplaced (it is difficult to tell which in a lot of cases).

I compared these 180 images with the 81 images I captured of the springtail shown in the first two images in this post using the MSN-202 on the Panasonic G3+45-175, working hand held. As I mentioned in that post, almost all of them were pretty much in focus. By and large the dof seemed to be better placed, and more of it. These differences were reflected in the fact that I chose to keep 13 of the 81 shots (and in choosing some of those 13 I had some other very similar shots to choose amongst), as against the 2 from 180 with the 70D+105.

Apart from the possibility of just having had an off day (that happens of course), I wondered if it had to do with the technique and settings I was using. It struck me that using f/16 may not have been a fair comparison. Even if the magnifications I was using had taken the effective aperture to f/22 (and a lot of them might not have taken it that far), f/22 on the 70D gives only about 70% of the dof that f/22 gives on the G3 with its smaller sensor. Also, I may have, on average, made the subject larger in the frame with the 70D session than with the G3 session, and cropped less, which would have increased the disparity in dof.

To test these ideas I went back the next day to the same wooded area looking for small subjects. This time I used f/22 on the 70D, and I tried to keep the subjects small in the frame. I also took the G3 so that if the opportunity arose I would try some real world comparison shots. In addition, I decided to use a tripod-assisted approach rather than working hand-held, so as to aid subject acquisition and framing consistency between shots, and to reduce the randomising effect of hand-shake in terms of focus placement.

Unfortunately, there was not a springtail to be seen. However, there was something else which I took lots of shots of with the 70D+Sigma 105 and the Raynox 250 and 150 stacked. It stuck around long enough that I decided to swap the achromats and the flash diffuser on to the G3 and try and get some comparison shots.

While capturing the 70D shots I once again thought they looked ok as they flashed up on the LCD. In contrast I was exasperated that I couldn't seem to get the G3 to focus on the subject at all.

On the other hand, switching directly from the 70D to the G3 it was obvious to me that the G3 was significantly smaller, lighter, easier and faster to handle, and it was far easier and faster to locate the subject with the G3. This was partly because of the ability to easily zoom in and out (3.75x) (out to locate the subject, back in to frame the shot), without moving the camera. With the 70D I had to combine rotating the focus ring with moving the camera back and forth to achieve this effect, and it was difficult to find the right combination, which resulted in a lot of frustrating floundering about. In addition, the ratio between max and min zoom of 2.1x was much less than the 3.75x with the G3, which further decreased the usefulness of the manoeuvre. In terms of usability, I found the G3 setup far superior to the 70D setup.

I also seemed to be able to make more sense of the out of focus areas on the LCD with the G3, which also helps with locating the subject. I'm not sure why this is, especially as the 70D has a much better (clearer, brighter) LCD. It may be the case that the G3 LCD shows two to three times more dof than the 70D. This would be the case if both cameras use the maximum available aperture for the live image on the LCD. This is f/2.8 for the 70D with the Sigma 105 and f/4 to f/5.6 depending on focal length for the G3 with the 45-175. f/4 provides 1.4 times the dof of f/2.8, and f/5.6 provides twice the dof. Added to this is a dof difference of about a stop (another multiple of about 1.4) because of the difference in sensor size.

Looking through the images on my PC I found one side-on pose that was very similar and similarly framed with the two cameras. I had about 25 attempts at it with the 70D and about 20 with the G3. Looking at them on the PC, almost 20 of the 70D shots seemed to be moderately well focused. Only one (or if I'm generous, perhaps 3) of the G3 shots were moderately well focused. Given how I had felt at capture time, this didn't surprise me.

What did surprise me, was when I compared the best of the shots from each camera. Here they are. (By my calculations the subject is about 1.5mm long, body and head.)

70D + Sigma 105 with Raynox 250 and 150 stacked

0641 02 2014_12_19 70D+Sigma 105+Raynox 250+150 IMG_8437 LR by gardenersassistant, on Flickr

G3 + 45-175 with Raynox 250 and 150 stacked.

0641 01 2014_12_19 G3+45-175+Raynox 250+150 P1880837 LR-3 by gardenersassistant, on Flickr

To my eye there is a big difference, and it is in favour of the G3. There are some caveats. The 70D used ISO 200, the G3 ISO 160. Both used f/22, but the effective aperture would have been smaller for the 70D (increasing dof but also increasing loss of sharpness/detail from diffraction). The images have both been heavily cropped (but by a very similar amount). The G3 image is a 100% crop. The 70D image was downsized to the same number of vertical pixels as the G3 image (this was needed as the 70D has more pixels than the G3, 20 megapixels vs 16). Both images have been post processed, including sharpening.

Apart from that subject, the only other suitable subject was one of the more cylindrical creatures I had seen the previous day. I only had the chance of using the 70D for this, and like the day before I didn't get any satisfactory images. This was the best I could manage.


0641 03 2014_12_19 IMG_8459 LR by gardenersassistant, on Flickr

I went home, not very happy, feeling I hadn't really made any progress (of course I hadn't examined the images at that point).

But, the day wasn't finished yet ...


Continued in next post.

 

[GardenersHelper]

Continued from previous post.

After having some lunch, and before looking at the morning's images, I decided to go out into our garden with the 70D and G3 while it was still light to see if there were any suitable, small subjects to work with.

