Wednesday, 2 June 2021

So is the A7Sii better?

I bought the A7sii on the grounds that it might produce better results than the A7ii because of its smaller number of larger pixels. The suggestion had been that post processing involves floating point arithmetic and this involves errors that can accumulate along a chain of operations. Starting with better data for individual pixels might cause the calculation errors to be smaller and accumulate slower, giving a better final result.

In this post at photomicrography.net Rik Littlefield agreed that starting with better pixel level data might be advantageous for some post processing operations, although not mainly for reasons of floating point error accumulation. 

So it doesn't seem like an intrinsically silly idea that my heavily processed images might end up better if starting from better pixel level data.

Within 10 days of getting the A7sii I had had seven sessions with it, and I was pleased with the results I was getting. I was curious as to whether these were better than I could get with the A7ii; had my purchase been worthwhile in terms of image quality? I know from experience that doing effective comparisons can be very difficult, even if you can reduce a lot of the variability by doing like for like tests in a controlled environment. However, I also know how difficult it is to draw real world conclusions from highly controlled comparisons. And it was real world results that mattered to me. So I tried a real world comparison.

Since image by image like for like comparisons were impractical out in the field, I decided to try comparisons at a session level. In the previous four days I had had three sessions with the A7sii with the 100mm Laowa 2X macro lens with 2X plus 1.4X teleconverters, using f/40 for the whole of two of the sessions, f/45 for one session and f/36 and f/32 for the other session. All four sessions were in the church grounds opposite our house.

I decided to do another session in the church grounds, using the same setup apart from using the A7ii rather than the A7sii. I would use f/40 throughout the session. The images I kept from the session are in this album at Flickr. The images from the four sessions with the A7sii are here, here, here and here at Flickr.

And my conclusion?

I can't convince myself that, at the session level, what I got from the A7ii session is any worse (or any better) than from the four A7sii sessions. There may be differences that someone more observant than I am might notice, but as it stands, I'm not seeing it.

So do I regret getting the A7sii?  No, I'm comfortable using it, I like the results I'm getting from it, I think it is giving me stronger focus peaking signals than the A7ii, although even there I'm not 100% sure about this; it is something else that is rather difficult to test in real world use. But if it is giving better focus peaking signals that will probably be improving my focus success rate. Having the A7sii means that the A7ii is freed up for other things. And if I hadn't got the A7sii I would have had a permanent niggle in the back of my mind that perhaps I could do better by using a low pixel count camera.

And I have this feeling, which may of course be no more than wishful thinking, that it may in fact be giving me better results, at least some of the time, but I just can't put my finger on it. Anyway, I'm entirely content to go forward with the A7sii as my primary camera for photographing invertebrates.


Using less teleconversion

 Using the A7ii and subsequently the A7sii with the Laowa 100mm 2X macro and a pair of 2X teleconverters opened up new possibilities for me. Previously I had occasionally managed to get a half decent image of small subjects such as globular springtails and aphids, but it was difficult and had a very high failure rate, so for the most part I avoided subjects that small. Now I found I was able to capture better images of these small subjects, and to do so with a much higher success rate. My photographic opportunities had widened significantly.

However, there turned out to be a downside at the other end of the spectrum; photographing larger subjects had become problematic, especially more "environmental" shots of larger subjects from further out. The problem was working distance. The problem is usually having not enough of it. Now the problem was having too large a working distance. At 1:1 the working distance was around 380mm and was increasing fast as the magnification decreased any further. This large working distance made it difficult to throw enough light on to the scene (and I need to throw a lot of light on to the scene because of the very small apertures I'm now using).

There were two problems. One was simply the distance. Illumination decreases with the square of the distance between the light source and the subject, and by 1:1 the illumination had become very diluted. And that was if the flash heads were pointing at the subject. But at lower magnifications they were not, and that was the second problem. If the flash heads were pointing so as to illuminate small, close subjects then they were far away from pointing at distant, larger subjects and not providing them with much light, necessitating uncomfortably high ISOs. And if the flash heads were adjusted for larger subjects the smallest subjects got no illumination at all. It is true that adjusting the direction the flash heads were pointing had become much more feasible with the Yongnuo YN24EX than it had been with the Venus Optics KX800, but making any adjustments to the physical flash setup was very unattractive; it would be very disruptive for some of the sequences I like to capture, where I move freely between a wide range of magnifications, sometimes taking shots as frequently as every two seconds or so while change magnification from shot to shot.

