Pages: 1 2 [3] 4 5 ... 11   Go Down

Author Topic: New profile making software: Lumariver Profile Designer  (Read 16638 times)

Alexey.Danilchenko

  • Full Member
  • ***
  • Offline Offline
  • Posts: 145
Re: New profile making software: Lumariver Profile Designer
« Reply #40 on: May 28, 2017, 06:52:00 AM »

SSF data is not supported at this time, as there are extremely few users, the problem being that it's very expensive or advanced DYI/ebay project to actually shoot SSF.

I am sorry but that is wrong on both counts. For a good quality profile with the card you need spectrophotometer to measure your particular instance of the card (using reference values is a dead approach - there are substantial variations and even the same card will vary over time). The cost of monochromator approach is actually comparable with a new CC SG card - it is a bit higher but gives you non comparable advantage of measuring sensor once as opposed faffing about with shoting targets correctly under different lighting and dealing with all that target shooting involves. Add to that the price of spectrophotometer for the target measuring and the price for spectral sensitivity curves approach will look more than favourable. Not to say it will produce more consistent results with less efforts.

The whole point of my and Iliah project was to show how easy it is and give away free tools to use to make it even easier to construct.

I also would point out that there are commercial products out there that allegedly allow you to take spectral curves from the cameras as well so it this is not limited just to DIYers like me.
« Last Edit: May 30, 2017, 04:36:19 AM by Alexey.Danilchenko »
Logged

torger

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 3243
Re: New profile making software: Lumariver Profile Designer
« Reply #41 on: May 28, 2017, 11:17:12 AM »

I am sotty but that is wrong on both counts. The cost us actually comparable with a new CC SG card - it is a bit higher but gives you non comparable advantage of measuring sensor once as opposed faffing about with shoting targets correctly under different lighting and dealing with all that target shooting involves.

The whole point of my and Iliah project was to show how easy it is and give away free tools to use to make it even easier to construct

Is there a link to this project? Could be nice to have in this thread.

Anyway, with the upcoming patch release you will be able to load custom targets with raw values, so you can load CGATS-like files with raw values prepared in DCamProf or other software, with or without SSF. There will be no direct support for SSF in LRPD at this time though, that is you need to prepare a CGATS file in some other software. It's not unlikely that it will appear eventually though, just because I can. I have nothing against SSF, but I just don't see many using it. If it requires an ebay purchase of rare equipment it's not going to be any volumes.
Logged

torger

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 3243
Re: New profile making software: Lumariver Profile Designer
« Reply #42 on: May 28, 2017, 01:56:48 PM »

do you override this with the control in GUI or the value dialied in GUI is added to the value from DNG ?

In the GUI it's the baseline exposure offset that is set, and thus added to the value in the DNG.

For the custom raw value files I've now decided to disregard from the baseline exposure in the DNG though, so the values provided are the values used regardless what the DNG baseline exposure in the image is. It only matters for 3D LUTs in any case...
Logged

scyth

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 738
Re: New profile making software: Lumariver Profile Designer
« Reply #43 on: May 28, 2017, 02:23:29 PM »

Is there a link to this project? Could be nice to have in this thread.

https://www.dpreview.com/forums/thread/4144719

but you need to hunt down cheap used integrating sphere and monochromator @ ebay... otherwise they cost a lot... some people DIY integrating spheres, coating them with ~barium sulfate inside
Logged

torger

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 3243
Re: New profile making software: Lumariver Profile Designer
« Reply #44 on: May 29, 2017, 01:52:50 AM »

https://www.dpreview.com/forums/thread/4144719

but you need to hunt down cheap used integrating sphere and monochromator @ ebay... otherwise they cost a lot... some people DIY integrating spheres, coating them with ~barium sulfate inside

My current view on SSF is that it's good for research, testing targets, how large impact different color filter shapes has etc, but that it seems to often suffer from quite large measurement errors. To the profiler SSF always looks like "perfect", there's never glare or uneven light or those kind of non-linear measurement errors. When shooting targets the profiler gets to see the measurement errors, with SSF there are no errors as seen by the profiler -- but is the SSF provided actually matching the camera's real SSF or is it differing substantially due to measurement error when making the SSF? I have no gear to measure SSF myself, so I have just tested the stuff found on the net from some academic projects, and for my old 5DMk2 I say that shooting targets leads to a more accurate profile than using those SSF, even if you have moderate measurement errors on the target. But the theory is if you have proper accurate SSF, they should be great... I just don't know how easy it is to get accurate SSF measurements...
« Last Edit: May 29, 2017, 01:57:43 AM by torger »
Logged

scyth

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 738
Re: New profile making software: Lumariver Profile Designer
« Reply #45 on: May 29, 2017, 09:22:12 AM »

I just don't know how easy it is to get accurate SSF measurements...

neither is to get a proper shot of a target ...
Logged

torger

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 3243
Re: New profile making software: Lumariver Profile Designer
« Reply #46 on: May 29, 2017, 10:00:16 AM »

neither is to get a proper shot of a target ...