As in the morning I was disappointed. No springtails, barkflies or similar. I was reduced to capturing images of seeds, pieces of muck, blemishes on leaves and odd bits of detritus. I was working in tripod-assisted (hands on the camera) mode, with the MSN-202 on the Sigma 105 on the 70D and on the 45-175 on the G3.

There was only one subject I did comparison shots of before I got sidetracked from doing comparisons (see below). I don't know what this spherical thing was, but by my calculations it was about 0.7mm in diameter.

I had four attempts at capturing it with the 70D, using manual focus. I think I stopped at that point because I was working carefully, tripod-assisted, and what I was seeing on the LCD looked similar from shot to shot and so I thought was probably about as good as I was going to get.

I then had 10 attempts with the G3. I think the first four attempts used manual focus but I was disappointed in what I was seeing on the LCD compared to what I had seen for the 70D. I then tried autofocus. I was not expecting this to work as I was working at the APS-C equivalent of 5:1 (4.5mm scene width). However, it did work and the sharpness looked better with all six of the autofocused shots.

The best shot from each camera is shown below. There are similar caveats as with the previous example concerning ISO, aperture and post processing. In terms of cropping, in this case it was the 70D image that had to be cropped most, because the 70D could not reach the (APS-C equivalent) 5:1 magnification that I was using with the G3. The 70D version below is a minute fraction less than a 100% crop. The G3 version was downsized to the same, 1100 pixel, height.

70D + Sigma 105 with Raynox MSN-202

0642 01 2014_12_19 IMG_8474 LR by gardenersassistant, on Flickr

G3 + 45-175 with Raynox MSN-202

0642 03 2014_12_19 Lightroom CA correction + Green Hue defringing P1880853 LR by gardenersassistant, on Flickr

To my eye, and to my surprise, once again it is the G3 version that is best.

As a side issue, the G3 version had a great deal of green fringing around the edge of the subject. The set at Flickr from which the posted versions are taken also contains versions of the G3 image showing what it looked like:

• With no chromatic aberration correction in Lightroom, showing a heavy green fringe
• With chromatic aberation correction, which left a fairly faint green fringe
• With chromatic aberration correction and also Lightroom Green Hue Defringing, which is the version posted above
• With chromatic aberration correction and also Lightroom Green Hue Defringing and Purple Hue Defringing. The Purple Hue defringing was an attempt to remove what looked to me like a possible reddish fringe around the subject, but using this changed the colour of the background unacceptably.

I don't know if this is an issue with the G3 in particular. I suspect it is an issue with the MSN-202 (as previously discussed I think), and I suspect there would have been a green fringe around the subject with the 70D if there had been a 70D image which had been in better focus. But that is a guess.

Anyway, that was the point at which the first distraction turned up. While I was photographing the spherical object I noticed a small orange blob moving around, and then two of them. I had 58 attempts at capturing an image of whatever it was, which subsequently turned out to be about 0.2mm in length by my calculations.

This was a big ask. Operating at 5:1 in APS-C terms on a small subject that was skittering around, fast, and by the way using autofocus, and contrast detect autofocus at that, which of course is relatively slow. And it wasn't just that it was moving around a lot in the plane of focus (I should be so lucky), it was moving nearer and further from the camera as well, which of course complicates matters considerably at that magnification.

When I looked at the images on my PC it turned out that 5 of them were somewhat in focus on the subject. None of them was what I would consider usable, which I don't think is surprising in the circumstances. This is a (post processed) 100% crop of the best of them. Not great, but I'm interested that I could get anything at all in those circumstances. With a subject that wasn't moving so much, better results might be possible.

G3 + 45-175 with Raynox MSN-202

0642 06 2014_12_19 P1880905 LR by gardenersassistant, on Flickr

Which led on to the second distraction. With the rig still set up in the same way (G3 with MSN-202, autofocus with the smallest available focus point box, camera on the tripod for tripod-assisted use), I noticed something a lot bigger in what we laughingly refer to as the “grass” or “lawn” at the edge of the flower bed I had been working on. It was a hopper of some sort I think, a little over 3mm long by my calculations. It was moving around more or less continuously in a small area of vegetation, and as it moved around over the next 20 minutes I captured images at an average rate of about one every 4.7 seconds, for a total of 253 images.

I changed the magnification quite often. For example, there was one passage where I shot with different magnifications for ten captures in a row, and at one point I changed the magnification from 1:5 to 5:1 from one shot to the next shot four seconds later. Overall, I used the following magnifications:

• 18% at 1.5:1
• 36% at 1.5:1 to 2:1
• 43% at 2:1 to 3:1
• 1% at 3:1 to 4:1
• 2% at 4:1 to 5:1

Almost all the shots had part of the subject in focus, and in the majority of cases the centre of focus was where I intended it to be. This included several shots where a significant proportion of the subject was obscured behind vegetation, including one where only a very small slice of the subject was visible. It was possible to get good focus in these cases because the G3 allows you to use a very small focus area. The smallest (live view) focus area available with the 70D is much larger.

Here are five of the images, post processed. I have included 76 of the 253 images in the Flickr album containing the images used in this post.



