rscheffler wrote:
Sorry for the slow reply. You're right, the image I was thinking about is still there... in the Sigma gallery. When going back to your links I only found the Sony gallery, hence why I couldn't find it.
---------------------------------------------
I agree there is likely some additional nervousness due to the thicker Sony/Sigma sensor cover glass, but the image I'm thinking about was probably focused near to MFD where sensor thickness will negatively affect image quality less than at infinity.
See upper left electrical cables showing double line effect and weird double image effect in upper right OOF signage.
I agree that in that shot the background is fairly busy/nervous, but I don't really see the double lining, maybe just a little but only with a lot of magnification. Definitely something to look into more. Can that busyness and nervousness be abated if you stop down a little? Does the Sigma fp look different from a Leica camera? Basically for my taste I don't like the background in this shot, but I not convinced by this shot that the rendering difficulties seen here will be a major problem.
Steve Spencer wrote:
I agree that in that shot the background is fairly busy/nervous, but I don't really see the double lining, maybe just a little but only with a lot of magnification. Definitely something to look into more. Can that busyness and nervousness be abated if you stop down a little? Does the Sigma fp look different from a Leica camera? Basically for my taste I don't like the background in this shot, but I not convinced by this shot that the rendering difficulties seen here will be a major problem.
Unfortunately that image is showing 'induced' field curvature where rendering starts to sharpen up off-axis. This lens is very sensitive to sensor stack thickness and I would only recommend it on the Leica or modded Sony. Unless using it stopped down at f/5.6 or smaller on a thicker sensor.
Keeping up with the bike theme, here is one from sunny Hawaii:
Fred Miranda wrote:
Unfortunately that image is showing 'induced' field curvature where rendering starts to sharpen up off-axis. This lens is very sensitive to sensor stack thickness and I would only recommend it on the Leica or modded Sony. Unless using it stopped down at f/5.6 or smaller on a thicker sensor.
Keeping up with the bike theme, here is one from sunny Hawaii:
(M9 SOOC Jpeg)
Feels like I can almost jump up on one of those bikes!
Steve Spencer wrote:
I agree that in that shot the background is fairly busy/nervous, but I don't really see the double lining, maybe just a little but only with a lot of magnification. Definitely something to look into more. Can that busyness and nervousness be abated if you stop down a little? Does the Sigma fp look different from a Leica camera? Basically for my taste I don't like the background in this shot, but I not convinced by this shot that the rendering difficulties seen here will be a major problem.
I will try to do a background rendering test with different apertures on my Sigma fp later on with a similar scene. So far I've only shot the lens wide open and at f5.6. We are in the middle of a rainy season in Tokyo and weather hasn't been very nice for photography in recent weeks most of the time so I've not been able to shoot as much as I'd like. The weather forecast for next 10 days is still rainy on most days.
I also don't think this background rendering on wide open shots will be a major problem for me as I will generally use 28mm more for deep dof infinity shots, and for close-ups of flowers & objects where the background is much closer and in those scenarios the background rendering looks better and not distracting. I think the bicycle shot showed one of the worst case rendering scenarios based on subject distance and background distance and was potentially affected by unoptimized camera/lens combination to some extent.
I can't directly compare results between Sigma fp and Leica since I don't have any Leica M bodies. Wouldn't mind having a digital full frame body with Leica M sensor characteristics so that my CV RF glass optimized for digital Leica M would work optimally on it but the camera would have to come with an EVF rather than rangefinder and with reasonable price tag (up to $3000) and a smallish / light body to be appealing to me. I don't suppose Leica will ever make a camera that matches those characteristics. Perhaps an A7C with Kolari Ultra Thin sensor cover modification would do the trick but as I understand it Kolari UT sensor cover glass is about 0.2 mm and Leica digital Ms are somewhere around 0.8-9mm (and Sigma fp is 1.3mm). Therefore I'm not sure UT modification would be optimal solution for getting best results with lenses that are optimized for digital M cameras. Seems like UT would be better match for old lenses designed for film cameras. Currently I'm sticking to my existing options with standard Sony A7C and Sigma fp with their native lenses + my adapted RF glass even though there are some compromises. If CV put out a nice native 28mm for E-mount the problem would be pretty much solved already as my other favorite focal lengths are already pretty well covered with native lenses. For 28mm (one of my very favorite focal lengths) I primarily rely on adapted RF glass (I have this CV 28/2 II, 2 x CV 28/3.5 LTM (Black and Silver), 2 x Avenon 28/3.5 (Black and Silver), Minolta M-Rokkor 28/2.8).
