p.10 #1 · First Sighting: A7rii Thin filter in action
dpap1978 wrote:
Uhoh7, I have a zm 21/4.5 to lend you if you want to see how it peforms on the thin filter vs the M9
PM sent. That would be very interesting
I'm pink, I think....M9 below with 28cron: M9_28cron by unoh7, on Flickr
In this shot the M9 was pretty dark in the building as it often is in high contrast. So you must push the shadows alot to get here, shooting into the veiled sun which is on the right not far out of the frame.
p.10 #2 · First Sighting: A7rii Thin filter in action
uhoh7 wrote:
Yes, looks just as bad as a M wide on the stock camera
Shows just how much the lenses must be tweaked to work with the many layers of glass over the stock sensor.
I think Illija is calling this version the "ultra-thin" (UT). V4 may get confusing as 1-3 are standard thickness IR cuts, and Kolari may try another of that thickness in the future, like the STC. Would be nice to have a short designation which makes it clear which is the new IR technology. My samples are designated UT. Can we use that so readers are not confused? There may actually be more UT versions, also.
The UT is fundamentally different in manufacture, and may present the photographer with equally different IR and WB issues. It is completely new, and this is the first time ever on any FF camera or large sensor camera.
p.10 #4 · First Sighting: A7rii Thin filter in action
hiepphotog wrote:
Derek, nice comparison! Exactly what I'm looking for. Since you have the G 28 and 21, would you post some similar comparison? Much appreciated.
here is the g28 comparison between stock, kolari UT, and reversed optosigma pcx 1.5m filter on stock a7. i shot the kolari UT a7 + g28 combo focused in the center, outer midzone, and corner. i only focused on pcx filter in the center because i didn't see any real difference between focus locations moving outward with the pcx filter. the kolari is clearly better in the center and midzone, but it's quite close in the extreme corner.
p.10 #5 · First Sighting: A7rii Thin filter in action
sebboh wrote:
here is the g28 comparison between stock, kolari v4 (UT), and reversed optosigma pcx 1.5m filter on stock a7. i shot the kolari v4 a7 + g28 combo focused in the center, outer midzone, and corner. i only focused on pcx filter in the center because i didn't see any real difference between focus locations moving outward with the pcx filter. the kolari is clearly better in the center and midzone, but it's quite close in the extreme corner.
WOW, that is a very good result. From my experience, the G21 is about similar to the G28 (probably a tad better). Will wait for your G21 results, but I think I might just send mine in for the V4. Thank you Derek.
p.10 #6 · First Sighting: A7rii Thin filter in action
Damn! There's gonna be a long queue of cameras waiting at Kolari's shelves for the v4 mod. I bet that when I send mine, they've run out of UT glass and I'll have to wait for weeks
Great job sebboh and uhoh. Your pictures are worth more than thousand words.
p.10 #7 · First Sighting: A7rii Thin filter in action
Superb Tests, Derek. Thank you so much for those
Those PCX really seem case by case, so people should not assume UT is beating PCX in all cases, like 35/1.4 ZM, or the 21 they like. It might be better, but we don't know, yet.
artur5 wrote:
Damn! There's gonna be a long queue of cameras waiting at Kolari's shelves for the v4 mod. I bet that when I send mine, they've run out of UT glass and I'll have to wait for weeks
Great job sebboh and uhoh. Your pictures are worth more than thousand words.
TY sir
Regarding what we call it:
From Ilija at Kolari:
"I like Ultra Thin.
Oh the certificate. I forgot to update that, will change"
So please let's not use V4. UT is good. If we get another version of it it can be UT2. This way years from now we don't have to remember which "V" is the thin one. Kolari might even replace base coverglass someday, like NK.
This was a pretty big gamble by Illja, and maybe it's going to work
I shot a large group with M9 and A7.ut at 35mm and let my client choose. He picked the UT, to my shock. So WB is not completely out there. Once Derek has shot the UT a few weeks I think we will really get an idea of how challenging it is. What I notice is, the AWB has greater variation in response to lighting with UT than M9 or V2. However it does seem to be able to be "brought back" to quite a degree. Long run perhaps multiple camera profiles will make it easier.
p.10 #8 · First Sighting: A7rii Thin filter in action
hiepphotog wrote:
WOW, that is a very good result. From my experience, the G21 is about similar to the G28 (probably a tad better). Will wait for your G21 results, but I think I might just send mine in for the V4. Thank you Derek.