I went back to my close-up lens setups and did some measurements to give me some context. I used my FZ330 small sensor bridge camera with Raynox 150 and Raynox 250 close-up lenses, as this was my most used setup, and the one with the widest range of magnifications.

The Raynox 150 was by far my most used setup with the FZ330 (sometimes I would have sessions with the FZ330 where I only used the Raynox 150 and didn't use the Raynox 250 at all). This let me photograph scenes from around 75mm wide down to 13mm wide. Using the Raynox 250 instead of the Raynox 150 gave me scene widths from 43mm down to 8mm. So by changing between the two close-up lenses I could photograph scenes from 75mm wide down to 8mm wide.

In contrast with the A7ii, the 100mm 2X macro and a pair to 2X teleconverters:

  • I could go down to a scene width of 4.5mm, much smaller than with the Raynox setups. (I had more powerful Raynox close-up lenses, and I could stack Raynoxes for example a Raynox 150 and a 250, or two 250s. However, I had found it difficult to get good results with these setups.) 
  • But I was limited for larger scenes to around 36mm scene width, which was very restrictive compared to the 75mm scene width I could get with my much used Raynox 150.
I did some measurements on the A7sii with 100mm 2X macro with one 2X and one 1.4X teleconverter. This would give me a minimum scene width of 6.4mm, which was still slightly smaller than I could get with the FZ330 and Raynox 250. At the maximum feasible working distance of around 380mm it gave me a scene width of around 53mm. That was not as good as the 75mm scene width I could get with the Raynox 150 on the FZ330, but it seemed like swapping out one of the 2X teleconverters and using a 1.4X in stead would give me a closer fit to what I needed.

So I experimented. It turned out that there was an adjustment I could make to the bendy arms holding the Yongnuo flash heads that brought them down a bit and let them point a bit more forwards and a bit less downwards and that was enough to let me use a slightly longer working distance, and that was all I needed to get to the sort of scene width I could get with the FZ330 and Raynox 150. So with the 2X plus 1.4X teleconverter setup I could cover the full range of scene sizes that I could with the FZ330 and Raynox 150 and 250, and could do so seamlessly, unlike with the Raynoxes which required changing close-up lenses to go between magnification ranges.

The other question was whether the 2X plus 1.4X teleconverter setup would give me enough magnification for what I wanted to do.  Although the double 2X teleconverter setup gave me up to 8X magnification, I had been feeling for a while (without doing any careful measurements) that I was rarely using much beyond 6X or so. It turned out in practice with the 2X plus 1.4X setup that I didn't often feel I was missing the extra 5.6X to 8X magnification range that the double 2X setup gave me.

So I decided to transition to a 2X plus 1.4X setup. So far it has worked well. Indeed, for reasons I don't understand, it turns out that I am using lower ISOs with this setup than the double 2X setup. I think that may be down to the adjustments I made to the position and direction of the flash heads. I'm sometimes using ISOs as low as 400, which gives me a three stop noise advantage over the FZ330 setup,  for which base ISO 100 is similar to ISO 3200 on full frame. And I'm rarely using ISOs as high as 3200, so there is at least a little noise advantage over the FZ330 setup for almost every shot, and sometimes a rather larger noise advantage.


Monday, 24 May 2021

Reaching the limit of depth of field

 I think I have reached the limit of the depth of field I can get from single shots. I might get more in due course if the software I am using improves or I get better at using the software. For the moment though the limit appears to be at around f/45 when using a pair of 2X teleconverters with the Laowa 100mm 2X macro. (That f/45 is for the camera/lens - the lens is set to f/11 and the camera knows that there is 4X teleconversion in use and adjusts that f/11 to f/(11*4), and shows an f-number of f/45).

I have two reasons for thinking that this is the limit.

