I find this color cast issue very interesting. Back over a year ago when I was considering getting the 50mm APO I noticed color cast, especially towards the corners in some of the photos people where posting here. I actually mentioned seeing a color cast and multiple people told me it's not something they have noticed using the lens. I did end up purchasing the lens and have enjoyed using it.
Anyway, seeing this color cast conversation I decide to purchase the Zeiss UV filter in both 49mm (for the 50mm APO) and 77mm for other lenses to experiment with. I have only done basic outdoor testing with it so far but, it looks like it definitely helps.
While it definitely helps the 50mm APO, I also noticed it helped my Sony 24-105mm f4. On both lenses the color just look more natural to my eye and have less of a magenta cast, especially in the clouds with the filter on. I also noticed a small shift in the red histogram when using the Zeiss UV filter.
For a while I have been in the camp of Sony colors not being as nice as Nikon colors. I have spent many hours in LR playing with color settings trying to Sony colors to where I like them. I wonder if this magenta casting has something to do with it. Going to have to also test what happens when I put the Zeiss UV filter on a Nikon lens mounted to my Sony camera. I also want to try and understand if the Zeiss UV filter maybe tinting the colors in a way I maybe find more pleasing out of camera, or is purely just fixing spectrum pollution. Since if it's causing a tint maybe I can replicate that tint in post for lenses I have which I can't put a Zeiss UV filter on.
This is blasphemy, but I look at it this way. The argument would be that this filter should make no difference, as the Sony stack has a hard cutoff at 420nm. Given that it seems to have impact, at least in lenses where there is something obviously going on...doesn't mean it wouldn't have some impact on shots from lenses where it would normally be even harder to notice. The only harm being done is there is another piece of glass in the optical path, arguably from a color perspective, the results you are seeing with the filter is what Sony intended. Given that the stack should be filtering at 420nm.......how can it be argued that a filter that cuts at 410nm is somehow impacting color in a negative way (it should be doing nothing to color really.)
So yeah, I get the whole another piece of glass in the optical path, and you should avoid it. But if you feel the results are better, then so be it, use it. I mean we have people using PCX filters to make some short flange distance legacy lenses perform better. The results are what matter, at least to me. If you find it really doesn't make a tangible difference on a lens, then no need to use it.
-Tim
DWOfPaul wrote:
I find this color cast issue very interesting. Back over a year ago when I was considering getting the 50mm APO I noticed color cast, especially towards the corners in some of the photos people where posting here. I actually mentioned seeing a color cast and multiple people told me it's not something they have noticed using the lens. I did end up purchasing the lens and have enjoyed using it.
Anyway, seeing this color cast conversation I decide to purchase the Zeiss UV filter in both 49mm (for the 50mm APO) and 77mm for other lenses to experiment with. I have only done basic outdoor testing with it so far but, it looks like it definitely helps.
While it definitely helps the 50mm APO, I also noticed it helped my Sony 24-105mm f4. On both lenses the color just look more natural to my eye and have less of a magenta cast, especially in the clouds with the filter on. I also noticed a small shift in the red histogram when using the Zeiss UV filter.
For a while I have been in the camp of Sony colors not being as nice as Nikon colors. I have spent many hours in LR playing with color settings trying to Sony colors to where I like them. I wonder if this magenta casting has something to do with it. Going to have to also test what happens when I put the Zeiss UV filter on a Nikon lens mounted to my Sony camera. I also want to try and understand if the Zeiss UV filter maybe tinting the colors in a way I maybe find more pleasing out of camera, or is purely just fixing spectrum pollution. Since if it's causing a tint maybe I can replicate that tint in post for lenses I have which I can't put a Zeiss UV filter on. ...Show more →
Color matching of buildings based on Zeiss T* UV，and increase saturation and contrast，it can be seen that the corners of Rodenstock and Sony are shift to red（Inside the box is the color of the upper right corner）
Juha Kannisto wrote:
Cosina announced the official release date for the E-mount version of APO 35/2 today. It’s 9th of April, earlier than what I expected.
I don't want to hear any more complaints about the 35mm FL for the E-mount.
We now have Sigma 35/2 DG DN, Sony 35mm f/1.4 GM and Voigtlander 35/2 APO. All solid choices. Plus many other previous 35mm models selling for bargain right now.
Fred Miranda wrote:
A few days before the Sony GM in the US. :-)
I don't want to hear any more complaints about the 35mm FL for the E-mount.
We now have Sigma 35/2 DG DN, Sony 35mm f/1.4 GM and Voigtlander 35/2 APO. All solid choices. Plus many other previous 35mm models selling for bargain right now.
GM started selling in Japan today. Not picking one up now though as it's relatively expensive here and I've got the Sigma 35/2 DG DN, Sony Zeiss 35/1.4, CV 35/1.2 SE & 35/1.4 classic already and I've preordered the 35/2 APO as well... Feeling well covered for 35mm.
Fred Miranda wrote:
A few days before the Sony GM in the US. :-)
I don't want to hear any more complaints about the 35mm FL for the E-mount.