Jul 01, 2021 at 10:23 PM
Steve Spencer Offline • • • • • • • Upload & Sell: On
Juha Kannisto wrote:
I will try to do a background rendering test with different apertures on my Sigma fp later on with a similar scene. So far I've only shot the lens wide open and at f5.6. We are in the middle of a rainy season in Tokyo and weather hasn't been very nice for photography in recent weeks most of the time so I've not been able to shoot as much as I'd like. The weather forecast for next 10 days is still rainy on most days.
I also don't think this background rendering on wide open shots will be a major problem for me as I will generally use 28mm more for deep dof infinity shots, and for close-ups of flowers & objects where the background is much closer and in those scenarios the background rendering looks better and not distracting. I think the bicycle shot showed one of the worst case rendering scenarios based on subject distance and background distance and was potentially affected by unoptimized camera/lens combination to some extent.
I can't directly compare results between Sigma fp and Leica since I don't have any Leica M bodies. Wouldn't mind having a digital full frame body with Leica M sensor characteristics so that my CV RF glass optimized for digital Leica M would work optimally on it but the camera would have to come with an EVF rather than rangefinder and with reasonable price tag (up to $3000) and a smallish / light body to be appealing to me. I don't suppose Leica will ever make a camera that matches those characteristics. Perhaps an A7C with Kolari Ultra Thin sensor cover modification would do the trick but as I understand it Kolari UT sensor cover glass is about 0.2 mm and Leica digital Ms are somewhere around 0.8-9mm (and Sigma fp is 1.3mm). Therefore I'm not sure UT modification would be optimal solution for getting best results with lenses that are optimized for digital M cameras. Seems like UT would be better match for old lenses designed for film cameras. Currently I'm sticking to my existing options with standard Sony A7C and Sigma fp with their native lenses + my adapted RF glass even though there are some compromises. If CV put out a nice native 28mm for E-mount the problem would be pretty much solved already as my other favorite focal lengths are already pretty well covered with native lenses. For 28mm (one of my very favorite focal lengths) I primarily rely on adapted RF glass (I have this CV 28/2 II, 2 x CV 28/3.5 LTM (Black and Silver), 2 x Avenon 28/3.5 (Black and Silver), Minolta M-Rokkor 28/2.8)....Show more →
Thanks for all the samples you have provided and for your take on the particular shot with the bike. I think what you say makes a lot of sense.
As for cover glass thickness, as I understand it the Kolari ultra thin (UT) leaves you with a camera with two pieces of glass between the mount and the sensor. One is a bit off the sensor and is .2mm thick, but there is another piece of glass epoxied onto the sensor that Kolari doesn't touch. That piece of glass is thicker, I believe something like .8mm making the two pieces of glass about 1mm thick and very close to the thickness of the glass on the CMOS Leica M cameras. So, I do think the Kolari mod is quite close in cover glass thickness to the Leica M240 and M10.
I know much less about the Sigma fp (and I don't even know if I am right about the Kolari UT conversion), but I am not sure what Kolari means when they say (if it was them) that the Sigma fp has 1.3mm in cover glass. Is that a piece of glass a bit in front of the sensor? Is there another piece of glass epoxied on the sensor like in the Sony cameras? What I do know is that Kolari said the thickness of the cover glass on the Nikon Z camera is 1.1mm, but this did not include the piece of glass epoxied on the sensor for these cameras. When you add the two pieces of glass together for the Nikon camera you have something around 2mm of glass and these Nikon Z cameras perform very similarly to Sony cameras with Leica M glass (and I have tested that out with 5 or 6 different lenses) and with Voigtlander lenses in which there are both Sony E mount and Leica M mount (both of which can be used on Nikon Z cameras) I have always preferred the Sony E mount lenses on the Nikon Z cameras. Given that history and my lack of knowledge about what the cover glass is like on the Sigma fp I worry that the cover glass is actually quite a bit thicker than it is on Leica M. We also know that the Panasonic S1, S1R, and S5 cameras seem to have thicker cover glass as Leica M lenses don't do much better on those cameras than they do on Sony, so it does not seem to be that the L mount alliance has specified a thinner cover glass. Even the Leica SL camera don't seem to do quite as well with Leica M glass as Leica M cameras, although Leica SL cameras clearly do better than Sony.
