I think they can fix it (magenta cast) by fixing the lens. I am not sure about the Leica images. Leica are not that reliable when it comes to colors. Just look at their digital history. Anyway, I am more concerned about the Sony version.
It might be a little worse than the 50, it's hard to say. The Zeiss UV would probably take a good bit of that away, but probably not all of it. I've been playing with the Radial Gradient tool (inverted) in Lightroom, and I can make a good bit of it go away on the CV 50 APO and Tamron 35. I have to muck around with the color afterwards though, a little more than I'd like.
If the cast was consistent across the frame, it wouldn't be a big deal. The fact that intensity changes towards the edges and corners, and is impacted (I believe) by the type of light, it makes it more painful to deal with.
And part of this makes me shake my head, did anyone take the lens out and shoot it on a sunny blue sky day? I mean it's hard not to notice.
-Tim
Fred Miranda wrote:
thanks Juha,
I'm surprised by the strong magenta color cast from the 35/2 APO samples on the A7R IV. Perhaps even worse than the 50/2 APO? The color cast is not present on the Leica M10-R samples...
Fred Miranda wrote:
thanks Juha,
I'm surprised by the strong magenta color cast from the 35/2 APO samples on the A7R IV. Perhaps even worse than the 50/2 APO? The color cast is not present on the Leica M10-R samples...
Yeah, I can see the cast in those sample images too, but I've not noticed color casts with my 50/2 APO on my A7C or A9.
I've had blueish color shading issues with my Sigma fp and rangefinder lenses, and I created C1 Pro Lens Cast Calibration profiles with different shooting apertures and focusing distances for each lens to tackle those (~100 profiles for each lens). I shot reference images towards sun or bright sky through an ExpoDisc for the LCC profile creation and those showed a clear shift on the RGB values towards blue when moving to towards the edges and corners. Those profiles work pretty well but they are a bit of extra work to use in PP, always have to choose and apply a profile that matches the aperture and shooting distance for best results.
When I took similar reference images with my 50/2 APO on my A7C through an ExpoDisc there is no shift in RGB values anywhere and I've not noticed any magenta shifts in my normal photos by eye up to now...
Anyway, need to see more samples with the 35/2 APO and I hope the final production version doesn't have such issues on Sony cameras...
My 50/2 APO does on my a7R III, and it seems to be the worst when in the presence of strong near UV spectrum (blue sky.) Your ExpoDisc may mask that.
Those samples look similar, albeit worse, than I'd expect when shooting my 50/2 APO. It mimics the pattern I see where the corners show the strongest shift.
I guess we'll see. Hopefully there is some other explanation for what is seen in those sample shots.
I wonder if it's only noticeable with A7R series cameras then... I still have the very original A7R in the closet and it was always more prone to color shifts than the later (especially BSI) models. I'll try my ExpoDisc thing with that tomorrow using my 50/2 and will try some blue sky shots when we get very clear skies next time.
Juha Kannisto wrote:
I wonder if it's only noticeable with A7R series cameras then... I still have the very original A7R in the closet and it was always more prone to color shifts than the later (especially BSI) models. I'll try my ExpoDisc thing with that tomorrow using my 50/2 and will try some blue sky shots when we get very clear skies next time.
I find this very interesting so I tested the lens (CV 50/2 APO) with both A7R/III and the original A7R. On A7R/III, there is very slight amount of color shift at the corners of the image (for morning blue sky). But it's very small amount, barely noticeable. But on the original A7R, the color shift becomes very clear. I did have a Heilopan UV filter mounted on the lens.
For curiosities, I also tested the Zeiss 55/1.8 on the A7R, the Zeiss lens also exhibited some color shift but not very noticeable.
If you have the desire, and time, take a look at these. Do you see any noticable difference between A and B, and A1 and B1? Talking about difference in magenta between center and corner.
tsdevine wrote:
If you have the desire, and time, take a look at these. Do you see any noticable difference between A and B, and A1 and B1? Talking about difference in magenta between center and corner.
I can just barely see, or think I see, a difference. To my eyes, it is the smallest difference I could possibly distinguish. It is right up there next to "no difference." This will have zero impact on my use of the lens.
No, that's absolutely reasonable if you see nothing that bothers you. I do find it annoying in some of my shots, as my eye is drawn to that color that looks out of place in the shot.
I have another lens that shows the behavior more. I'm hoping that the 35/2 APO is not comparable or better.
Bertrick wrote:
I can just barely see, or think I see, a difference. To my eyes, it is the smallest difference I could possibly distinguish. It is right up there next to "no difference." This will have zero impact on my use of the lens.
tsdevine wrote:
No, that's absolutely reasonable if you see nothing that bothers you. I do find it annoying in some of my shots, as my eye is drawn to that color that looks out of place in the shot.
