p.19 #1 · A thinner sensor stack may be possible after all!
I'm puzzled why you would tweak the AWB setting to compensate for the mod. I have not seen any significant changes there. AWB is an imperfect science anyway, since the algorithm in the camera is trying to guess at the proper WB based on the image content. How good it is depends on the camera algorithm and the subject matter - and certainly some algorithms are better than others under some circumstances. I think you should just need to tweak the presets, which do make assumptions about the filter stack. You might be creating more work for yourself chasing AWB nudges around.
I always work from RAW files and adjust white balance for each photo anyway, unless the 'as shot' from auto looks fine, which is the exception rather than the norm for me with any camera. At least when I shoot color, which is a small percentage of my photos, being an old-timer.
p.19 #2 · A thinner sensor stack may be possible after all!
mdemeyer wrote:
I'm puzzled why you would tweak the AWB setting to compensate for the mod. I have not seen any significant changes there. I think you should just need to tweak the presets, which do make assumptions about the filter stack. You might be creating more work for yourself chasing AWB nudges around.
I always work from RAW files and adjust white balance for each photo anyway, unless the 'as shot' from auto looks fine, which is the exception rather than the norm for me with any camera. At least when I shoot color, which is a small percentage of my photos, being an old-timer.
p.19 #3 · A thinner sensor stack may be possible after all!
Hello Michael and Uhoh,
Do you think 21 color skopar or 15mm heliar would make good choices for an A7"m" now?
I'm looking for a "cheap" and light M mounted option to combine with nokton 35 and future 75.
The new heliar seems attractive but bigger and 2 to 3 times the price of used heliar.
So, is the 21 skopar still the "worst" WA on A7? (colour shift, angles)
p.19 #4 · A thinner sensor stack may be possible after all!
A revealing tell-tale would be testing a modded camera with 'hardboiled' wides such as the Biogon G 21/2.8 or the Biogon ZM 25/2.8, which are by themselves excellent, regarding corner sharpness. Most of the possible softness with these two would be due to the interaction with the sensor's filter.
p.19 #5 · A thinner sensor stack may be possible after all!
uhoh7 wrote:
You may be right. I'm experimenting. But I would like to be close OOC.
Anyway, I hope my current WB issues don't cloud the central issue: edges!! We have some nice edges now
Just for kicks, you might try a different raw converter with the same image. Often, I don't see the same issue with Sony's own IDC or even C1 so there is something going on with the Adobe profile with certain Sony's.
p.19 #6 · A thinner sensor stack may be possible after all!
I bought a 21mm Skopar for the project, since it was IMO essentially unusable owing to smearing on the original A7. I figured it as a worst-case example and it certainly did improve a lot with the mod. It's not a lens I kept since my go-to lens in that range is a 24mm Elmar 3.8.
Fall off and color shift are pretty extreme with that lens, which are optics or sensor issues, and create too much PP work for me to tolerate. That said, for around $300 (used), there are not many choices that would be better except maybe some SLR wides, but that blows the size advantage. A compromise, somewhat larger, would be the Olympus wides, which are still not too large.
p.19 #8 · A thinner sensor stack may be possible after all!
FWIW, the C-Biogon 21/4.5 performs quite good with the mod (sharp corner-to-corner by f/8-f/11) a tiny bit soft at WO (I think on M9 it has the same problem too).
p.19 #9 · A thinner sensor stack may be possible after all!
Beni wrote:
How do you feel about the Huelight? I'm loathe to put down the money without being able to see or test the differences.
I only just got it, but so far I like it. My world isn't terribly colorful at the moment, though. It does seem to be better for skin tones, for me. With the Adobe Standard profile, it seems I was often struggling with getting it to look right, especially indoors and in poor light. Too pink, red, green, yellow, or pale. I could usually fix it by playing with the sliders in Camera Calibration, but I never ended up with a standard set of tweaks. I'm hoping Huelight will get me close enough that I don't have to bother with them very often.
I could try to do a test shot if no one else has yet.
p.19 #10 · A thinner sensor stack may be possible after all!
vitix68 wrote:
Hello Michael and Uhoh,
Do you think 21 color skopar or 15mm heliar would make good choices for an A7"m" now?
I'm looking for a "cheap" and light M mounted option to combine with nokton 35 and future 75.
The new heliar seems attractive but bigger and 2 to 3 times the price of used heliar.
So, is the 21 skopar still the "worst" WA on A7? (colour shift, angles)
Thanks for all previous feedbacks
Forget 24/4 CV and all skopars, my personal advice. The new CV15 which is coming out should be good, and the older ones might be OK aside from colorshift, I don't have one. They should be better than the tiny 21/4.
I'm surprised and skeptical the ZM21/4.5 is OK with mod. But perhaps others know better.
At 21 the Leica SEM is good, but pricey at 2100. Colroshift is very mild. I'd bet the ZM21/2.8 is pretty good as well, but cannot verify. The ZM18 is good but with colorshift.
Tariq Gibran wrote:
Just for kicks, you might try a different raw converter with the same image. Often, I don't see the same issue with Sony's own IDC or even C1 so there is something going on with the Adobe profile with certain Sony's.
I just bring the raws straight into LR. All specific advice about an alternative very welcome
p.19 #11 · A thinner sensor stack may be possible after all!