All that is to say the performance of Leica M glass on the Sigma fp is an interesting question. I don't really know anything but I have my doubts that Leica M lenses perform as well on that camera as on a Leica M body, and I would be surprised if they even perform as well as they do on a Leica SL body--but they might I don't really know. I have found this cover glass stuff extremely hard to sort out as it is really hard to get factual information and when you find things that looks like factual information it is then challenged by others with conflicting information that also looks factual.
Steve Spencer wrote:
Thanks for all the samples you have provided and for your take on the particular shot with the bike. I think what you say makes a lot of sense.
As for cover glass thickness, as I understand it the Kolari ultra thin (UT) leaves you with a camera with two pieces of glass between the mount and the sensor. One is a bit off the sensor and is .2mm thick, but there is another piece of glass epoxied onto the sensor that Kolari doesn't touch. That piece of glass is thicker, I believe something like .8mm making the two pieces of glass about 1mm thick and very close to the thickness of the glass on the CMOS Leica M cameras. So, I do think the Kolari mod is quite close in cover glass thickness to the Leica M240 and M10.
I know much less about the Sigma fp (and I don't even know if I am right about the Kolari UT conversion), but I am not sure what Kolari means when they say (if it was them) that the Sigma fp has 1.3mm in cover glass. Is that a piece of glass a bit in front of the sensor? Is there another piece of glass epoxied on the sensor like in the Sony cameras? What I do know is that Kolari said the thickness of the cover glass on the Nikon Z camera is 1.1mm, but this did not include the piece of glass epoxied on the sensor for these cameras. When you add the two pieces of glass together for the Nikon camera you have something around 2mm of glass and these Nikon Z cameras perform very similarly to Sony cameras with Leica M glass (and I have tested that out with 5 or 6 different lenses) and with Voigtlander lenses in which there are both Sony E mount and Leica M mount (both of which can be used on Nikon Z cameras) I have always preferred the Sony E mount lenses on the Nikon Z cameras. Given that history and my lack of knowledge about what the cover glass is like on the Sigma fp I worry that the cover glass is actually quite a bit thicker than it is on Leica M. We also know that the Panasonic S1, S1R, and S5 cameras seem to have thicker cover glass as Leica M lenses don't do much better on those cameras than they do on Sony, so it does not seem to be that the L mount alliance has specified a thinner cover glass. Even the Leica SL camera don't seem to do quite as well with Leica M glass as Leica M cameras, although Leica SL cameras clearly do better than Sony.
All that is to say the performance of Leica M glass on the Sigma fp is an interesting question. I don't really know anything but I have my doubts that Leica M lenses perform as well on that camera as on a Leica M body, and I would be surprised if they even perform as well as they do on a Leica SL body--but they might I don't really know. I have found this cover glass stuff extremely hard to sort out as it is really hard to get factual information and when you find things that looks like factual information it is then challenged by others with conflicting information that also looks factual. ...Show more →
As I recall from either Kolari Vision or Vivek (realvivek), the total thickness of the Kolari Vision UT sensor modification at least for the Sony cameras works out to be about 1.1mm thick for the complete sensor stack.
Steve Spencer wrote:
Thanks for all the samples you have provided and for your take on the particular shot with the bike. I think what you say makes a lot of sense.
As for cover glass thickness, as I understand it the Kolari ultra thin (UT) leaves you with a camera with two pieces of glass between the mount and the sensor. One is a bit off the sensor and is .2mm thick, but there is another piece of glass epoxied onto the sensor that Kolari doesn't touch. That piece of glass is thicker, I believe something like .8mm making the two pieces of glass about 1mm thick and very close to the thickness of the glass on the CMOS Leica M cameras. So, I do think the Kolari mod is quite close in cover glass thickness to the Leica M240 and M10.