i have strange behavior with the g21 – it's slightly decentered with one corner looking bad on the kolari mod and the opposite corner looking bad with the pcx filter. this makes it impossible to compare the two in a way that i think is fair. i tried recentering the pcx filter, but that didn't make a difference and i tried placing the lens back in it's original contax g body, but that also didn't make a difference.
p.10 #9 · First Sighting: A7rii Thin filter in action
@ uhoh7 :
Looking at your two samples with the A7-UT and the Leica lenses, I see a color difference. The one with the Cron 28 looks reddish while the SEM21 tends clearly to pink. Maybe the sky conditions changed between the shots and you used Auto WB?. If not, this is rather weird. I’d think that modern Leica lenses should have a very similar color rendering.
Instead, your shot of the color checker card has an excellent color balance. The grey scale is almost perfectly neutral and the black anorak looks black and nothing else. Did you use a custom profile ?.
I say that because I tried to correct the reddish cast in your picture of the Opera house with the A7-UT+ SEM21. Using a cold filter in Photoshop the snow came perfectly white but the green foliage was still more brown than green, specially if compared with the greens that we see in the M9 shot. Therefore, I suspect that a custom WB for the UT won’t be enough to get accurate colors. A custom profile seems also compulsory.
p.10 #10 · First Sighting: A7rii Thin filter in action
sebboh wrote:
i have strange behavior with the g21 – it's slightly decentered with one corner looking bad on the kolari mod and the opposite corner looking bad with the pcx filter. this makes it impossible to compare the two in a way that i think is fair. i tried recentering the pcx filter, but that didn't make a difference and i tried placing the lens back in it's original contax g body, but that also didn't make a difference.
p.10 #11 · First Sighting: A7rii Thin filter in action
artur5 wrote:
@ uhoh7 :
Looking at your two samples with the A7-UT and the Leica lenses, I see a color difference. The one with the Cron 28 looks reddish while the SEM21 tends clearly to pink. Maybe the sky conditions changed between the shots and you used Auto WB?. If not, this is rather weird. I’d think that modern Leica lenses should have a very similar color rendering.
Instead, your shot of the color checker card has an excellent color balance. The grey scale is almost perfectly neutral and the black anorak looks black and nothing else. Did you use a custom profile ?.
I say that because I tried to correct the reddish cast in your picture of the Opera house with the A7-UT+ SEM21. Using a cold filter in Photoshop the snow came perfectly white but the green foliage was still more brown than green, specially if compared with the greens that we see in the M9 shot. Therefore, I suspect that a custom WB for the UT won’t be enough to get accurate colors. A custom profile seems also compulsory.
p.10 #12 · First Sighting: A7rii Thin filter in action
Of course, jpgs aren’t raws (they’re already "cooked" ) but applying a photography filter to a jpg file, it works almost as modifying the color temperature of the shot. Basically the same thing that in film days. when we used CC filters on the lens to compensate for different light conditions. In this case, a used a weak cyan filter that rendered the snow quite neutral, but the greens of the foliage still looked a dull green-brown instead of shinny green like in the M9 sample.
p.10 #13 · First Sighting: A7rii Thin filter in action
artur5 wrote:
Of course, jpgs aren’t raws (they’re already "cooked" ) but applying a photography filter to a jpg file, it works almost as modifying the color temperature of the shot. Basically the same thing that in film days. when we used CC filters on the lens to compensate for different light conditions. In this case, a used a weak cyan filter that rendered the snow quite neutral, but the greens of the foliage still looked a dull green-brown instead of shinny green like in the M9 sample.
i've never been able to accurately correct WB with jpegs, so i wouldn't worry too much about that aspect. i've also never been able to replicate m9 greens with an a7 even shooting raw. i'm sure there is a little difference in terms of what one can get colorwise with the stock a7 vs the a7 kolari UT, but it's not jumping out at me comparing them side by side after correcting WB.
is there any kind of color comparison you'd like me to do while i have both cameras?
p.10 #14 · First Sighting: A7rii Thin filter in action
Thanks for your feedback and offer of a color comparison. I really appreciate that.
As you said, for fine tuning the color output of a camera, jpgs aren’t the best thing. It takes a raw. Now, if you had a color checker card, like the one uhoh7 posted on top of this page, a raw from the stock A7 and one from the A7 Kolari-ut, both taken with the same light ( say sunny-daylight or tungsten bulbs, not fluorescents ) it would be very revealing. That is, if you can upload those raw files somewhere, so that we can download them, as Guy did with his Front Filter testing.
p.10 #15 · First Sighting: A7rii Thin filter in action
artur5 wrote:
Thanks for your feedback and offer of a color comparison. I really appreciate that.
