Pages: [1]   Go Down

Author Topic: WB for every lens when profiling the camera/lens?  (Read 1527 times)

Hening Bettermann

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 853
    • landshape.net
WB for every lens when profiling the camera/lens?
« on: May 28, 2017, 05:28:15 PM »

Hi!

When I have  shot the X-Rite Passport target for camera profiling, I have used the off-white page of the target to record the white balance of the Solux lamp, run at 15 Volts. This was then stored as a custom WB in the camera. Now I want to make profiles for every single lens I have. Do I need to shoot and store a custom WB for every lens? Or is it enough with 1 for the camera (a7r2)? I can't look through this :-( I will use DCamProf or Lumariver PD for creating the profile.

Background: I have hitherto believed, that the difference between lenses was neglegible compared to the difference between cameras. Wrong again. Recently, I compared the CY Sonnar 135 to the Olympus OM 135 for sharpness, using Bart's resolution target. Compared to the physical target, the Zeiss looked red, and the Oly looked green.

Thank you for your help!

torger

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 3253
Re: WB for every lens when profiling the camera/lens?
« Reply #1 on: May 29, 2017, 03:36:50 AM »

Hi!

When I have  shot the X-Rite Passport target for camera profiling, I have used the off-white page of the target to record the white balance of the Solux lamp, run at 15 Volts. This was then stored as a custom WB in the camera. Now I want to make profiles for every single lens I have. Do I need to shoot and store a custom WB for every lens? Or is it enough with 1 for the camera (a7r2)? I can't look through this :-( I will use DCamProf or Lumariver PD for creating the profile.

Background: I have hitherto believed, that the difference between lenses was neglegible compared to the difference between cameras. Wrong again. Recently, I compared the CY Sonnar 135 to the Olympus OM 135 for sharpness, using Bart's resolution target. Compared to the physical target, the Zeiss looked red, and the Oly looked green.

Thank you for your help!

Are you doing profiles for reproduction or general-purpose? In reproduction you have controlled environment and the target setup is generally the same setup as when shooting the actual goods. With a general-purpose profile the light used when shooting the target is often not the same as when shooting real scenes, unless it's flash photography, but even with flash the temperature vary a bit shot to shot. I'm just wondering that maybe it's overkill, as white balance will be adjusted to taste anyway? Or do you intend to use a fixed white balance? If so and you really want to match the lenses I guess a custom wb for each lens would be the way to go...

Regarding the profilers mentioned they don't need the white balance, they can use the white patch or the reference file as reference white rather than the white balance in the image. If making an ICC profile it's desirable that the WB is close to ideal for technical reasons, but a small variation is no problem.
Logged

daicehawk

  • Jr. Member
  • **
  • Offline Offline
  • Posts: 66
Re: WB for every lens when profiling the camera/lens?
« Reply #2 on: May 29, 2017, 05:49:11 AM »

Hi!

When I have  shot the X-Rite Passport target for camera profiling, I have used the off-white page of the target to record the white balance of the Solux lamp, run at 15 Volts. This was then stored as a custom WB in the camera. Now I want to make profiles for every single lens I have. Do I need to shoot and store a custom WB for every lens? Or is it enough with 1 for the camera (a7r2)? I can't look through this :-( I will use DCamProf or Lumariver PD for creating the profile.

Background: I have hitherto believed, that the difference between lenses was neglegible compared to the difference between cameras. Wrong again. Recently, I compared the CY Sonnar 135 to the Olympus OM 135 for sharpness, using Bart's resolution target. Compared to the physical target, the Zeiss looked red, and the Oly looked green.

Thank you for your help!
Lenses may be non-neutral and give color-casts (I have experienced a yellow-green one). Basically you want to create a profile for each lens that leaves the gray axis gray when white-balanced on the neutral patch. More staurated colors are usually not a problem.   
The lens color-cast may cause another problem when a canned camera profile was made for a neutral lens: when you white-balance a shot made with green-yellow, this causes the pink-magenta overall shift for the whole shot, which is remarkable particularly on skintones. So you should take care of this when profiling.   
« Last Edit: May 29, 2017, 05:53:40 AM by daicehawk »
Logged

Hening Bettermann

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 853
    • landshape.net
Re: WB for every lens when profiling the camera/lens?
« Reply #3 on: May 29, 2017, 02:16:33 PM »

Hi Anders, thank you for chiming in.

My profiles are for general purpose, landscape.

> With a general-purpose profile the light used when shooting the target is often not the same as when shooting real scenes

No. But the Sonnar and the Oly 135 were shot with the same Solux light, so they should look the same. - I need ICC profiles, because Helicon doesn't handle dcp's.

I do not adjust WB to taste visually for every image. The camera is set to 5000 K, which fits most of the time. If not, I sometimes use the sky as visual reference. Or I adjust WB to what I believe, based on sky or time of day. Regardless, 2 lenses shot under the same light should give the same result, IMHO.

So if I understand you right, I have to shoot the target with the Solux run at 15 V, which gives close to 5000K; and make a profile for every lens. So I will.

Hening Bettermann

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 853
    • landshape.net
Re: WB for every lens when profiling the camera/lens?
« Reply #4 on: May 29, 2017, 02:22:00 PM »

Hi daicehawk, thank you for your reply.

>Basically you want to create a profile for each lens that leaves the gray axis gray when white-balanced on the neutral patch.

Yes this seems to be the answer.

