I've noticed a strange phenomena on at least 2 of my lenses with my a7R III. My Voigtlander 50mm f/2 APO Lanthar and Tamron 35 f/2.8 Di III seem to have some magenta shading towards the corners. I have all in camera corrections turned off, and I don't believe it's the "magenta ring" phenomena that appears to be related to some shading corrections that can't completely be turned off.
I had been using Cornerfix with profiles I created to try to remove it, which for the most part worked ok, but it's painful from a workflow perspective.
So I was researching UV filters as a way to reduce PF in some lenses, a completely unrelated topic and in no way was I thinking of a UV filter as a solution to the shading I was seeing. For PF it was suggested that a Tiffen UV 2A filter can help, but it's not multicoated. In my "research" of UV filters, I found this test that Lensrentals did.
And then I looked at the Zeiss datasheet and it corroborated the Lensrentals test. It has a pretty hard cutoff around 410nm, which is more aggressive than most filters.
So I bought one to play around with, and it just so happened to be 49mm. On a complete whim, I tried it on the CV 50 and it seemed to address the magenta shading I was seeing. It also helped with the Tamron, maybe just a hair less though.
Now supposedly camera sensors block UV, but looking at this article, and it seems to show that Sony filters at 409nm which is less aggressive compared to some other manufacturers.
So is it possible that the Zeiss UV filter starts to block some of the higher frequency UV light? And that magenta I'm seeing is really the manifestation of UV light slightly above where Sony's cutoff is....and the Zeiss filter knocks it down?
Here are some before and after shots. I basically picked "Daylight" WB in Lightroom, to compare on equal footing.
Anyone have any theories? Is it plausible that both these lenses allow more near UV light through, and maybe it isn't fully filtered, and the cover glass causes some type of shift, making it stronger in the corners?
My Tamron 17-28 has a strong magenta cast over the whole image - clearly visible when I photograph the same scene with my Pentax-K 28mm or my CY lenses. I don't like that cast. So, instead of editing the white balance, perhaps buying one of these Zeiss filters is a good idea?
Maybe download the RAW files I posted for my Tamron 35? You can get a feel how my Tamron looks without the filter and then with the filter. I would guess you might be able to extrapolate to your 17-28, at least to some extent.
My Tamron 17-28 has a strong magenta cast over the whole image - clearly visible when I photograph the same scene with my Pentax-K 28mm or my CY lenses. I don't like that cast. So, instead of editing the white balance, perhaps buying one of these Zeiss filters is a good idea?
tsdevine wrote:
Maybe download the RAW files I posted for my Tamron 35? You can get a feel how my Tamron looks without the filter and then with the filter. I would guess you might be able to extrapolate to your 17-28, at least to some extent.
My Tamron 17-28 has a strong magenta cast over the whole image - clearly visible when I photograph the same scene with my Pentax-K 28mm or my CY lenses. I don't like that cast. So, instead of editing the white balance, perhaps buying one of these Zeiss filters is a good idea?
Thoughts from the experts about this idea?
Tim,
Where is see this exaggerated is when I take a photo that has been greatly underexposed and then try to bring it back up to proper exposure. Mainly when trying to single expose for highlight protection. The magenta rings are more noticeable in these cases. In fact there is not only magenta rings, but also some geometrically shaped ‘vignetting’ which appears to be some attempt at crude shading correction.
From what I’ve read, mainly from several dpreview threads (search for magenta rings Sony), is that this occurs even with the camera corrections turned off - unless the contacts are taped over and the camera can’t recognize the lens. Then it doesn’t occur.
I haven’t had the time to do extensive testing on this myself but it does reoccur periodically in this type of scenario. Ultimately I haven’t tried the zeiss uv filter but would be interested in a stress test if you can manage it with and without the filter when the lens is underexposed. I will also take a look at your raws for now, thank you for providing them.
If I am off target with what phenomenon you are experiencing, please let me know.
If the colour cast is in the corners rather than everywhere in the frame... and if the colour cast is a real problem for you and there aren't easier solutions...
It's a little bit of work, since it involves taking a reference frame/image of something plain and white, for each lens + camera sensor combination you want to fix corner colour casts on (and chances are nobody has already created a profile for you, for your lens/camera). But it's an option.
