This topic is in the 2015 archive. You can read it and like its posts here. It is read only.
Archive 2015(updated) Lens colour difference test
p.2 #1(updated) Lens colour difference test
jcolwell wrote:
Update: the first set of colour slices shown below is modified, as suggested by many replies to the original posting. The mid-grey panel above the "TM" mark in the ColorChecker was used to balance colour and exposure across all eight original RAW files, to RGB(127,127,127). There is slight variation in illumination vertically. The centre areas of all mid-grey panels are within +/-1 of 127 for all lenses.
What seems very odd is that the two chart versions (equalized exposure and non-equalized exposure) seem identical with regard to value readings in each respective pair (either color or converted to gray scale).
Edit. Now they look different (the first version now alphabetically labeled). That looks more like I would expect.
One thing I have noticed running similar tests (and as also mentioned by others) is that the subtle hues can differ more than the stronger, saturated primaries. On the color checker chart, that is of course the first two rows.
I found for myself, still better, is to place the colorchecker chart in a natural setting that includes a lot of subtle shades of green foliage, tree bark and so on (I also include something earthy and fleshy like clay pots/ heads).
Update: the first set of colour slices shown below is modified, as suggested by many replies to the original posting. The mid-grey panel above the "TM" mark in the ColorChecker was used to balance colour and exposure across all eight original RAW files, to RGB(127,127,127). There is slight variation in illumination vertically. The centre areas of all mid-grey panels are within +/-1 of 127 for all lenses.
What seems very odd is that the two chart versions (equalized exposure and non-equalized exposure) seem identical with regard to value readings in each respective pair (either color or converted to gray scale).
Edit. Now they look different (the first version now alphabetically labeled). That looks more like I would expect.
One thing I have noticed running similar tests (and as also mentioned by others) is that the subtle hues can differ more than the stronger, saturated primaries. On the color checker chart, that is of course the first two rows.
I found for myself, still better, is to place the colorchecker chart in a natural setting that includes a lot of subtle shades of green foliage, tree bark and so on (I also include something earthy and fleshy like clay pots/ heads).
Sep 09, 2015 at 07:44 PM
p.2 #2(updated) Lens colour difference test
Ha! This is amazing, i can't believe how much work was put in here, very well done. The results pretty much look how I'd expect, and for what it's worth, studio strobes are pretty similar to daylight spectrally, so i think your findings here are very relevant. To me, checking saturated colors like this makes good sense as well, but evaluating full spectrum light coming through a lens on a 3 channel system is always going to be tricky. Anyways, I appreciate what's been done here, this is cool.
I guess I'm curious though, no one seems to have drawn any strong conclusions here. Personally, I'd say these differences in the equalized version (i agree with that logic too, mostly) are pretty slight, but I'm sure some would disagree. We get into all kinds of murky territory with adjusting exposure and Adobe's hue shifting tactics as well (See chromosoft blog). But then, if you changed your strobes instead, the color balance of the lights could change a little. So, doing this test "perfectly" could be dang near impossible, but to me it's all close enough here to establish that the differences in color between lenses aren't dramatic. But, the jumping off point between art and science is different for everyone and my subtle is probably someone else's dramatic.
Thank you for doing this, such a cool test and conversation.
I guess I'm curious though, no one seems to have drawn any strong conclusions here. Personally, I'd say these differences in the equalized version (i agree with that logic too, mostly) are pretty slight, but I'm sure some would disagree. We get into all kinds of murky territory with adjusting exposure and Adobe's hue shifting tactics as well (See chromosoft blog). But then, if you changed your strobes instead, the color balance of the lights could change a little. So, doing this test "perfectly" could be dang near impossible, but to me it's all close enough here to establish that the differences in color between lenses aren't dramatic. But, the jumping off point between art and science is different for everyone and my subtle is probably someone else's dramatic.
Thank you for doing this, such a cool test and conversation.
Sep 10, 2015 at 02:07 AM
p.2 #3(updated) Lens colour difference test
Thanks for the update, Jim.
The difference in lens contrast is surprisingly high.
I downloaded your updated comparison add ajusted levels for equal luminance in the green patch. What I see is an obvious difference in stauration, which I suppose is explained by the difference in contrast. I still don't see any differences in colour hue.
The difference in lens contrast is surprisingly high.
I downloaded your updated comparison add ajusted levels for equal luminance in the green patch. What I see is an obvious difference in stauration, which I suppose is explained by the difference in contrast. I still don't see any differences in colour hue.
Sep 10, 2015 at 03:13 AM
p.2 #4(updated) Lens colour difference test
Something strange is happening with the two colour slice composite images; the same image of old slices is being shown for both cases. I've had similar problems in the past with the 'replace image' edit option, which I used to replace the old-slices image with the new balanced-slices image. Later, I added the old-slices as a new third image. I'm going to delete both slices images and then re-load them...