As you said, for fine tuning the color output of a camera, jpgs aren’t the best thing. It takes a raw. Now, if you had a color checker card, like the one uhoh7 posted on top of this page, a raw from the stock A7 and one from the A7 Kolari-ut, both taken with the same light ( say sunny-daylight or tungsten bulbs, not fluorescents ) it would be very revealing. That is, if you can upload those raw files somewhere, so that we can download them, as Guy did with his Front Filter testing. ...Show more →
does dropbox work by you? i might have a color checker chart at work, will check.
p.10 #17 · First Sighting: A7rii Thin filter in action
no dice on finding a color chart at work, but i can take some photos with a broad range of color and something properly grey and upload the raws.
in the meantime it's sonnar test day!
the zm 50/1.5 definitely shows improvement in the midzone at infinity with the ultra thin sensor mod (you can see i did a poor job of framing for this comparison ): http://c1.staticflickr.com/3/2830/33408355706_8e8e0c6c13_o.jpg here is a link to the full sized comparison.
i think i would be willing to use the zm 50/1.5 for landscape at f/8 now in a pinch, which puts it back in the running for the position as my main 50mm.
tried to do the same test with my jupiter-9, but it doesn't reach infinity on the stock a7 i borrowed from nehemiah. it seems there is some truth to the rumors of flange distance variation across sony cameras. it hit infinity on my camera perfectly prior to the conversion.
i am pretty sure both sonnars show less spherical aberration at mid and close distance with the UT sensor conversion, but am yet to find a good test for this...
p.10 #18 · First Sighting: A7rii Thin filter in action
artur5 wrote:
Of course, jpgs aren’t raws (they’re already "cooked" ) but applying a photography filter to a jpg file, it works almost as modifying the color temperature of the shot. Basically the same thing that in film days. when we used CC filters on the lens to compensate for different light conditions. In this case, a used a weak cyan filter that rendered the snow quite neutral, but the greens of the foliage still looked a dull green-brown instead of shinny green like in the M9 sample.
Sorry, I've been swamped. For sure, Jpegs are worthless to play with color. You can PM me for any RAW on my flickr. It might take me a bit to get them in a DB for you.
Leica Lenses colors are dependent on the camera to considerable degree, though perhaps editing RAWs can bring things close. M9 vs M240 vs M10, there are differences, let alone stock sony, or various mods.
For anyone getting the UT I would suggest getting a color checker and making camera profiles in with the free adobe program. It's not hard at all and will export directly into LR. Petapixel has a good tutorial.
That's AWB, no adjustments. For sure there are special WB issues, however I'm increasingly thinking they can be dealt with as Derek says. I shot a play tonite with M9 A7.v2 and A7ut
At the moment even making a base profile as described in the petapixel article is hard here, as the sun is never overhead,and we have snow everywhere throwing huge blue cast, etc.
Nevertheless the UT is often doing pretty well to my eye: UT_CV35/1.2_f4ish by unoh7, on Flickr
In the long run, I would make myself a series of profiles, for different light, now that I see it's pretty easy and would be REALLY easy with practice. Like the tutorial guys says, WB is all over the place with most cameras, and that's why there are so many choices in body and in PP. I've been lazy, but this whole process of vetting the UT is boning me up a bit
p.10 #20 · First Sighting: A7rii Thin filter in action
uhoh7 wrote:
Sorry, I've been swamped. For sure, Jpegs are worthless to play with color. You can PM me for any RAW on my flickr. It might take me a bit to get them in a DB for you.
These things are WB nightmare, but it was equal nighmare though with various quirks to each body.
At the moment even making a base profile as described in the petapixel article is hard here, as the sun is never overhead,and we have snow everywhere throwing huge blue cast, etc.
You should relocate to sunny Arizona
I use a normalized D50 light source for making daylight profiles (small basic models are not expensive at all). For indoors light. a 3000K tungsten bulb is fine. As you say, outdoors light isn't consistent.
) but applying a photography filter to a jpg file, it works almost as modifying the color temperature of the shot. Basically the same thing that in film days. when we used CC filters on the lens to compensate for different light conditions. In this case, a used a weak cyan filter that rendered the snow quite neutral, but the greens of the foliage still looked a dull green-brown instead of shinny green like in the M9 sample. 