I have another lens that shows the behavior more. I'm hoping that the 35/2 APO is not comparable or better.
These are very similar to what I saw in my images this morning. I am little bothered by the corner/edge color shift. I also hope that CV 35/2 APO will do better than the CV 50/2 APO in that regard.
'Late April' here, Fred - but no solid ETA as yet. Australia often does well with early lens releases.
The colour issue, we don't know how these images were prepared. I have no complaints re the 50/2, in fact quite the contrary. It could be Cosina's way of indicating the quite real differences in colour palettes across systems, perhaps with defaults. You have to like the colours from a lens, obviously. Time will tell.
I'm not a big fan of the modern Leica colours, you can see them in Jonathan Slack's pre-release review of the new APO-Summicron-M 35/2:
tsdevine wrote:
If you have the desire, and time, take a look at these. Do you see any noticable difference between A and B, and A1 and B1? Talking about difference in magenta between center and corner.
A has quite noticeable color shift towards the edges and corners in both .arw and .jpg. It's getting milder on B in both .arw and especially in .jpg. When looking at the RBG levels, blue channel is getting higher values than Red and Green increasingly towards the edges and corners, but in b.jpg the gap is only about half of the gap in a.jpg...
I also think B1 is looking better than A1. A1 has some noticeable shift but I can' really see it in B1.
Both A/B and A1/B1 were shot on my a7R III a few days apart. A/B is a picture of snow on the ground on a very drab and cloudy day, while A1/B1 is blue sky (probably obvious) on a very clear and sunny day.
A/A1 has no filter attached. B/B1 has a Zeiss UV filter, which has a hard cut around 410nm (which by my understanding is within the visible spectrum, I guess you could call it near UV spectrum.)
If I can get the 35/2 APO to perform like B and B1 without a filter, I'd be ecstatic. Although if I could get similar performance with the filter, I'd still be happy. If it's noticeably worse, even with the filter, I'll be bummed.
-Tim
Juha Kannisto wrote:
A has quite noticeable color shift towards the edges and corners in both .arw and .jpg. It's getting milder on B in both .arw and especially in .jpg. When looking at the RBG levels, blue channel is getting higher values than Red and Green increasingly towards the edges and corners, but in b.jpg the gap is only about half of the gap in a.jpg...
I also think B1 is looking better than A1. A1 has some noticeable shift but I can' really see it in B1.
I took some ExpoDisc shots with my CV 50/2 APO today aimed at sky (it was cloudy today and the light seemed a bit blueish) with 4 cameras: A7C, A9, A7R, A7II. They can be found here:
All were focused at infinity and taken at f2, f2.8, f4, f5.6 and f8. However, with A7II I seem to have taken 2 shots at f4 and missed the one at f5.6 somehow. It seems to me that A7C and A9 were more consistent and there were more color fluctuations with A7R overall and A7II also had some.
tsdevine wrote:
Both A/B and A1/B1 were shot on my a7R III a few days apart. A/B is a picture of snow on the ground on a very drab and cloudy day, while A1/B1 is blue sky (probably obvious) on a very clear and sunny day.
A/A1 has no filter attached. B/B1 has a Zeiss UV filter, which has a hard cut around 410nm (which by my understanding is within the visible spectrum, I guess you could call it near UV spectrum.)
If I can get the 35/2 APO to perform like B and B1 without a filter, I'd be ecstatic. Although if I could get similar performance with the filter, I'd still be happy. If it's noticeably worse, even with the filter, I'll be bummed.
Those all look good to me, except the a7R which I see a maybe a little. Thanks for the test Juha. I used to have an ExpoDisc, but it's probably from like the mid-2000's. I may have to dig around and see if I can find.
Not sure why I'm seeing what I see.
Juha Kannisto wrote:
I took some ExpoDisc shots with my CV 50/2 APO today aimed at sky (it was cloudy today and the light seemed a bit blueish) with 4 cameras: A7C, A9, A7R, A7II. They can be found here:
All were focused at infinity and taken at f2, f2.8, f4, f5.6 and f8. However, with A7II I seem to have taken 2 shots at f4 and missed the one at f5.6 somehow. It seems to me that A7C and A9 were more consistent and there were more color fluctuations with A7R overall and A7II also had some.
Thanks! My ExpoDisc is also a relic from mid-2000's, it has (c) 2005 on the box and I had bought it for some use with my old Canon DSLRs around 2005 or 2006. It's been useful again in the recent years...
tsdevine wrote:
Those all look good to me, except the a7R which I see a maybe a little. Thanks for the test Juha. I used to have an ExpoDisc, but it's probably from like the mid-2000's. I may have to dig around and see if I can find.
...will update.