0642 27 2014_12_19 P1890041 LR by gardenersassistant, on Flickr


0642 11 2014_12_19 P1880998 LR by gardenersassistant, on Flickr


0642 13 2014_12_19 P1890013 LR by gardenersassistant, on Flickr


0642 07 2014_12_19 P1880962 LR by gardenersassistant, on Flickr


0642 08 2014_12_19 P1880976 LR by gardenersassistant, on Flickr

While I was capturing these images it struck me that I wouldn't have been able to do the same with the 70D. And on reflection subsequently, I'm even more convinced of that. It needed more flexibility, speed of operation, capability to (re-)locate the subject, rapid and accurate autofocus centred on a very specific, small area than the 70D could provide.


Conclusions?

I suppose it is a rather limited amount of evidence thus far, but it is looking like the G3 with its 45-175 can hold its own quality-wise compared to what I can get from the 70D and Sigma 100 when both are used with achromats to capture images of fairly small subjects. And in terms of usability, given my preferences and working methods, the G3 seems clearly superior. This being the case, and until I can summon up the motivation to do more comparison tests on this sort of subject, I foresee me using the G3 rig for small subjects like springtails, barkflies and fruit flies.

I still have an open mind about larger invertebrate subjects. For example, for snails in motion in natural light the 70D with the Sigma 105 or the 500D on the 55-250 may be the best option, because of the 70D's superior handling of higher ISOs. The nice quality the 70D sensor can deliver may also make the 70D better for snails in motion with flash using the Sigma 105, and/or early morning natural light shots of stationary dew-covered invertebrates such as non-tiny flies, damselflies, crane flies etc using the Sigma 105 by itself or with an achromat to get a bit beyond 1:1. And the 70D with the Sigma 105 my turn out to be better in general for invertebrates subjects/scenes needing magnifications no larger than 1:1. However, I won't be able to compare the rigs for those subjects until the Spring.


And next? ...

My next move is to go back to those flowers I captured hand held a few days ago with the Sigma 105 and see if there is anything to learn from them. However, I think I'm probably going to need to do some more comparisons. As with the non-tiny invertebrates, that may need to wait a while until some better subjects come into view.

 

[LCPete]

Very interesting Nick the last set, the hoppers are really good
it does look like the G3 is better for fiddly small subjects
as you say though I'm convinced that when you find some larger insects in the spring like early morning damselflies the 70D and Sigma will be perfectly suited to that
I'm normally at ISO 800 for that sort of thing and the image quality of the 70D will be excellent at 800 or even higher
I'm not sure but wonder if the larger sensor of the DSLR is more prone to diffraction effects at small apertures than the G3
Can't get my head around it at this time of night!
Definitely worth using the 70D and Sigma 105 for 1:1 subjects and closeups though you will get wonderful results with that setup

 

[GardenersHelper]

Very interesting Nick the last set, the hoppers are really good

Thanks Pete. I was quite pleased (and a bit surprised actually) at the clarity and the way the backgrounds worked.

I think the surprise came from the fact that the G3 can be quite misleading in terms of what you see (unmagnified) on the LCD after taking a shot. Whereas on the 70D everything tends to look really good (even when it isn't), things can look rather poor and washed out on the G3 (even when they aren't). Hmmm.... maybe I'm over-simplifying, or over-reacting. I'll try to do some explicit side by side LCD comparisons - I've never really done that.

Also, before post processing the G3 images tend to look worse on the PC than the 70D images. The 70D images don't seem to need so much (or much at all) doing to them a lot of the time.

it does look like the G3 is better for fiddly small subjects

It's certainly looking that way. But ...

I'm not sure but wonder if the larger sensor of the DSLR is more prone to diffraction effects at small apertures than the G3
Can't get my head around it at this time of night!

... the business about image quality with the 70D+Sigma 105+strong achromat is really bugging me. Maybe there is some hardware difference with the sensors, or possibly the lens characteristics. I feel some more tests coming on to at least try to pin down the circumstances in which these differences arise. I've even wondered if there is a fault with the Sigma, but it's really too early to be thinking about that. I'd need to show it was only the Sigma that had the issue before going down that road.

as you say though I'm convinced that when you find some larger insects in the spring like early morning damselflies the 70D and Sigma will be perfectly suited to that
I'm normally at ISO 800 for that sort of thing and the image quality of the 70D will be excellent at 800 or even higher

Yes, chroma noise on the 70D is a non-issue and the luminance noise is fine and non-clumpy.

I think it is working distance that may be more of an issue for some of my subjects. I think damselflies should be no problem in terms of working distance. A quick test suggests I might be working at 40-50cm or so with the Sigma 105 for the sort of framing I typically use (whole body, side on, nowhere near filling the frame). With flies though, it's a different matter. I often need to go in beyond 1:1, even for whole body shots. For example, this one (which is uncropped) was captured using an achromat on the 55-250 STM on the 70D. If (as I suspect) it was captured using the Raynox 150 it had a scene width of about 19mm, which of course is a bit beyond 1:1, and it wasn't by any means a small fly. And if it happened to have been the Raynox 250, then the scene width would have been about 13mm.