Good light!

torger

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 3253
Re: WB for every lens when profiling the camera/lens?
« Reply #5 on: June 04, 2017, 05:37:06 AM »

If you are using polarizers, it could be a good idea to make profiles with polarizer on too. It can have a quite strong effect on colors. Usually people are satisfied with just adjusting white balance to taste to compensate for the errors and it's okay, but if you're a perfectionist, and if you make a profile per lens you are ;), making a profile with the polarizer on and off would be a natural thing to do.
Logged

Hening Bettermann

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 853
    • landshape.net
Re: WB for every lens when profiling the camera/lens?
« Reply #6 on: June 04, 2017, 03:04:08 PM »

Hi Anders,
a good reminder. Also, one would actually need a profile for every angle of the polarizer, or? - I missed (lacked) a polarizer once in the 1980s for some shots on a lake in northern Sweden with nice rock formations under water, bought it after that and have never used it.
Good light!

Ernst Dinkla

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 3519
Re: WB for every lens when profiling the camera/lens?
« Reply #7 on: June 05, 2017, 04:17:23 AM »

With a dozen non native lenses on a Sony Axxxx I was considering to measure the spectral plots of the lenses just to have an idea of their color transmission. An old EOS film body, D50 approx diffuse flash at the film gate side, 3D printed adapter for the  x-rite i1Pro to fit on the lens front. Which raised the question whether the measurements could be used to alter a camera profile. I understand that making profiles per lens is probably easier but more often I discovered that spectral plots tell me more what actually is the cause of color reproduction differences.


Met vriendelijke groet, Ernst

http://www.pigment-print.com/spectralplots/spectrumviz_1.htm
March 2017 update, 750+ inkjet media white spectral plots
Logged

Hening Bettermann

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 853
    • landshape.net
Re: WB for every lens when profiling the camera/lens?
« Reply #8 on: June 05, 2017, 06:16:10 AM »

Hi Ernst,
indeed profiling the lenses sounds easier than making spectral plots, which is out of my reach. - What were your experiences with your analysis?

Ernst Dinkla

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 3519
Re: WB for every lens when profiling the camera/lens?
« Reply #9 on: June 05, 2017, 06:39:08 AM »

Hi Ernst,
indeed profiling the lenses sounds easier than making spectral plots, which is out of my reach. - What were your experiences with your analysis?

There is the word "considering" in my message :-) Have to finish my 3D printed masked lens hoods first, simulated the masks needed by projecting the film gate of an obsolete camera through the mounted infinity set lenses to millimeter draft film at the possible mask position. Odd mask shapes sometimes, add some opening tolerance though. Also still have to fix one lens infinity setting with my recently made collimator.

If you mean what I have seen in other spectral plots then it was in the inkjet paper white spectral plots. In spectral plots of viewing lights, calibration white/grey cards/stuff, copy stand lamps, the dichroic filters in the enlarger head I converted to a slide duplicator. The LED lamps that replaced the halogens + dichroic filters of that head. Not totally replaced and still searching for an affordable, efficient, 540NM green LED though.

Met vriendelijke groet, Ernst

http://www.pigment-print.com/spectralplots/spectrumviz_1.htm
March 2017 update, 750+ inkjet media white spectral plots
Logged

xpatUSA

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 375
    • Blog
Re: WB for every lens when profiling the camera/lens?
« Reply #10 on: June 10, 2017, 04:33:52 PM »

Hi!

When I have  shot the X-Rite Passport target for camera profiling, I have used the off-white page of the target to record the white balance of the Solux lamp, run at 15 Volts. This was then stored as a custom WB in the camera. Now I want to make profiles for every single lens I have. Do I need to shoot and store a custom WB for every lens? Or is it enough with 1 for the camera (a7r2)? I can't look through this :-( I will use DCamProf or Lumariver PD for creating the profile.

Background: I have hitherto believed, that the difference between lenses was neglegible compared to the difference between cameras. Wrong again. Recently, I compared the CY Sonnar 135 to the Olympus OM 135 for sharpness, using Bart's resolution target. Compared to the physical target, the Zeiss looked red, and the Oly looked green.

Thank you for your help!

Always good to run nominal 12V equipment at 15V . .

Kind of like running a nominal 120V halogen lamp at 150V ;)
Logged
best regards,

Ted

Hening Bettermann

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 853
    • landshape.net
Re: WB for every lens when profiling the camera/lens?
« Reply #11 on: June 10, 2017, 06:01:42 PM »

Hi Ted,
well this is what brings the lamp up to 5000 K. It doesn't have to run so long...

Ethan Hansen

  • Jr. Member
  • **
  • Offline Offline
  • Posts: 79
    • Dry Creek Photo
Re: WB for every lens when profiling the camera/lens?
« Reply #12 on: June 10, 2017, 07:20:56 PM »

Not totally replaced and still searching for an affordable, efficient, 540NM green LED though.

If your SMT soldering patience is up to the task, Osram makes good greens.

Ernst Dinkla

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 3519
Re: WB for every lens when profiling the camera/lens?
« Reply #13 on: June 11, 2017, 04:23:16 PM »

If your SMT soldering patience is up to the task, Osram makes good greens.

Ethan, I know and Lumileds (ex Philips) has them too. They published the research for them some years back, independent from one another.  But the SMT soldering is another matter.

http://nl.mouser.com/Optoelectronics/LED-Lighting/LED-Emitters/High-Power-LEDs-Single-Color/_/N-8usfn?P=1z0z7ptZ1yw8xrs

Met vriendelijke groet, Ernst

http://www.pigment-print.com/spectralplots/spectrumviz_1.htm
March 2017 update, 750+ inkjet media white spectral plots
Logged
Pages: [1]   Go Up