First, I did a session with f/57, which is two thirds of a stop up from f/45, and I didn't want to keep any of the images from the session. Despite the heavyweight processing they were still too soft for my liking and lacked some detail that I had been getting in similar shots at f/45, for example in ants' eyes.

Second, a very helpful and insightful analysis of one of my tiny aperture images by Rik Littlefield. Rik is immensely knowledgeable in this area. He is I believe the author of Zerene Stacker, which is one of the two highly specialised focus stacking software applications. He also runs the photomacography.net web site. Rik is very actively involved with discussions on the web site and has a talent for explaining complicated issues in accessible language for people (like me) who don't have the background to understand the underlying optics and mathematics.

Rik's analysis was in this thread, in which I had posted five of my recent tiny aperture images. (I don't post much on that site, but I'm very glad I did this time round.) Another forum member asked me to do a comparison between a focus-stacked image and one of my tiny aperture images of the same subject, which I did. (I'm not very skilled at focus stacking at the macro scale, but fortunately what I managed was good enough for the purpose in hand.)

In this post Rik did a detailed comparison of a couple of areas in the images and showed that even at the reduced image size of 1300 pixels high it was not just that the tiny aperture image was losing fine detail, but that for the detail that remained it was rearranging some elements of it so that it was different from the actual structure of the subject. However, you had to look very carefully at a like for like comparison with a more accurate rendition to tell this was going on. As Rik put it, 

I don't see that any of this impacts Nick's use of this technology for his purposes. Small aperture one shot is the only technique I know for capturing high DOF images of small active critters. But it does seem clear that this image has reached beyond the limit of accurately capturing fine detail, even at the 1300 pixels size.

I'm content to live with this fine detail inaccuracy as it doesn't in my view significantly alter the overall impression of the scene and the subject. And given my approach of "pretty pictures" and "credibility" rather than "accurate records" and "authenticity" I think that is good enough. But, coupled with my f/57 experience, I do think I'll draw the line at f/45 for now. 

The curious issue remains that using a constant camera/lens f-number of f/45 seems to work even though this means that the effective f-number is changing as the magnification changes, and so the depth of field and diffraction softening are changing too. Why this constant f-number should work is a mystery to me. It may be that I haven't looked at this carefully enough. I'll try some more variations away from f/45 to see if I can get more depth of field at some magnifications. If I can then I can adopt a more nuanced approach to setting the camera/lens f-number, although to be honest the simplicity of using a single f-number is very appealing because it lets me react faster to what is going on.

Update 28 May 2021. 

User ThrillaMozilla has commented very helpfully on this mystery in this post at dpreview.com:

ThrillaMozilla wrote:

gardenersassistant wrote:

It is very convenient to not adjust the aperture as the magnification changes. Given that this means the effective aperture changes with magnification, I don't understand why this works as well as it does in terms of apparently (difficult to test) getting a similar depth of field in relation to subject size whatever the magnification.

Image-side geometry to the rescue.

It's late, but if I did the math. correctly on the small scrap of paper, for a given focal length the diffraction blur diameter and the depth of field are both proportional to 1+1/M. How convenient. You want them both to stay proportional to the size of the sample, hence constant with changes magnification. For M between 1.5 and 5.6 that varies between 1.18 and 1.67 -- thus constant enough. It's just your good luck that it works out that well. 


Tuesday, 18 May 2021

New kit - Collecting evidence Part 2 - flowers etc

 The third of the three new pieces of kit I was thinking about was a micro four thirds Panasonic Leica 50-200. In this post I said that "For flowers etc I need to do some practical comparisons between the 60mm macro and the 45-175." 

Three weeks ago I photographed 9 flowers hand-held in our garden using a Panasonic 45-175 on a G9, with and without a Canon 500D close-up lens depending on the size of the subject, and with an Olympus 60mm macro on a Panasonic G9. 

I used aperture bracketing. With the Olympus 60mm macro this gave me a set of 7 shots for each shutter press. With the 45-175 each aperture bracket set contained 5 shots.

For each subject I selected the image that appealed to me most from each camera and processed the selected images from raw using my then current flower workflow of Lightroom followed by Topaz DeNoise AI.

For each subject I selected the image that I liked best. This turned out to be four captured with the 45-175 and five captured with the 60mm macro. (They are in this album at Flickr.)