0552 046 Myopa 2014_04_09 IMG_3579-Edit-2 PS1 PSS3 by gardenersassistant, on Flickr

With the Sigma 105 this would have needed cropping (and the greater the amount of cropping the more noise becomes important) or an achromat on the 105 (which would reduce the working distance from the 140mm for 1:1 on the Sigma 105, versus 160mm or more with the Raynox 150 on the 55-250 to get a bit beyond 1:1, as - probably - in this example).

Definitely worth using the 70D and Sigma 105 for 1:1 subjects and closeups though you will get wonderful results with that setup

I'm definitely hoping you are right about that (subject to any working distance issues of course).

 

[Tintin124]

I need to read the 2 posts prior to the last one but I needed to mention don't trust the LCD on camera any camera they are cr@p. You are better off judging your image by the histogram number of times I have been caught out by the LCD back when I started and it always comes out underexposed on PC compared to looking magnificent on camera LCD. Not really worth your effort compared to just checking histogram.

 

[GardenersHelper]

I need to read the 2 posts prior to the last one but I needed to mention don't trust the LCD on camera any camera they are cr@p. You are better off judging your image by the histogram number of times I have been caught out by the LCD back when I started and it always comes out underexposed on PC compared to looking magnificent on camera LCD. Not really worth your effort compared to just checking histogram.

He he. Different strokes ....

Don't know if it makes a difference, but I almost always use the LCD and almost never use the viewfinder on any of my cameras. (Maybe a couple of times a year, for a few shots.)

I expose based on controlling blown and bleached areas, not how good the image looks on the LCD. I find the LCD works ok for that on all my cameras, once I've learnt how to "read" the LCD in terms of what the outcome is likely to be after post processing.

From the point of view of my cameras' evaluative metering I am usually underexposing, quite often by a stop or more.

My OOC histograms often have very long flat tails at the top, indeed, they often have long flat tails at the top after post processing. The length of the tail, or more generally the shape and distribution of the histogram, tells me nothing at all about bleaching and nothing of use about blown areas (I can see them on the LCD). I have the histogram on the camera turned off - I have no use for it.

Each to his own, and vive la différence.

 

[GardenersHelper]

I have been saying for months now that I can't get flash to work properly for close-ups on my 70D. I've also said, if not here then certainly elsewhere, that I can be very slow on the uptake sometimes. Well here is a case in point.

I have been trying to use flash for close-ups with achromats mounted on a 55-250 STM on my 70D. I have been using a Metz 58-AF2, exactly the same model (apart obviously from being Canon fit rather than Olympus/Panasonic fit) that I use with my G3 and FZ200, and they work ok with flash. On the 70D though the Metz has been producing wildly inappropriate exposures, radically over-exposing or under-exposing, yet sometimes working fine. I couldn't understand what was going on, and it proved impossible to use, which is why I turned to my Panasonic cameras for flash work. Was it an issue with the camera, or the flash, or some combination of the two? I had no idea.

Here is the way I would expect flash to work. First, the exceptions. If the aperture is small enough/the ISO low enough/the subject far enough away from the flash, then the flash may not be able to provide enough light and the image will be under-exposed. Similarly, if the aperture is large enough/the ISO high enough/the subject near enough to the flash, then the flash may be unable to provide a small enough amount of the light and the image will be over-exposed. Within these limits, with the flash set to use TTL metering and the flash exposure compensation set to whatever seems appropriate, you ought to be able to alter the ISO and the aperture as you wish and the exposure should remain unchanged.

This, it turns out, is exactly how flash works with the Metz on the 70D with the Sigma 105. Having discovered that, I found that it also worked in this way with the Canon 18-55 STM (which is a lens I have never tried before with flash). And then I tried the 55-250 STM (as with the 105 and the 18-55, just the bare lens, without an achromat). It turns out (how I didn't spot this before I can't imagine) that with the 55-250 STM, even when there is no question of the flash being able to provide enough light or a small enough amount of light, if you increase the aperture and/or increase the ISO then the exposure goes up, and if you decrease the aperture and/or decrease the ISO then the exposure goes down. So, for any particular aperture, there is only one ISO that will give the required exposure, and if you change either the aperture or the ISO you must change the other one in the opposite direction to balance things out. It is like this not just with the Metz flash - it is the same with the on-board flash.

Presumably therefore, the 55-250 is faulty (I can't believe that this is a "feature", but maybe I haven't read the small print!)

I bought the lens, along with the 70D and the 18-55, from a camera shop in Bristol in February last year. I will be going back there in the New Year.

I don't think this changes anything in terms of small subjects which need the use of flash. My recent posts here have covered that ground. But it does open up the options for larger subjects that require flash, for example using the 105, by itself or with one of the milder achromats, or using a replaced/repaired 55-250 with one of the milder achromats.

 

[GardenersHelper]

I need to work out what kit to use for botanical work next season, almost all of which will be working with natural light. Before Christmas I decided to try some comparisons in carefully controlled conditions. I started by considering three options:

• The Sigma 105 Macro on the 70D
• Low powered achromats (Canon 500D and Raynox 150) on the 55-250 STM on the 70D
• Low powered achromats on the 45-175 on the G3

I collected some small flowers and some leaves and twigs from the garden. I usually do test shots in my study but this has a suspended floor and this ruins long exposure natural light shots of the type these comparisons would involve. So I arranged the subject matter on a table in the kitchen, which has a solid floor. The table was near a large window. It was the middle of the day, but heavily overcast.