Sep 10, 2015 at 05:10 AM
p.2 #5(updated) Lens colour difference test
jcolwell wrote:
Something strange is happening with the two colour slice composite images; the same image of old slices is being shown for both cases. I've had similar problems in the past with the 'replace image' edit option, which I used to replace the old-slices image with the new balanced-slices image. Later, I added the old-slices as a new third image. I'm going to delete both slices images and then re-load them...
Thanks again, Jim. Now I see the colour differences
Something strange is happening with the two colour slice composite images; the same image of old slices is being shown for both cases. I've had similar problems in the past with the 'replace image' edit option, which I used to replace the old-slices image with the new balanced-slices image. Later, I added the old-slices as a new third image. I'm going to delete both slices images and then re-load them...
Thanks again, Jim. Now I see the colour differences
Sep 10, 2015 at 05:19 AM
p.2 #6(updated) Lens colour difference test
Tariq Gibran wrote:
What seems very odd is that the two chart versions (equalized exposure and non-equalized exposure) seem identical with regard to value readings in each respective pair (either color or converted to gray scale).
Edit. Now they look different (the first version now alphabetically labeled). That looks more like I would expect. ...
Thanks Tariq, that verifies my memory, as discussed just above. The first attempt to upload a new version of the slices didn't "take". The secons attempt appeared to work, but it had reverted to the original old-slices image when I checked again this morning. Hopefully, it'll be stable this time.
What seems very odd is that the two chart versions (equalized exposure and non-equalized exposure) seem identical with regard to value readings in each respective pair (either color or converted to gray scale).
Edit. Now they look different (the first version now alphabetically labeled). That looks more like I would expect. ...
Thanks Tariq, that verifies my memory, as discussed just above. The first attempt to upload a new version of the slices didn't "take". The secons attempt appeared to work, but it had reverted to the original old-slices image when I checked again this morning. Hopefully, it'll be stable this time.
Sep 10, 2015 at 06:41 AM
p.2 #7(updated) Lens colour difference test
jcolwell wrote:
Edit. Now they look different (the first version now alphabetically labeled). That looks more like I would expect. ...Thanks Tariq, that verifies my memory, as discussed just above. The first attempt to upload a new version of the slices didn't "take". The secons attempt appeared to work, but it had reverted to the original old-slices image when I checked again this morning. Hopefully, it'll be stable this time.
That has happened with some of my uploads in the past. Very frustrating I know.
Tariq Gibran wrote
What seems very odd is that the two chart versions (equalized exposure and non-equalized exposure) seem identical with regard to value readings in each respective pair (either color or converted to gray scale).Edit. Now they look different (the first version now alphabetically labeled). That looks more like I would expect. ...
That has happened with some of my uploads in the past. Very frustrating I know.
Sep 10, 2015 at 09:32 AM
p.2 #8(updated) Lens colour difference test
Jim -- Thanks for the comparisons and sparking discussion on lens color characteristics.
I've found in my comparisons of different lenses and designs shot wide open, that without inherent aberrations nullified by smaller aperture settings, the color cast is much more pronounced. Extreme examples include the old Pentax thorium elements aged to a near amber color. Modern Zeiss vs. Canon colors and contrast differences are often striking, producing visible differences at wider apertures.
Cinematographers are extremely conscientious about lens testing, even within brands, for most consistent color and contrast to maintain accuracy. It's almost a religion in film making and videography.
I've found in my comparisons of different lenses and designs shot wide open, that without inherent aberrations nullified by smaller aperture settings, the color cast is much more pronounced. Extreme examples include the old Pentax thorium elements aged to a near amber color. Modern Zeiss vs. Canon colors and contrast differences are often striking, producing visible differences at wider apertures.
Cinematographers are extremely conscientious about lens testing, even within brands, for most consistent color and contrast to maintain accuracy. It's almost a religion in film making and videography.
Sep 10, 2015 at 10:45 AM
p.2 #9(updated) Lens colour difference test
jcolwell wrote:
Tough crowd.
Pfft. Any competent internet poster would know that.
Tough crowd.
Pfft. Any competent internet poster would know that.
Sep 10, 2015 at 05:56 PM
p.2 #10(updated) Lens colour difference test
jcolwell wrote:
Tough crowd.
freaklikeme wrote:
Pfft. Any competent internet poster would know that.
Competent? What's that, a race to set up tents?
Tough crowd.
freaklikeme wrote:
Pfft. Any competent internet poster would know that.
Competent? What's that, a race to set up tents?
Sep 10, 2015 at 06:45 PM
p.2 #11(updated) Lens colour difference test
jcolwell wrote:
Competent? What's that, a race to set up tents?
Yes, sister competition to the charity event Compotent, in which they race to set up tents for underprivileged campers.
jcolwell wrote
Tough crowd.freaklikeme wrote
Pfft. Any competent internet poster would know that.Yes, sister competition to the charity event Compotent, in which they race to set up tents for underprivileged campers.
Sep 10, 2015 at 06:56 PM