Cornerfix is normally/originally intended for use with ultra-wide-angle lenses, where this colour-tint-in-corners issue is commonplace (the light angles are v steep, it makes it hard to prevent in the lens itself, especially since the camera's own filter stack is often involved). But it would work for your normal focal-length lenses too.
Indeed, creating an evenly-lit plain white reference frame for an ultrawide lens is potentially a bit of a headache, so at least one aspect of this is easier for the lenses you have in mind.
I definitely am not discounting that phenomena, but if that were solely the cause, then why would I get better results with the Zeiss UV filter? Assuming there is the magenta ring problem, it takes some effort to coax it out of an image. I can't fix that.......but if I can fix what I'm seeing in let's say 95% of my shooting, I'll take it.
And my understanding of the magenta ring problem, it mainly is WO or near WO. This happens on these lenses at any aperture.
-Tim
abadger wrote: tsdevine wroteMaybe download the RAW files I posted for my Tamron 35? You can get a feel how my Tamron looks without the filter and then with the filter. I would guess you might be able to extrapolate to your 17-28, at least to some extent.
My Tamron 17-28 has a strong magenta cast over the whole image - clearly visible when I photograph the same scene with my Pentax-K 28mm or my CY lenses. I don't like that cast. So, instead of editing the white balance, perhaps buying one of these Zeiss filters is a good idea?
Thoughts from the experts about this idea?Tim,
Where is see this exaggerated is when I take a photo that has been greatly underexposed and then try to bring it back up to proper exposure. Mainly when trying to single expose for highlight protection. The magenta rings are more noticeable in these cases. In fact there is not only magenta rings, but also some geometrically shaped ‘vignetting’ which appears to be some attempt at crude shading correction.
From what I’ve read, mainly from several dpreview threads (search for magenta rings Sony), is that this occurs even with the camera corrections turned off - unless the contacts are taped over and the camera can’t recognize the lens. Then it doesn’t occur.
I haven’t had the time to do extensive testing on this myself but it does reoccur periodically in this type of scenario. Ultimately I haven’t tried the zeiss uv filter but would be interested in a stress test if you can manage it with and without the filter when the lens is underexposed. I will also take a look at your raws for now, thank you for providing them.
If I am off target with what phenomenon you are experiencing, please let me know. ...Show more →
tsdevine wrote:
I definitely am not discounting that phenomena, but if that were solely the cause, then why would I get better results with the Zeiss UV filter? Assuming there is the magenta ring problem, it takes some effort to coax it out of an image. I can't fix that.......but if I can fix what I'm seeing in let's say 95% of my shooting, I'll take it.
And my understanding of the magenta ring problem, it mainly is WO or near WO. This happens on these lenses at any aperture.
-Tim
abadger wrotetsdevine wroteMaybe download the RAW files I posted for my Tamron 35? You can get a feel how my Tamron looks without the filter and then with the filter. I would guess you might be able to extrapolate to your 17-28, at least to some extent.
My Tamron 17-28 has a strong magenta cast over the whole image - clearly visible when I photograph the same scene with my Pentax-K 28mm or my CY lenses. I don't like that cast. So, instead of editing the white balance, perhaps buying one of these Zeiss filters is a good idea?
Thoughts from the experts about this idea?Tim,
Where is see this exaggerated is when I take a photo that has been greatly underexposed and then try to bring it back up to proper exposure. Mainly when trying to single expose for highlight protection. The magenta rings are more noticeable in these cases. In fact there is not only magenta rings, but also some geometrically shaped ‘vignetting’ which appears to be some attempt at crude shading correction.
From what I’ve read, mainly from several dpreview threads (search for magenta rings Sony), is that this occurs even with the camera corrections turned off - unless the contacts are taped over and the camera can’t recognize the lens. Then it doesn’t occur.
I haven’t had the time to do extensive testing on this myself but it does reoccur periodically in this type of scenario. Ultimately I haven’t tried the zeiss uv filter but would be interested in a stress test if you can manage it with and without the filter when the lens is underexposed. I will also take a look at your raws for now, thank you for providing them.
If I am off target with what phenomenon you are experiencing, please let me know.
The easiest way for anyone curious to see what we are talking about is to take the RAW and crank up the dehaze slider to 100%. You can pretty easily see the rings as well as a very exaggerated sense of the tint problem.