I used a tripod and wired remote release. Image stabilisation was off. I tried to align the shots in the same way for each rig, but the G3 and the 70D have different aspect ratios so I could not capture identical shots. I attempted to make the vertical coverage of the scene the same for each rig, with the 70D shots then having some extra material on each side. I was not entirely successful with the alignments, but on reviewing the images I thought the alignments were good enough for my purposes.

I used manual focus, using the magnified view in live view. I took shots at three magnifications of 1:1 (in APS-C terms), 1:1.2 and 1:2. I took the shots first with the Sigma 105 on the 70D, setting the magnification on the barrel, and then tried to arrange the other shots by referring to the images created with the Sigma 105. Each shot had a small daisy in it, and I tried to focus on the same part of the daisy every time.

I tried to get the exposure level similar for each shot by adjusting the exposure compensation a little when a shot appeared to have drifted from the level I was working at. However, that didn't work very well and I had to apply additional exposure adjustments in post processing to bring the images to a better consistency of brightness.

I was interested in sharpness, dof, clarity, bokeh, colour rendition and noise. I took twelve sequences of shots:

70D+Sigma 105 at 1:1, ISO 100 at f/2.8 to f/22
70D+Sigma 105 at 1:1, f/14, ISO 100 to 12800
70D+Sigma 105 at 1:1.2, ISO 200, f/2.8 to f/22
70D+Sigma 105 at 1:2, ISO 400, f/2.8 to f/22

70D+55-250+Raynox 150 at 1:1, ISO 100, f/5.6 to f/22
70D+55-250+Raynox 150 at 1:1, f/14, ISO 100 to 12800
70D+55-250+Raynox 150 at 1:1.2, ISO 200, f/5.6 to f/32
70D+55-250+Raynox 150 at 1:2, ISO 400, f/5.6 to f/28

G3+45-175+Raynox 150 at 1:1 (APS-C equivalent), ISO 160 (G3 base ISO), f/5.6 to f/22
G3+45-175+Raynox 150 at 1:1, f/14, ISO 160 to 6400
G3+45-175+Raynox 150 at 1:1.2, ISO 200, f/5.6 to f/22
G3+45-175+Raynox 150 at 1:2, ISO 400, f/5.6 to f/22

The images were captured as RAW and imported into Lightroom using my standard import settings. None of the images has had any luminance noise reduction. The import settings include import sharpening, which is set slightly differently for 70D and G3 images (the G3 images being sharpened slightly more, but I checked the effect of adjusting a 70D image to the G3 sharpening and the effect was inconsequential – pretty much indiscernible actually, at 100% viewing.)

For the noise test sequences (the ones with varying ISO) I took one shot of each setting. For the other sequences I took three shots at each setting. I looked at several of these triplets and could not discern any difference between the shots, so I simply took the third shot from each sequence and used that.

Apart from the (very mild) import adjustments, the images had a camera profile applied and a white balance taken from a white card reference shot (separate reference shots for the 70D and G3).

As mentioned above, I made some alterations to the Exposure in Lightroom to try to even up the brightness as between shots in a sequence and between the various sequences so as to aid comparisons.

The images were exported from Lightroom, resized to a common height of 3448 pixels. This was the pixel height of the G3 images, the 70D images being 3670 pixels in height. The resizing was done to make like for like comparisons more meaningful at 100% or other magnifications. The images had no output sharpening. They were uploaded to this album at Flickr.

I compared pairs and triplets of images and ran through sequences of images using Faststone Image Viewer. This let me pick pairs, triplets etc to work on and flick from one to the next of the selected images just using the arrow keys, keeping my eyes on the images, and I could do this with the images fitted to the screen or viewing 100%.

In this way I could do comparisons such as

• One image from each rig at 1:1.2, comparing the clarity, sharpness, colour etc looking at the whole image, and comparing sharpness and detail at 100%
• For a particular rig and magnification, running through the apertures to see how the dof and sharpness/detail changed
• For a particular rig, at 1:1 running through all the available ISO values to see how the noise characteristics changed with ISO
• Comparing the noise between rigs for a particular ISO

This produced some unexpected results, and after Christmas I decided to do another test run. I used the same technique, but this time I added a fourth rig – my FZ200 bridge camera, used with low power achromats. This time I captured all the shots at base ISO (ISO 100 for the 70D and FZ200, ISO 160 for the G3). I used four magnifications of 1:1, 1:1.5, 1:2 and 1:4. I obviously needed new material for the test scene, but I made it fairly similar to the scene used in the first run. In particular, it had a small daisy in all the scenes and I tried to focus on the same part of it for all the shots.

Here are the four test scenes from the second test run, as captured by the 70D and Sigma 105.


0644 5.01 Scene magnifications by gardenersassistant, on Flickr

I used magnified view in live view for focusing for the 70D and G3, and also for the 1:1 shots for the FZ200, but after that I used autofocus for the FZ200 shots because manual focus proved too awkward and tiresome to be useful.

For the second run I used the same exposure compensation of -1/3 for every shot. On importing to Lightroom every shot had the same, mild, colour noise reduction and the same, mild sharpening, and no luminance noise reduction or output sharpening. I adjusted the Exposure of a few images which seemed much darker than the rest.