It is possible that the 50-200 would render colours, textures and out of focus areas better than either the 60mm macro or the 45-175, but I obviously could not test this. What I did expect to see was a difference in the way backgrounds were rendered, with the longer focal lengths I typically use with the 45-175 throwing the backgrounds more out of focus, but looking at the test shots nothing of that sort jumped out at me.

Basically, in this test I did not see any evidence of a particular advantage of using a telezoom with/without a close-up lens compared to the macro lens. It looked like any significant benefit of a 50-200 would have to come from better rendering of colours, textures and out of focus areas. But about that I had no evidence to go on.

That was for stills. For focus stacking the 50-200 might be pretty much as good as the 60mm macro, although a lot of the time it would be using a slightly smaller aperture than the f/2.8 that I use with the 60mm macro for focus stacking, and so backgrounds would be slightly less out of focus than with the 60mm macro (possibly counteracted by the effect of using a long focal length) and in some cases shutter speed would be slightly slower or ISO slightly higher.

What was definitely the case was that at 655 grams the 50-200 is (to me) significantly heavier than the 60mm macro at 186 grams.  It is also expensive, at almost £1,500.

To my mind the potential benefits are too much in doubt to outweigh the definite disadvantages, especially as I have no particular problems I need/want to solve with the Olympus 60mm macro. In fact, I rather like using it and am content with the image quality I get from it. I will therefore not be pursuing the option of a Panasonic Leica 50-200.

So for now, I have no additional hardware under consideration.

Monday, 17 May 2021

A more substantial test of the A7sii

 Yesterday I spent two hours in the Church grounds opposite our house testing the A7sii. For the first time this year there were plenty of subjects around. Mind you, I wouldn't necessarily have thought that if I had been using one of my close-up lens setups, because a lot of the subjects were rather small compared to what I found comfortable to deal with using my close-up lens setups.

The session convinced me that I had not been my imagination - I really did seem to be getting stronger focus peaking signals, and easier focusing because of it, including a good degree of success with subjects that were moving around, two of which were rather small and moving a lot.

The post processing was unproblematic. I have a feeling that the raw files may be a bit more malleable than with the A7ii. Looking at the outputs I also get the impression that the A7sii may be picking up more from dark areas that are far from the flash. I did not raise shadows for any of the images beyond what is done by my presets. Nonetheless, I'm not seeing the deep black areas that I dislike.

I captured 529 shots in the session. The 93 that kept (over 17%) are in this album at Flickr. Here are several images from the session.

I don't have much experience yet with the A7sii, so perhaps there are problems I haven't registered yet, but as it stands I expect the A7sii to take the place of the A7ii as my camera of choice for invertebrates.

 













A surprising new camera

 In this post a month ago I mentioned that I was considering two new cameras, a Sony A7r ii and an Olympus E-M1 III. In this post 10 days ago I wrote "So far it is going well enough with the A7ii that I'm not inclined towards the Olympus E-M1 III or the Sony A7rii at this stage. I'm going to put those possibilities on hold for now."  

I've gone further than that now; I really don't think either would work for me. 

As far as the A7rii goes, yes I could use larger apertures to get less diffraction softening than with the tiny apertures I'm using at the moment, but I would want to capture from further away and crop in order to get back the depth of field I had lost by using the larger aperture. With the image quality loss from a big crop (and it would have to be a big crop to make sense), I strongly suspect this approach would simply take me back to roughly where I am with the A7ii. And yes, I could use larger apertures to get better resolution, and put up with the reduced depth of field. But I know I wouldn't be happy with reduced depth of field, and in any case if I want to do this I can do it with the A7ii. All in all, I really don't think the A7rii would be useful for me.

And the Olympus E-M1 III? The potential benefit is with focus stacking for invertebrates using flash, for which none of my current kit is suitable, at least not for my preferred way of working - hand held. The thing is though, I am continuing to get results I like with tiny apertures, and finding that it works with subjects that are moving around, or on foliage that is blowing around in the breezes that are so common here, or grooming, or dealing with prey, or struggling to release themselves from a spider's web, or blowing bubbles ...... in fact anything that involves movement or action of some sort; those are the opportunities that attract me the most, and focus stacking isn't suitable for any of them. The more I think about it, the more I have to conclude that an E-M1 III would not be useful for me.