Tim, in doing this I can see that it does indeed impact the magenta cast when using said UV filter, albeit not completely resolved. That is interesting.
I will try to shoot a set with the CV 50 not fully making contact electrically and also where it is. I’ll try to do a with and without filter shot in both those scenarios.
Hopefully in the next few days.
-Tim
abadger wrote: tsdevine wroteI definitely am not discounting that phenomena, but if that were solely the cause, then why would I get better results with the Zeiss UV filter? Assuming there is the magenta ring problem, it takes some effort to coax it out of an image. I can't fix that.......but if I can fix what I'm seeing in let's say 95% of my shooting, I'll take it.
And my understanding of the magenta ring problem, it mainly is WO or near WO. This happens on these lenses at any aperture.
-Tim
abadger wrotetsdevine wroteMaybe download the RAW files I posted for my Tamron 35? You can get a feel how my Tamron looks without the filter and then with the filter. I would guess you might be able to extrapolate to your 17-28, at least to some extent.
My Tamron 17-28 has a strong magenta cast over the whole image - clearly visible when I photograph the same scene with my Pentax-K 28mm or my CY lenses. I don't like that cast. So, instead of editing the white balance, perhaps buying one of these Zeiss filters is a good idea?
Thoughts from the experts about this idea?Tim,
Where is see this exaggerated is when I take a photo that has been greatly underexposed and then try to bring it back up to proper exposure. Mainly when trying to single expose for highlight protection. The magenta rings are more noticeable in these cases. In fact there is not only magenta rings, but also some geometrically shaped ‘vignetting’ which appears to be some attempt at crude shading correction.
From what I’ve read, mainly from several dpreview threads (search for magenta rings Sony), is that this occurs even with the camera corrections turned off - unless the contacts are taped over and the camera can’t recognize the lens. Then it doesn’t occur.
I haven’t had the time to do extensive testing on this myself but it does reoccur periodically in this type of scenario. Ultimately I haven’t tried the zeiss uv filter but would be interested in a stress test if you can manage it with and without the filter when the lens is underexposed. I will also take a look at your raws for now, thank you for providing them.
If I am off target with what phenomenon you are experiencing, please let me know.The easiest way for anyone curious to see what we are talking about is to take the RAW and crank up the dehaze slider to 100%. You can pretty easily see the rings as well as a very exaggerated sense of the tint problem.
Tim, in doing this I can see that it does indeed impact the magenta cast when using said UV filter, albeit not completely resolved. That is interesting.
Consider these preliminary as I did them in a hurry and realized afterward that partially dismounting the lens (so the contacts don't make connections) changes the exposure. I ran these all through Lightroom, chose "Daylight" color temp and tried to equalize the exposure. All were taken with the CV 50 on my a7R III.
Electrical contact between the lens and the camera, no filter
There is a bit of a precedent for something like this happening. In the earlier days of micro 4/3 some Panasonic lenses did not behave all that well on Olympus bodies. As far as I remember the reason was that Panasonic was doing more of the UV IR filtering in their stack and the glass was letting more pollution through the Olympus sensor.
It would follow that the sharper cut off of the size filter is preventing the issue that you’re seeing here. The filter is basically blocking the wavelength that’s causing the mild issue.
Thanks, that does seem to me to be the most obvious answer. For some reason a hint of magenta really catches my eye, and this turns out to be a very workable solution for me.
AdaptedLenses wrote:
There is a bit of a precedent for something like this happening. In the earlier days of micro 4/3 some Panasonic lenses did not behave all that well on Olympus bodies. As far as I remember the reason was that Panasonic was doing more of the UV IR filtering in their stack and the glass was letting more pollution through the Olympus sensor.
It would follow that the sharper cut off of the size filter is preventing the issue that you’re seeing here. The filter is basically blocking the wavelength that’s causing the mild issue.
tsdevine wrote:
Consider these preliminary as I did them in a hurry and realized afterward that partially dismounting the lens (so the contacts don't make connections) changes the exposure. I ran these all through Lightroom, chose "Daylight" color temp and tried to equalize the exposure. All were taken with the CV 50 on my a7R III.