I created a camera profile for each of the four rigs using a Macbeth colour chart shot, and as well as applying the appropriate camera profile to each shot I set the white balance for each shot using the white square on the appropriate camera profile shot. I exported the images from Lightroom 3000 pixels high, which is the pixel height of the FZ200 sensor, the G3 and 70D having larger pixel heights.

The images from the second run (which includes the composite images in this and subsequent posts) are in this album at Flickr.

It turned out that despite using camera profiles and white balance shots the colours were significantly different between the 70D on the one hand and the two Panasonic cameras on the other.

I saw that the Temp and Tint value set from the white balance shots varied between cameras (and between the 70D with the Sigma and the 70D with the 55-250). I tried using the same Temp and Tint values rather than camera-specific values, but even when I did this significant colour differences remained, as illustrated here.


0644 5.02 1to1 Equalised White Balance by gardenersassistant, on Flickr


I am surprised, and disappointed, about this. I thought I had got colour issues sorted.

Anyway, on now to what I saw when I compared the images.


Continued in next post

 

[GardenersHelper]

Continued from previous post


There are several well known facts of relevance to capturing close-ups with different equipment

• Beyond a lens' “sweet spot” aperture, image sharpness decreases as aperture decreases, because of the effects of diffraction
• A prime macro lens will produce sharper, more detailed images than a general purposes zoom lens, especially if that zoom lens has an additional piece of glass on the front of it such as an achromat
• If the same scene is captured with two rigs using the same aperture, the rig using the smaller sensor will have a greater depth of field.

It turns out that none of these statements is quite as straightforward as it seems.

Let's start by looking at the 1:1 scene. (I have used visual examples from the post-Christmas, second test run, but what I saw was consistent with what I saw in the first test run.)

Here is the 1:1 scene as captured by the 70D+Sigma 105 at f/22. The three red rectangles show the areas used below for 100% comparisons.


0644 5.03 1to1 showing crops by gardenersassistant, on Flickr

Here are crops of the area on the daisy. This is the area on which I tried to focus each shot. The FZ200 crops are 100%, the others are less than 100% because of the other cameras' larger number of pixels, and my priority was to be able to compare the same areas of the images, looking at the areas the same size from each rig.


0644 5.04 1to1 Max focus crop x aperture by gardenersassistant, on Flickr


Notes/caveats. In the following I am using “sharpness” as shorthand for sharpness/detail/definition/microcontrast as appropriate, and I am describing what I see, on balance, when comparing crops from these images (it is “on balance”, because with some comparisons it depends where you look in the crops as to which seems the sharpest – and there are judgement calls in producing an overall “which is sharpest” assessment in such cases). I am also not comparing sharpness between 100% crops, but sharpness when looking at crops which are the same shape and the same proportion of an image's size, but where the pixel height has been adjusted in most cases to allow for easier “like to like” comparisons between rigs with different sensor sizes.


For the daisy crop, for the 70D+Sigma 105, the sharpest crops are around f/5.6 to f/8. The Sigma 105 goes to 2.8, but opening up the aperture beyond f/5.6 does not improve the sharpness. Here are some 100% crops from the 70D+Sigma.


0644 5.04a 70D+Sigma105 1to1 100pc crop largest apertures by gardenersassistant, on Flickr

For the 70D+Sigma, f/11 is not quite as sharp as f/8, and f/16 is clearly less sharp than f/11. F/22 is very much less sharp than f/16.

For the 70D+55-250 with the Raynox 150, sharpness increases from f/5.6 to f/11 and then declines very slowly from f/11 to f/16 to f/22. At f/5.6 and f/8 it is less sharp than the 70D+Sigma 105. By f/16 it is slightly sharper than the 70D+Sigma 105, and at f/22 it is significantly sharper.

The G3_45-175 with the Raynox 150 has similar sharpness at f/8, f/11 and f/16, and is less sharp at f/5.6 and f/22. It is less sharp than the 70D+Sigma 105 at f/5.6 and f/11, and sharper than the 70D+Sigma 105 at f/16 and f/22.

The FZ200+Raynox 150 is sharpest around f/4 and f/5.6. Comparisons with the other rigs are tricky because of the big difference in sensor size, but at its minimum aperture of f/8 the FZ200+Raynox 150 is sharper than the 70D+Sigma 105 at its minimum aperture of f/22, and similarly for f/5.6 on the FZ200 compared to f/16 on the 70D+Sigma. At its maximum aperture of f/2.8 the FZ200 is less sharp than the 70D+Sigma at f/5.6, as it is at f/4 compared to f/8 to f/11 on the 70D+Sigma.

In summary, when looking at the area of the images which is in sharpest focus, the 70D+Sigma 105 is sharper than the other rigs at large apertures, but is less sharp than them at smaller apertures.

I was surprised by this finding. I found it in the first run and it was the main reason why I decided to do the second run, to check that it was genuine. It does seem to be.


Next we look at a crop from the area below the flower and to the left of it. This is what I think of as a “midground” area – not fully sharp, but not extremely soft.