With my mind released from thinking about those two bits of kit it wandered off to a claim I had heard 8 months ago. It was in this thread that I started in the Photographic Science and Technology forum at dpreview.com. (Scary place, there are some fearsomely knowledgeable people there. It was the first time I had ever dared to post there.) I wanted to know whether, compared to my Sony A7ii, a camera with the same sensor size but fewer, larger pixels might be better for my purposes. This generated a lot of discussion, most of which I didn't understand, with much of it, as far as I could tell, not being relevant to my particular issue. 

However, in this post a fascinating possibility was raised. It was argued about, as to whether it was true or not, but I got the impression that the general idea was agreed by at least one person who is deeply knowledgeable about such things. The claim was (in my translation) that when dealing with an image captured at a low light level, what you can get out of post processing depends on how large the pixels are; the larger the better.

Now, my A7ii has a 24 megapixel sensor. In contrast, the Sony A7sii has half the number of pixels. Despite this, for a given scene and a given aperture and shutter speed (or in my case for a given aperture and flash level), both cameras will capture the same amount of light, and the image quality, the noise particularly, depends on the amount of light captured. 

However, post processing is done on individual pixels. Post processing involves arithmetic using floating point numbers. Each time a calculation is done for a pixel there will be one or more rounding errors. With repeated calculations these rounding errors will accumulate and the result may go further and further away from what it should be. Critically, the extent of this degradation depends on how good the data was that was captured by each pixel. 

Each pixel in the A7sii will capture twice as many photons as a pixel in the A7ii. There will therefore be a better starting point for the calculations, and the degradation will be less.

That at least is what I thought I could infer from those parts of the argument that I thought I understood. As you will gather from the way I phrased the previous sentence, I was by no means certain about this, but it did seem to make sense to me.

The thing is, because I use such small apertures I can't throw enough light on my scenes to get anywhere near using base ISO. I use quite high ISOs. This means that the camera is collecting less light than it would at base ISO, so even though I'm using flash I'm essentially working in a low light situation. And in addition to the low light, I use very strong post processing on my images to try to mitigate the effect of severe diffraction softening, and also to control highlights and bring up dark areas. Strong post processing very likely has a lot of potential for accumulating calculation errors and degrading the images. I couldn't help thinking that with an A7sii I might be able to get better results with the peculiar approach that I'm using. 

There is of course the fact that the A7sii is "only"12 megapixels. However, I know from experience with my bridge cameras that 12 megapixels is plenty for what I do, with my outputs being only 1300 pixels high. I sometimes do substantial cropping, but it isn't extreme and 12 megapixels is enough for the cropping I want to do. And in any case I seem to be cropping rather less with my new, tiny aperture approach with its greater than usual depth of field.

I idly wondered how much an A7sii would cost, so I went over to eBay to have a look. They looked affordable, and there was one with a low-ish shutter count (under 5,000) and less than 3 hours video capture, which is not much for a camera which is generally considered as a video camera rather than a stills camera. For £1,000 it also had an almost unused Sigma MC-11 Canon EF to Sony E mount converter. These are good. I have one. Having another would let me use non-Sony lenses on both the A7ii and the A7sii if I wanted. They cost around £200 new. And there were three batteries, two of which were Sony branded (around £50 each new). The seller had owned it all from new.

I bought it. 

It arrived very quickly, two days ago, and from the way it was packed I would guess that the owner was the type who looks after his kit very carefully (unlike me). 

The sensor was a bit spotty, but not as bad as I expected it to be. One wet clean got rid of most of the spots and I decided to leave it at that  and go and test it in the garden rather than trying to clean it more and risk getting it worse again and getting into a cleaning loop rather than getting out and testing it.

The A7sii has exactly the same body shape and button arrangement as the A7ii. I set it up in the same way as the A7ii. I used the Yongnuo flash setup that I had sorted out for the A7ii, described in the previous post.