Thank you for doing this test! So in reviewing these, the magenta remains in all photos, but is dampened by the UV filter which does improve the look as you said (and adds a very slight green tint). In the contactless photos, the concentric rings have disappeared. So I would call the shading compensation theory correct (the camera bakes in shading correction even with all corrections turned off), while also noting that the magenta tint probably is independent of shading correction.
Also of note that the magenta seems to appear near the corners of the images most prominently, and in some cases flares to one side or the other. So the direction of light input does appear to impact it, and that would make some sense if it was from UV filtering.
Yeah, it gets me close enough that when I do my color balance, nothing sticks out as jarring.
abadger wrote: tsdevine wroteConsider these preliminary as I did them in a hurry and realized afterward that partially dismounting the lens (so the contacts don't make connections) changes the exposure. I ran these all through Lightroom, chose "Daylight" color temp and tried to equalize the exposure. All were taken with the CV 50 on my a7R III.
I'll try to do a more controlled test later this week or this weekend.
-TimThank you for doing this test! So in reviewing these, the magenta remains in all photos, but is dampened by the UV filter which does improve the look as you said (and adds a very slight green tint). In the contactless photos, the concentric rings have disappeared. So I would call the shading compensation theory correct (the camera bakes in shading correction even with all corrections turned off), while also noting that the magenta tint probably is independent of shading correction.
Also of note that the magenta seems to appear near the corners of the images most prominently, and in some cases flares to one side or the other. So the direction of light input does appear to impact it, and that would make some sense if it was from UV filtering....Show more →
abadger wrote: tsdevine wroteConsider these preliminary as I did them in a hurry and realized afterward that partially dismounting the lens (so the contacts don't make connections) changes the exposure. I ran these all through Lightroom, chose "Daylight" color temp and tried to equalize the exposure. All were taken with the CV 50 on my a7R III.
I'll try to do a more controlled test later this week or this weekend.
-TimThank you for doing this test! So in reviewing these, the magenta remains in all photos, but is dampened by the UV filter which does improve the look as you said (and adds a very slight green tint). In the contactless photos, the concentric rings have disappeared. So I would call the shading compensation theory correct (the camera bakes in shading correction even with all corrections turned off), while also noting that the magenta tint probably is independent of shading correction.
Also of note that the magenta seems to appear near the corners of the images most prominently, and in some cases flares to one side or the other. So the direction of light input does appear to impact it, and that would make some sense if it was from UV filtering....Show more →
How can we find out whether it's the camera or the raw conversion software that is adding the rings? Does it happen with all raw converters? I remember reading some time ago that Lightroom was applying some "corrections" that one couldn't disable (the buttons were grayed out) and the only possible cure was indeed to tape the lens contacts. Is it still the case with LR?
vdo1 wrote: abadger wrotetsdevine wroteConsider these preliminary as I did them in a hurry and realized afterward that partially dismounting the lens (so the contacts don't make connections) changes the exposure. I ran these all through Lightroom, chose "Daylight" color temp and tried to equalize the exposure. All were taken with the CV 50 on my a7R III.
I'll try to do a more controlled test later this week or this weekend.
-TimThank you for doing this test! So in reviewing these, the magenta remains in all photos, but is dampened by the UV filter which does improve the look as you said (and adds a very slight green tint). In the contactless photos, the concentric rings have disappeared. So I would call the shading compensation theory correct (the camera bakes in shading correction even with all corrections turned off), while also noting that the magenta tint probably is independent of shading correction.
Also of note that the magenta seems to appear near the corners of the images most prominently, and in some cases flares to one side or the other. So the direction of light input does appear to impact it, and that would make some sense if it was from UV filtering.How can we find out whether it's the camera or the raw conversion software that is adding the rings? Does it happen with all raw converters? I remember reading some time ago that Lightroom was applying some "corrections" that one couldn't disable (the buttons were grayed out) and the only possible cure was indeed to tape the lens contacts. Is it still the case with LR?
I can’t immediately find the thread but I do recall people having the same issue with Capture One, for what it’s worth. So I think the correction has already been baked into the file in camera.
I think I see it faintly... These were processed though Photo Ninja.
-Tim
Both exhibit concentric circles from shading compensation. The second exhibits clearly less magenta tinting than the first. So I assume the second has the UV filter attached. A big improvement if we want to partially remove unwanted magenta tinting.