0644 5.05 1to1 Mid focus crop x aperture by gardenersassistant, on Flickr

Here the comparisons are very different. The 70D+Sigma 105 is at its sharpest at around f/16 and is sharper than all the other rigs at every comparable aperture. Indeed, the least sharp of the 70D+Sigma images is as sharp as the sharpest of the 70D+55-250+Raynox 150 images. While its sharpness advantage over the G3 and FZ200 is not quite so extreme, it is still very marked.


Now we look at a crop from the area below the flower and to the right of it. This is what I think of as a background area – extremely soft.


0644 5.06 1to1 Min focus crop x aperture by gardenersassistant, on Flickr

Here the 70D+Sigma is at its sharpest at f/22. As with the midground, it is sharper than all the other rigs at every comparable aperture. At f/16 and f/22 it is markedly sharper than any of the other rigs at any aperture.

In summary, when looking at areas outside of the areas of sharpest focus, the 70D+Sigma 105 is sharper than other rigs at every comparable aperture, in many cases markedly sharper.

Going back to the three “well known facts” mentioned above, the first was this:

Beyond a lens' “sweet spot” aperture, image sharpness decreases as aperture decreases, because of the effects of diffraction

This is clearly the case with the Sigma 105 on the 70D as far as the sharpest areas in the image are concerned. However, it is not the case with other areas of the image, either with the Sigma 105 on the 70D or with the other rigs – they all get sharper as aperture decreases. And on the other three rigs, the decline in sharpness of the sharpest areas in the image as aperture decreases is much less marked than with the Sigma 105 on the 70D.

The second well known fact was this:

A prime macro lens will produce sharper, more detailed images than a general purposes zoom lens, especially if that zoom lens has an additional piece of glass on the front of it such as an achromat.

At least with these rigs, this turns out not to be true for all apertures in the areas of the image that are most in focus. At the smallest apertures, in the areas of the image that are most in focus, this particular prime macro lens, the Sigma 105, is less sharp than general purpose zoom lenses with achromats on them, on the same camera as the Sigma 105, and on different cameras.

For these rigs the prime macro lens is sharper than all the other rigs for areas outside of the areas that are most in focus.

The third well know fact was this:

If the same scene is captured with two rigs using the same aperture, the rig using the smaller sensor will have a greater depth of field.

We'll look at this one in the next post.


So, I was expecting the 70D+Sigma 105 to be produce the sharpest/best defined/most detailed images across the board. It doesn't. The situation is more complicated. So how do these complications pan out in terms of images seen as a whole, rather than examined in the minutest detail? Should I drop the Sigma 105, like I dropped the Canon 100L? (which I dropped for different, and less carefully researched, reasons)


Continued in next post

 

[GardenersHelper]

Continued from previous post


Notes: As well as looking at the composite images posted here (which can be seen larger over at Flickr) I have compared the individual images in these composites, with the individual images fitted to my quite large screen so as to get a better look at them. All the individual images used here (and a number of others) are available in the Flickr sets linked in the previous but one post.

Here is the whole 1:1 image as captured by each of the four rigs at broadly comparable, largish apertures. (There is a fair degree of approximation here in terms of comparability, otherwise known as “equivalence”. The G3's micro-four thirds sensor is smaller than the 70D sensor, so a comparable aperture in terms of dof and diffraction would be a bit smaller for the 70D than the G3, rather than the same aperture as used here. On the other hand the FZ200 probably needs an aperture larger than f/2.8 to be comparable with f/5.6 on the other rigs, but f/2.8 is the largest aperture available on the FZ200.)


0644 5.07 1to1 large aperture comparison by gardenersassistant, on Flickr

To my eye, the 70D+Sigma 105 image is preferable. It looks to have better clarity, and looks somehow “cleaner”.

The 70D+Sigma image also appears to have more dof than the G3 image. But the G3 image used the same aperture, and the G3 has a smaller sensor than the 70D, so by the third “well known fact” the G3 image should have more dof. Now, there may be a techical argument about what dof really means, and that what we are seeing there isn't actually greater dof (because the not in focus areas may be better defined, but they aren't in focus, so they aren't in the dof). But in practical terms it looks like more dof to me. And more importantly, irrespective of the correct terminology, it looks more pleasing to my eye.


Here is the whole of the 1:1 image as captured by each of the four rigs at broadly comparable, smallish apertures.

0644 5.08 1to1 Small aperture comparison by gardenersassistant, on Flickr

To my eye, exactly the same comments apply. I prefer the Sigma 105 version.


Here is the whole 1:4 image as captured by each of the four rigs at broadly comparable, largish apertures.


0644 5.09 1to4 Large aperture comparison by gardenersassistant, on Flickr

This time the smaller sensor cameras do seem to have slightly more dof. For example, the purple flower is softer on both of the 70D rigs compared to the Panasonic rigs, while the patterns of the green leaves in the background at the top of the image look better defined in the Panasonic images.

As expected given the better sharpness of the Sigma 105 in areas that are less in focus, the Sigma 105 seems to have more dof on the 70D than the 55-250+Canon 500D.

I find it difficult to pick a “winner” from these four images. On balance, I think I prefer the G3 and FZ200 images because of the better definition of the purple flower, but there really isn't much in it.


Here is the whole 1:4 image as captured by each of the four rigs at broadly comparable, smallish apertures.