There was not much around in the garden so I had a relatively short session of 40 minutes or so, but there was enough to show that the camera appeared to be working properly. I did notice one difference from the A7ii; I seemed to be getting stronger focus peaking signals, which made focusing easier. But perhaps it was just my imagination, seeing what I wanted to see from my new purchase. Anyway, the post processing went fine and the results certainly seemed no worse than with the A7ii. Perhaps they were a bit better, but that could have been my imagination too.

The 34 images ( a rather high 20% keeper rate) I kept from the 174  shots I captured during the session are in this album at Flickr. Here are several of them, including one natural light botanical shot from an interlude when I couldn't find any animals to photograph.

My first impressions, of handling, processing and outputs were promising.











Tuesday, 11 May 2021

A random walk to a new flash setup, possibly

 My experiments tend to be a bit haphazard. One consequence of this is that it can take me much longer to realise something than it would have had I been more methodical. I know that, but it's just not my style to be buttoned down methodical, carefully pre-planned, thorough records etc. It's the same with my style of shooting. Persistent, yes. Very persistent. But methodical, no. I try to be methodical sometimes, but I get bored with it, and wander off, lose track of the plan, can't be bothered to keep records, forget what happened, don't see the patterns. Bumbling along. Rather inefficient.

There can be an upside to this though Sometimes I wander along from one thing to another, and suddenly realise that I've ended up somewhere rather unexpected, but rather promising. Somewhere that had I been methodical, logical, rational, I probably wouldn't have reached. 

So it was yesterday.

I intended to go over the road to the church grounds to do some aperture variation tests as mentioned at the end of the previous post.  However, it kept raining on and off, which put me off. So I decided to do some indoor testing. That way I could be more, yes, methodical, and should be able to home in quickly on good apertures for different magnifications. I did need a suitable subject though. None of the man-made things I tried had the appropriate type and level of detail. But I searched the window sills and found a dead wasp, about 10mm long. That would do nicely, it's eyes in particular.

Until it's head fell off.

But even headless it was useful. Until it fell apart even more. But I did do some comparisons. I couldn't do enough to be sure about it, but to my surprise the comparisons I did manage to do seemed to support the continued use of the constant f/45 that I had happened upon by .... by accident? Out of laziness? Through some sort of unconscious evaluation? Who knows. Anyway, I should look at this more thoroughly when I can find another suitable subject. Should. Will? Maybe. We'll see.

Anyway, With the very small apertures I'm using I'm finding it difficult to get enough light on to the subject, especially when testing even smaller apertures than the f/45 I've been using recently. One thing that can lead to over-optimistic results is putting the subject on a somewhat reflective surface that can bounce light on to the subject, especially its underside. To try to avoid this I'll sometimes put the subject right on the edge of the table, bookshelf or whatever. A related problem can arise when the subject's surrounding are much lighter than they might be out in the field. To try to avoid dropping into that trap I put subjects on to dark (and not very reflective) doormats. I have several of these spread around in my study to try to protect the light coloured carpet from the muck that I bring in from the garden on my shoes. 

The doormat I used for the dark surroundings tests was in a position where it wasn't getting much light. That made it impossible to see enough to focus properly. The obvious solution would have been to move the mat into the light. But (fortunately) that never entered my mind. Instead the obvious thing seemed to be to turn on the KX800's focusing light. I realised that the KX800 I was using (I have three) was the one with a faulty focusing light which only comes on for a moment every second or so. It is unusable. (The flash itself is fine. It is the one I have been using for a while now. Just not the focusing light.)

One of the other two KX800s has a faulty arm. The KX800s have loc-line (or similar) arms. One of the arms on this one has broken several times, in different places, and each time I have superglued the broken ball and socket joint together. It is now rather inflexible and I wasn't keen on trying to bend the arms into the shape needed for the very long lens setup I'm using on the A7ii. So, I used the third KX800.

The focusing light started acting up. Not as bad as the first one, probably still usable, but I didn't like the way it was going. I decided to try a different direction.