We now have Sigma 35/2 DG DN, Sony 35mm f/1.4 GM and Voigtlander 35/2 APO. All solid choices. Plus many other previous 35mm models selling for bargain right now.
You don't understand how it works--now we complain because we have too many options
Juha Kannisto wrote:
GM started selling in Japan today. Not picking one up now though as it's relatively expensive here and I've got the Sigma 35/2 DG DN, Sony Zeiss 35/1.4, CV 35/1.2 SE & 35/1.4 classic already and I've preordered the 35/2 APO as well... Feeling well covered for 35mm.
What is the price difference of the 35i versus the 35 GM by you right now? Here (stateside) it's $600 vs. $1,400--the Sigma cost 43% of the GM...which is a major reason the GM is not tempting me. And I am flush with 35mm’s as well.
nehemiahphoto wrote:
What is the price difference of the 35i versus the 35 GM by you right now? Here (stateside) it's $600 vs. $1,400--the Sigma cost 43% of the GM...which is a major reason the GM is not tempting me. And I am flush with 35mm’s as well.
Referring to Map Camera for prices as that's where I would buy in most cases:
So Sigma is slightly below 39% of the GM price (GM is 2.57 times as expensive).
In the Covid-19 times there are no events that I would normally shoot (especially all kinds of Japanese summer festivals including dance performances) so I'm mostly keen on lighter and smaller lenses for my regular walkaround shooting purposes and that also makes faster / bigger and more expensive lenses less desirable. At other times I might be more tempted to pick up the GM to replace my Sony Zeiss 35/1.4.
tsdevine wrote:
This is blasphemy, but I look at it this way. The argument would be that this filter should make no difference, as the Sony stack has a hard cutoff at 420nm. Given that it seems to have impact, at least in lenses where there is something obviously going on...doesn't mean it wouldn't have some impact on shots from lenses where it would normally be even harder to notice. The only harm being done is there is another piece of glass in the optical path, arguably from a color perspective, the results you are seeing with the filter is what Sony intended. Given that the stack should be filtering at 420nm.......how can it be argued that a filter that cuts at 410nm is somehow impacting color in a negative way (it should be doing nothing to color really.)
So yeah, I get the whole another piece of glass in the optical path, and you should avoid it. But if you feel the results are better, then so be it, use it. I mean we have people using PCX filters to make some short flange distance legacy lenses perform better. The results are what matter, at least to me. If you find it really doesn't make a tangible difference on a lens, then no need to use it.
'the Sony stack has a hard cutoff at 420nm': where are you sourcing that from Tim? realViveks A7rII transmission graph on the previous page doesn't show that.
'how can it be argued that a filter that cuts at 410nm is somehow impacting color': because the UV filter doesn't have a hard cut at 410: the gradient of the curve is not vertical on the Zeiss UV filter shown on this page https://www.zeiss.com/content/dam/consumer-products/downloads/photography/brochures/en/brochure-zeiss-filter-en.pdf ,and it doesn't round off till about 420nm at 98% transmission - so it'll be taking some of the visible blues out of the image; the transmission curve also isn't flat, reaching only 100% between 575-650nm which will also change the overall tint of the image.
The Kenko transmission graph shown on this listing https://www.ebay.com/itm/Kenko-49mm-Zeta-L41-UV-ZR-Coated-Slim-Frame-Camera-Lens-Filters-from-Japan-/313278078882 takes even more of the visible blues out without degrading the reds like the Zeiss, so should look warmer still...
Juha Kannisto wrote:
GM started selling in Japan today. Not picking one up now though as it's relatively expensive here and I've got the Sigma 35/2 DG DN, Sony Zeiss 35/1.4, CV 35/1.2 SE & 35/1.4 classic already and I've preordered the 35/2 APO as well... Feeling well covered for 35mm.
Juha, what are your thoughts on 35/1.4 classic? I was thinking about picking one up as a tiny fun lens to play with, but some reviews are pretty bad for the lens. I am pretty happy wih my Bigma 35 and 35 ZA combo, but feeling an urge to try a manual lens in this range.
That came from Kolari via Lloyd Chambers, but I agree that Vivek's test doesn't support that. Maybe it starts up near 420nm...but it doesn't do a hard cut there. In published content, Kolari puts Sony around 409nm, which puts it lower than any other major manufacturer. Fuji is up at 422nm.
And still, if we go by Vivek's test, the stack filtering fits within the Kenko curve, or is so close to it in portions you can't differentiate. The only wavelengths where the Kenko seems to block more is between 380-400nm. And we're talking a few percentage points.
The published (transmission curves for the Zeiss and Kenko appear to be pretty close. Maybe the Kenko filters slightly more. If you go by Lensrentals and Viveks test, they look fairly close.
I do think the T&Y is probably overkill given Vivek's test. It will affect exposure and knock more visible spectrum from the end result.
Given Vivek's test, I stand by that the Zeiss and Kenko filters do not appear to be drastically different, or should filter much more than Sony intended via the sensor stack. The other wavelength you mention are already attenuated much more by the filter stack (via Vivek's) test. I don't think the few percentage points makes a huge difference IMHO.