Lloyd Chambers tested the ZM 21/4.5 as his "worst case scenario" and reported good sharpness stopped down but, of course, still with huge color shifts. It's a subscriber-only report if you want to see the sample images. Out of respect for his subscription model (which I have subscribed to for some time and highly recommend) I'm not quoting him further on the review - but if you are either considering the lens or the mod and looking for definitive info, the Leica subscription is worth the $50 IMO.
p.19 #14 · A thinner sensor stack may be possible after all!
Looks good to me Charlie. There might be a slight hint of green/cyan in the shadows, but could simply be the quality of the light.
From my tests of the CV15 on the various a7 models, it and the CV12 were generally pretty decent already in respect to edge sharpness. It's color shift that is the bigger problem and likely will remain an additional step to fix in post with modded cameras.
Matt - the ZM25 is such as sharp lens (as you know) that even if the post-mod edges aren't exactly as good as a Leica M, it will be miles ahead of what it was on a stock sensor. I can't really think of a ~24mm SLR lens I'd rather have over the ZM, other than something specialized like the Canon TS-E.
Have we seen anything yet that was a disappointment, in respect to edge smearing (not color shift), on the modded sensors? If so, I would almost bet it would be similar on a Leica. A lens like the CV21/4 might be an example - it's never going to be a technically spectacular lens like the 21 SEM or ZM21/4.5.
p.19 #15 · A thinner sensor stack may be possible after all!
It's a very good one at that FL, Ron, smooth slow decline MTF and best of all, excellent corners, and very low distortion for this FL. And very good wide open, a plus over the SEM pair in speed/versatility, also very light.
Now is a good time to check MTF for each lens of interest, see what each should be doing, easier to assess sensor effects.
CZ could grow the basic 25/2.8 Biogon design and enlarge/tune it, if it becomes one of the wide Loxias. It has more mid aperture fall-off in ZM form than the equivalent CZ DSLR lenses, a design issue. They would try to make it match the L35/50 pair in barrel dimensions - the Loxias are a great fit to the a7 cameras for MF lenses. I could see one going well if that is what Zeiss intend. Say a 25/2.8 at 350 grams with a common filter size to the L35/2. ZM is now (an under-priced) $1200, so a Loxia might be the same ballpark.
p.19 #16 · A thinner sensor stack may be possible after all!
At the risk of sounding a bit like a scratched record, could anyone check for field curvature? Has there been any changes? How about sensor reflections? Thanks so much!
p.19 #17 · A thinner sensor stack may be possible after all!
xbarcelo wrote:
At the risk of sounding a bit like a scratched record, could anyone check for field curvature? Has there been any changes? How about sensor reflections? Thanks so much!
It's the change in field curvature that results in smeared corners on an unmooded A7. So I believe there is a change toward the original lens' field curvature.
p.19 #18 · A thinner sensor stack may be possible after all!
xbarcelo wrote:
At the risk of sounding a bit like a scratched record, could anyone check for field curvature? Has there been any changes? How about sensor reflections? Thanks so much!
Regarding reflections, as I posted previously, I have shot some dusk images including streetlights. Not a before and after test, but have not had issues that troubled me. I can, if I shoot in the dark at night, cause reflections from streetlights, but I can make almost any digital camera do that with the right (or wrong) combination of lights and exposure. Not having an unmodified camera handy I can't do comparative shots, unfortunately, but will upload a couple to the Flickr group when I'm back home next week.
Regarding curvature and corners, the primary reason for the improvement from thinning the stack is not changes to the lens field curvature performance. Curvature doesn't smear - it just causes the image to be out of focus. The smearing results from a asymmetrical increase in astigmatism as you move off-center in the image field.
Again, I don't have a way to test before/after, but ALL film lenses should perform closer to their designed film performance regarding curvature. If adding glass to the stack improved field curvature vs. the performance on film, that would seem to be a very lucky accident for some particular lens. But it would be at the expense of smeared edges which, being an asymetrical distortion, tends to be much more disturbing that simply not being sharp. Yes, as the ray angle increases toward the edge, the filter glass increases the path length, and that would create some focal plane curvature. But a thinner stack would always has less effect than a thicker one. So performance on a film lens is closer to the designed performance. That's the fundamental goal of the mod. Getting closer to the as-designed performance of a film lens. Not to try to create a lucky combination of the stack with some particular lens. And I think all the results show that it does that.
I would be interested in any results or theory that suggest otherwise or show how adding glass plate to the path would generally improve curvature.
p.19 #19 · A thinner sensor stack may be possible after all!
mdemeyer wrote:
Regarding curvature and corners, the primary reason for the improvement is not changes to the lens field curvature performance. Curvature doesn't smear - it just causes the image to be out of focus. The smearing results from a asymmetrical increase in astigmatism as you move off-center in the image field.
Michael
Well Michael, it does change the field curvature. Since you have Lloyd's subscription, just check again the Zeiss 35/1.4 MTF on a mirrorless. However, you are right that the change in curvature does not result in smearing (my bad ), but mainly because of the increase in astigmatism. However, both of these effects would produce not only a smeary corners, but also very soft corners as well. Astigmatism with high MTF would just give a mild double image.
p.19 #20 · A thinner sensor stack may be possible after all!
Agree it changes it. My point is just that thinner changes it less than thicker and the only thing that can be generalized is that, with a thinner stack, performance is closer to the film performance in this regard. Whether that's good or bad for any particular lens takes the level of testing that Lloyd does (very well IMO) that few, if any, others do.
), but mainly because of the increase in astigmatism. However, both of these effects would produce not only a smeary corners, but also very soft corners as well. Astigmatism with high MTF would just give a mild double image.