0644 5.10 1to4 Small aperture comparison by gardenersassistant, on Flickr

In this case too I don't have a strong preference. I think I prefer the 70D+Sigma 105 version because it seems to have slightly more clarity/definition/micro-contrast or some such, but here too there really isn't much in it to my eye.


Here is the whole of the 1:2 image as captured by each of the four rigs at broadly comparable, middling apertures.

0644 5.11 1to2 Mid aperture comparison by gardenersassistant, on Flickr

In this case the 70D+Sigma 105 version looks much the best to me because of the better definition//detail/clarity/texture of the tracery of the decaying leaf, the purple flower and the flower head at the front. This crop illustrates the differences in definition/detail and clarity quite clearly to my eye.


0644 5.12 1to2 Mid aperture crop comparison by gardenersassistant, on Flickr


Overall

Comparing other pairs, triplets etc of images from both test runs it seems to me that, on balance, the 70+Sigma 105 produces a greater proportion of the images that I prefer, but I do prefer versions captured with other rigs some of the time.

In summary, these two test runs suggest that, for botanical subjects, the Sigma 105 on the 70D may more often than not produce results that please my eye more than equivalent results from the other rigs. However, the 70D+Sigma is not uniformly better, and it looks as though in many cases even where the 70D+Sigma is better (to my eye) it may not be by much.


Usability/ergonomics

I have concentrated in these posts on image quality/characteristics, and haven't discussed ergonomics. Briefly, it seems to me thus far at least that, for botanical work, ergonomic considerations favour the 70D+Sigma 105. The lens and camera work very well together for manual focusing, which I am finding increasingly important for botanical work, and allow for very fluid capture workflow.


Conclusion

I will keep the Sigma 105, and will start next season with it as my preferred rig for natural light shots of botanical subjects down to 1:1 (which is most of them). I am though a little disappointed that it does not seem to be as much better for those subjects as I had hoped (and to be honest had expected) it to be in terms of image quality.

 

[GardenersHelper]

I will keep the Sigma 105, and will start next season with it as my preferred rig for natural light shots of botanical subjects down to 1:1 (which is most of them).

Ho hum. I might have been a bit premature with that. I've done some more tests/comparisons. While these tests confirmed that the Sigma 105 is (IMO) capable of producing better image quality than any of my other rigs for botanical subjects using middling and large apertures, they have revealed some other factors which may stop me using the Sigma 105 as much as I thought I would. I'll discuss that in the next post. First, the recent comparisons of image quality.

Two days running I captured images of a Camellia flower in the garden. I used the same four rigs I have used before, 70D with 105 macro and 55-250, G3 with 45-175 and FZ200. This time I used no achromats. On the first day I captured images using various apertures, from smallest to largest. The second day I concentrated on the apertures that seemed (for my taste) most appropriate for the subject, centring on f/11 for the 70D and G3, and f/4 for the FZ200 which is roughly equivalent to f/11 on the other cameras for dof and diffraction.

It was breezy on both days and this made capturing precisely like for like images difficult. For example, I wanted to compare the rigs at base ISO, but despite the fact I had multiple attempts for each combination of rig, aperture and ISO, on the first day (when the light level was very low) I failed, because of the breeze, to get a sharp image from the G3 at base ISO, so I had to use an ISO 800 shot instead. More generally, I picked what seemed to me the best one from each combination of rig, aperture and ISO, but even so there are probably variations from the best of one combination to the best of another in terms of how much they were influenced by the breeze.

There are other reasons to be cautious about drawing too firm a conclusion from any individual comparison. For example, despite my best efforts I failed to get all the shots aligned the same, and on the second day the light level varied as the cloud cover came and went. Also, the flower was in a very different state on the two days. On the first day the anthers were heavily laden with pollen, which provided a lot of fine detail to examine, but on the second day there was virtually no pollen in evidence.

There are also issues to do with post processing, for example to do with sharpening. Different images from different rigs at different ISOs can benefit from different amounts of sharpening. Is it “like for like” to sharpen images individually, or should you just look at them with the same sharpening for each, or perhaps “out of the camera”? (but does “out of the camera” even make sense with raw captures, which these all were?)

When trying to see how much detail the images have captured pixel peeping is, IMO, appropriate. However, there are issues too. Should you use actual 100% crops or, since the images were deliberately similarly framed, should you look at the same area in each image and adjust those areas to a common size to aid comparison?

I chose to sharpen each image as seemed suitable for that image, so as to bring out the most detail that I could when looking at the image full size. (That revealed btw some oversharpening when viewing at 100%, but here too, is that relevant if it doesn't show up at normal viewing size?) I also chose to resize whole images to a common pixel height to aid comparisons, and similarly for the heavy crops.

Sets of whole images (with some mild cropping to align them better) and heavy crops (100% for the FZ200 and somewhat less for the other cameras) are here for shots of the single flower (captured on both days), and here for shots of the flower in its context of leaves and buds (as captured on the second day, because I was not methodical enough on the first day to have worthwhile comparisons to show).

After a lot of poring over the linked images, and others too, it seems to me, despite the sometimes rather significant variations, that when used around f/8, f/11, f/16, the 70D plus Sigma 105 does produce results that are more pleasing to my eye when viewed as whole images and do contain more detail when looked at closely (pixel-peeped).


Continued in next post