Take my comments with a grain of salt. If you look at my samples, and even more importantly, your own first hand experience...and you see nothing untoward, then no worries. If you do see something that bothers you, you might want to try the Zeiss or Kenko filter. But it's everyone's choice. I definitely wouldn't tell anyone they "had" to use a filter, and I expect no one to tell me I "absolutely" shouldn't.
-Tim
Petegh wrote:
'the Sony stack has a hard cutoff at 420nm': where are you sourcing that from Tim? realViveks A7rII transmission graph on the previous page doesn't show that.
'how can it be argued that a filter that cuts at 410nm is somehow impacting color': because the UV filter doesn't have a hard cut at 410: the gradient of the curve is not vertical on the Zeiss UV filter shown on this page https://www.zeiss.com/content/dam/consumer-products/downloads/photography/brochures/en/brochure-zeiss-filter-en.pdf ,and it doesn't round off till about 420nm at 98% transmission - so it'll be taking some of the visible blues out of the image; the transmission curve also isn't flat, reaching only 100% between 575-650nm which will also change the overall tint of the image.
The Kenko transmission graph show on this listing https://www.ebay.com/itm/Kenko-49mm-Zeta-L41-UV-ZR-Coated-Slim-Frame-Camera-Lens-Filters-from-Japan-/313278078882 takes even more of the visible blues out without degrading the reds like the Zeiss, so should look warmer still......Show more →
Tim, I do not agree with the speculation that there is some edge/corner leakage of the Sony filter stack. I already posted that in reply to inks&paper. The Sony stacks are virtually indistinguishable from that of a Kenko Zeta L41 filter when it comes to UV cutoff.
I used a calibrated Ocean optics spectrometer with a cosine corrector for my measurements. There are no details from the other reported measurements.
The STC595 in that graph refers to a STC drop in filter (meant to be used with a full spectrum camera).
Then what is the explanation? Is in the increase filtering between 380-400nm? Is there cumulative effect, where the filter and stack together are additive in nature? Could it be just my cameras? Is there some fluorescence in these lenses? Or with certain wavelengths the lenses are skewing the wavelengths towards the edges and corners? What are the hypothetical causes that haven't been ruled out?
Is your test average this across the whole frame or sampling a portion? This curve is exactly the same in the center and at the edges and corners, with no deviation, including with a lens mounted where ray angle may come into play more?
I'll revise my post and soften my comments around the "why". I will say "no one knows" from now on, that seems to be the safest bet.
My last question would be, based on your test, do you feel there should be as noticeable change between bare lens and Kenko filtered lens photos? In other words, is your test confirmed in the test shots I have posted and vice versa?
Here is more info about the Lensrentals test for the Zeiss filter graph I posted:
I only mentioned the manufacturer posted transmission graphs and the Lensrentals one. If the Lensrentals test isn't transparent enough in how the test was done, I guess we have to ignore their test.
realVivek wrote:
Tim, I do not agree with the speculation that there is some edge/corner leakage of the Sony filter stack. I already posted that in reply to inks&paper. The Sony stacks are virtually indistinguishable from that of a Kenko Zeta L41 filter when it comes to UV cutoff.
I used a calibrated Ocean optics spectrometer with a cosine corrector for my measurements. There are no details from the other reported measurements.
The STC595 in that graph refers to a STC drop in filter (meant to be used with a full spectrum camera).
Would you expect both lenses to have similar ray angle, so can that be ruled out? If not, could it just be a shift related to sensor stack/BSI in a certain portion of spectrum with certain ray angles?
Or is it truly something within the lens design itself that is shifting portions of the wavelength?
ramesesthe2nd wrote:
Juha, what are your thoughts on 35/1.4 classic? I was thinking about picking one up as a tiny fun lens to play with, but some reviews are pretty bad for the lens. I am pretty happy wih my Bigma 35 and 35 ZA combo, but feeling an urge to try a manual lens in this range.
I do like it quite a bit. It's funky wide open and I really like its classic rendering there in many shots, and it also works as a good corner-to-corner sharp lens stopped down to f8 (and is almost as good from f5.6). I usually keep the lens correction settings at auto for distortion and purple fringing and I think it helps with any distortion that the lens might otherwise have. In Japan it's been among the most popular CV E-mount lenses right after 50/2 and 65/2 APO-Lanthars. I think the size and weight are also pretty much perfect for smaller cameras like my A7C.
tsdevine wrote:
So yeah, I get the whole another piece of glass in the optical path, and you should avoid it. But if you feel the results are better, then so be it, use it. I mean we have people using PCX filters to make some short flange distance legacy lenses perform better. The results are what matter, at least to me. If you find it really doesn't make a tangible difference on a lens, then no need to use it.
-Tim
I agree at the end of the day it's the results that matter. I have to admit though I do find it a bit confusing to be buying UV filters after always thinking their basically a gimmick.
Another taught I had on this is I wonder if it has something to do with blocking the UVC light from going into the lens from the start, VS trying to block the UV light at the rear with the sensor stack. We know light passing through our lenses can get manipulated which leads to issues like CA. Maybe the UV light is getting mixed up to so that by the time it hits the sensor stack some of it is no longer in the frequency range that is blocked by the sensor stack.