I know much less about the Sigma fp (and I don't even know if I am right about the Kolari UT conversion), but I am not sure what Kolari means when they say (if it was them) that the Sigma fp has 1.3mm in cover glass. Is that a piece of glass a bit in front of the sensor? Is there another piece of glass epoxied on the sensor like in the Sony cameras? What I do know is that Kolari said the thickness of the cover glass on the Nikon Z camera is 1.1mm, but this did not include the piece of glass epoxied on the sensor for these cameras. When you add the two pieces of glass together for the Nikon camera you have something around 2mm of glass and these Nikon Z cameras perform very similarly to Sony cameras with Leica M glass (and I have tested that out with 5 or 6 different lenses) and with Voigtlander lenses in which there are both Sony E mount and Leica M mount (both of which can be used on Nikon Z cameras) I have always preferred the Sony E mount lenses on the Nikon Z cameras. Given that history and my lack of knowledge about what the cover glass is like on the Sigma fp I worry that the cover glass is actually quite a bit thicker than it is on Leica M. We also know that the Panasonic S1, S1R, and S5 cameras seem to have thicker cover glass as Leica M lenses don't do much better on those cameras than they do on Sony, so it does not seem to be that the L mount alliance has specified a thinner cover glass. Even the Leica SL camera don't seem to do quite as well with Leica M glass as Leica M cameras, although Leica SL cameras clearly do better than Sony.
All that is to say the performance of Leica M glass on the Sigma fp is an interesting question. I don't really know anything but I have my doubts that Leica M lenses perform as well on that camera as on a Leica M body, and I would be surprised if they even perform as well as they do on a Leica SL body--but they might I don't really know. I have found this cover glass stuff extremely hard to sort out as it is really hard to get factual information and when you find things that looks like factual information it is then challenged by others with conflicting information that also looks factual. ...Show more →
Yeah, after reading a bit more on this topic it does seem that Kolari UT modified A7 is quite close to M9 in terms of performance with several RF lenses that were tested in this review and especially modern rangefinder wide angle lenses that were designed for digital worked similarly well with the UT modification: https://phillipreeve.net/blog/review-kolari-ultra-thin-sensor-stack-modification/
As for the Sigma fp cover glass thickness being 1.3mm, this information was provided to @imagesfromobjects@ from Kolari after he had sent his fp for UT conversion and it was the first time for Kolari to work on a conversion for fp. https://www.fredmiranda.com/forum/topic/1633515/6#15332394
> By the way, the sensor stack is approximately ~1.3mm by their measurements.
I don't know if this 1.3mm refers to the whole sensor stack or whether there is something extra in addition there. In my personal experience fp has better edge and corner performance with RF lenses than my Sony E-mount cameras in terms of sharpness / smearing but on the other hand fp has more color shading and vignetting than e.g. A7C and A9 with the same lenses. Color shading and vignetting are easier to mitigate though and fp has functionality that allows creation of 10 color shading compensation profiles in-camera based on shooting reference images (I use this feature by shooting the sky through an Expodisc).
It would be interesting to see a comprehensive comparison review on how all these cameras other than digital M (e.g. various L-mount cameras, E-mount cameras, Nikon Z and Canon RF) perform with certain rangefinder glass. I'm sure they wouldn't do as well as digital M but certainly there would be some noticeable differences.
I have a Nikon Z6 with Kolari UT mod, and I got to test it side by side with a Leica M 240 using my 28mm Summicron-M Asph II. Regarding field curvature (and possible smearing), they behave indistinguishably.
Jul 02, 2021 at 03:25 AM
Steve Spencer Offline • • • • • • • Upload & Sell: On
LarsHP wrote:
I have a Nikon Z6 with Kolari UT mod, and I got to test it side by side with a Leica M 240 using my 28mm Summicron-M Asph II. Regarding field curvature (and possible smearing), they behave indistinguishably.
That makes good sense to me. I would think that both Sony E mount and Nikon Z mount cameras need the modification to get the best performance regarding field curvature/astigmatism.
I just purchased a Leica SL to test how the Voigtlander 28/2 Ultron II performs on its sensor. I will update the review once the M to L adapter shows up.