I have a Yongnuo YN24EX twin flash. Like the Canon twin flashes, it can be mounted on the front of the lens, which gets the flash heads near to the subject. That had been sitting on the shelf for many months because although the power of the flash was fine given its front mounting, despite numerous attempts I couldn't arrange diffusion for it which got me illumination I liked. This was despite help from John Kimbler (@dalantech), who uses this type of flash setup to very good effect. However, I could live with not so good illumination for this exercise; it was how detail came out at different apertures that I was concerned with, and the YN24EX has focusing LEDs built in to each flash head. 

The simple, quick and straightforward thing would have been to mount the YN24EX flash heads on the front of the lens as I had done previously. But I looked at the pathetic little two-polystyrene-layer diffusers that were still on the flash heads as the last thing I had tried before giving up on the YN24EX, and I decided on a whim to try something else.

I remembered that I had bought a flash bracket with bendy arms like this one. 



The arms on the one I bought are very stiff indeed. Also, the configuration of the base and the arms made it uncomfortable to use hand-held. It was not an appealing piece of kit. I had not found a use for it. For some reason I decided to try using it with the YN24EX. 

I reused the diffusers from the KX800, and after an hour or so of testing and adjustments, the setup ended up looking like this.



The piglets illustrate the working distances. From 8:1 down to 2:1 or so the working distances are similar enough that it is easy to arrange the flash heads so they can provide satisfactory illumination without needing to adjust the position of the heads. However, as you approach 1:1 the working distance gets bigger faster, getting to around 350mm by 1:1, at which point the working distance is increasing very fast indeed. This makes flash illumination difficult for low magnifications, partly because the flash heads are pointing in the wrong direction (or if they are pointing in the right direction they are pointing in the wrong direction for higher magnifications). And even if they are pointing in the right direction the illumination level decreases with distance according to the inverse square law, and 1:1 is a lot further away than 2:1. Given that I am using very small apertures, this makes it difficult to provide enough light at lower magnifications. This is a general problem with this double teleconverter setup, not specific to the YN24EX.

What surprised me was that it seemed, at least from indoor testing, that this setup could actually provide (just) enough illumination out at 1:1, and enough light to aid focusing too. In dark scene tests I had to let the ISO go up beyond the ISO 3200 that is equivalent to what I have been using with close-up lenses for years, but not by much, and only right out at 1:1. It only needed a quite small increase in magnification to get it back to ISO 3200.

There were also a couple of hints that the quality of illumination might be slightly better than with the KX800. This seems unlikely, given that I was re-using the KX800 diffusion arrangement. However, more by accident than design, the outer diffuser was mounted further away from the flash head boxes than with the KX800, and previous tests indicated that greater inter-layer distances increase the diffusion effect. 


Here is what it looks like from the subject's perspective.


Desktop tests are one thing. How things work in the field can be quite another, so earlier today I went across to the church grounds to see if I could find anything to give the setup a reality test. It did ok.

  • The flash bracket was not as uncomfortable to use as I thought it might be. I think that is because of the length of the lens setup and the fact that I keep my left hand at the front of the barrel, on the focus/magnification ring. With the left hand bendy arm running up close to the lens barrel, and my hand out at the front, the bendy arm didn't get in the way.
  • The flash provided enough illumination for me to run the flash at 1/4 power as usual to keep recycle times fast enough for my shooting rate without having to raise the ISO so far as to be troublesome.
  • The quality of the illumination was no worse than with the KX800, and might possibly have been just a little better with reflective surfaces. 
  • The way the outer diffuser was mounted, a bit higher than with the KX800, left just enough of a gap underneath which I could look through to help with lining up subjects.
  • I didn't need to use them, but it was nice to know I had focusing lights available should I need them.
  • If I really wanted to, I could alter the direction the flash heads were pointing because there are flexible connectors at the ends of the bendy arms. This is not possible with the KX800 because the flash heads are in a fixed position at the end of its bendy arms. In order to rotate the flash heads you have to manipulate the arms, which moves the flash head to a different location. It can be tricky to get the flash head back to the right place but pointing in a different direction.  
I captured 285 shots during the test session and ended up keeping 60 of them, a keeper rate of 21%. Given that it was distinctly breezy, and that the majority of the subjects were fairly small, that seemed very satisfactory. 

Here are some of the images from the session. The rest are in this album at Flickr.