Both the above are near focus (the grass), wide open; 1000/s ISO 64. Ooc jpg resized. Top is a 28mm Ultron f/1.9 LTM and bottom is the 28mm Ultron f/2 II (I have version I of the f/2 also). The difference in exposure is almost a third of a stop less underexposure for the LTM. For me this would ideally be an f/8 image, which the f/1.9 LTM handles with aplomb.
Edit:
28mm Ultron LTM at f/8. Note the upper corners...
That comparison indicates the Ultron II has quite a bit more vignette than the f/1.9 LTM. Not unsurprising considering the small front element of the new lens.
Both the above are near focus (the grass), wide open; 1000/s ISO 64. Ooc jpg resized. Top is a 28mm Ultron f/1.9 LTM and bottom is the 28mm Ultron f/2 II (I have version I also). For me this would ideally be an f/8 image, which the f/1.9 LTM handles with aplomb.
What are you using for a camera for these?
Thanks.
Rich
Jul 02, 2021 at 01:46 PM
Steve Spencer Offline • • • • • • • Upload & Sell: On
naturephoto1 wrote:
What are you using for a camera for these?
Thanks.
Rich
The ISO 64 is a give away to the camera. It probably is a Nikon Z7 or Z7 II. I don't know if the sensor has been modified, but having an unmodified Z7 I do think the sensor tends to exacerbate vignetting and that might be part of the reason it is so strong here.
Steve Spencer wrote:
The ISO 64 is a give away to the camera. It probably is a Nikon Z7 or Z7 II. I don't know if the sensor has been modified, but having an unmodified Z7 I do think the sensor tends to exacerbate vignetting and that might be part of the reason it is so strong here.
Hi Steve,
Thank you. I will have to check the lens on my Kolari Vision UT sensor modified A7rII wide open and stopped down somewhat. It may take me a number of days to get this however. I agree and suspect that a stock Nikon or Sony FF camera will exacerbate the effects of vignetting of the lens as well.
Compared to a digital Leica M camera, I expect most, if not all, wide angle lenses will have more vignetting with FF mirrorless cameras even after Kolari UT mod. I base this on my side-by-side tests with my 28mm Summicron Asph II on my Z6UT versus M 240 as well as theory. Without having measured it, I think the Z6UT has about one stop less in the corners compared to the M 240 (shooting wide open with the 28mm Summicron Asph II).
The reason is that the micro lenses in the sensor of Leica M cameras are designed to handle steep angles of light much better than other cameras. From the graphical drawing of the M 240 sensor design (see link below), the micro lenses of regular sensors are relatively flat whereas those in the M 240 are taller, which makes them catch angled light better.
I think all sensors have less effective pixels than there actually is. The Leica M sensors are not the same as those we find in Nikon, Sony and Canon, so we can't assume that M sensors are more cropped by pixel count. Leica specs the effective pixel area as 23.9 × 35.8 mm which is comparable to other full frame sensors.
mapgraphs wrote:
I wondered about the well design. The M-240 also crops 24 pixels on each side and 12 on the top and bottom...
Edit:
My bad. To explain more concisely, if one takes the same lens, for same shot, with both bodies, the M-240 shot is cropped (sides, top and bottom) compared to the 6000x4000 image...
mapgraphs wrote:
M-240: 5952 x 3968 Pixels (23.62 MPixels) (3:2)
A7ii: 6000 x 4000 Pixels (24.00 MPixels) (3:2)
Edit:
My bad. To explain more concisely, if one takes the same lens, for same shot, with both bodies, the M-240 shot is cropped (sides, top and bottom) compared to the 6000x4000 image...
Well, yes, because the M 240 has slightly fewer pixels. Similarly, the Nikon Z6 and Sony a7III has slightly more than a 24MP Sony a7II.
We usually don't call that "crop" though. That relates to sensor size, not pixel count.
I just checked the exact specs of the two mentioned cameras:
Nikon Z6 = 35.9 mm x 23.9 mm
Sony a7III = 35.6 x 23.8 mm
None of them are precisely 3:2. The Nikon is slightly longer, the Sony slightly shorter (both of which is annoying sometimes).