Indeed, but in that case you get, with some practice, a good indicator of the quality. You can generally see on the matrix result if the shot suffers from large measurement errors, huge lightness errors along the neutral axis especially on the black patch, or over-saturation = glare. With a badly measured SSF the profiler still sees a "perfect camera" (but the wrong camera) so you get no indication of measurement error in the matrix or LUT calculations, you just have your eyes to decide if the final rendered image seems to produce sane result.

Target measurement errors usually leads to some saturation errors, hues are still correct. Lightness is generally not compensated for, unless you do reproduction, so the typical workflow is very robust against that. As cameras make a pretty good matrix-only match these days, the result gravitate to something linear and sane in the relax stage, further enhancing robustness.

SSF measurement errors however, I think has a tendency to also produce hue matching errors (which I think is a worse problem), but maybe I'm wrong? Although I have made many theoretical experiments with SSF I actually don't have the experience from making SSF measurements, so I stay humble about the subject.
Logged

torger

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 3243
Re: New profile making software: Lumariver Profile Designer
« Reply #47 on: May 29, 2017, 12:33:12 PM »

Attention DxO Optics Pro users:

I've noticed that LRPD gets fooled by that the linear RAW tiffs still have an ICC profile embedded and then messes up the transfer function. I will fix that to the next patch release. However you can work around it currently by removing the ICC profile from the linear TIFF before loading it into LRPD, using exiftool like this:

exiftool -icc_profile:all= linear.tif

and then load it into LRPD and make your profile.

I've also noted that it seems that the "File -> Export Image for ICC Profile" option is only available in the Windows version of DxO Optics, don't know why. The mac version can load profiles still, but not export for profiling, so for that it seems like you need a Windows box, or complain to DxO so they fix the mac version.
Logged

daicehawk

  • Newbie
  • *
  • Offline Offline
  • Posts: 44
Re: New profile making software: Lumariver Profile Designer
« Reply #48 on: May 29, 2017, 03:21:40 PM »

Crashes when selecting all the following:
1) ICC
2) Neutral: from white patch
3)LUT Optimization: None
4) Curve:Linear
5) TRO:Neutral
6) Base Look:None
It seems that whatever the options are selecting ICC \Repro crashes under Win7 64.
Logged

daicehawk

  • Newbie
  • *
  • Offline Offline
  • Posts: 44
Re: New profile making software: Lumariver Profile Designer
« Reply #49 on: May 29, 2017, 03:34:28 PM »

OK, the demo does not allow to export profile, but how can I try applying it?
Logged

torger

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 3243
Re: New profile making software: Lumariver Profile Designer
« Reply #50 on: May 29, 2017, 03:36:54 PM »

OK, the demo does not allow to export profile, but how can I try applying it?

You can use the Profile Comparison tab and load custom non-target images there.
Logged

daicehawk

  • Newbie
  • *
  • Offline Offline
  • Posts: 44
Re: New profile making software: Lumariver Profile Designer
« Reply #51 on: May 29, 2017, 03:43:13 PM »

You can use the Profile Comparison tab and load custom non-target images there.
Using a DataColor chekr24 the skin tone is too redas in common profiling. Could you be so kind and quote your reference data for SpderCC24 or point to the source of it? I might send you mine if you need it.
Logged

torger

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 3243
Re: New profile making software: Lumariver Profile Designer
« Reply #52 on: May 29, 2017, 03:45:37 PM »

Crashes when selecting all the following:
...

Tried the same with v1.0.1 on Win without issues, but it could be related to target or target image rather than those settings. You could use File -> Archive one step away from crashing and PM me a link to have a look. Or sit around and wait for v1.0.2 which I'm working on currently, it fixes some crash things.
Logged

daicehawk

  • Newbie
  • *
  • Offline Offline
  • Posts: 44
Re: New profile making software: Lumariver Profile Designer
« Reply #53 on: May 29, 2017, 03:48:08 PM »

Tried the same with v1.0.1 on Win without issues, but it could be related to target or target image rather than those settings. You could use File -> Archive one step away from crashing and PM me a link to have a look. Or sit around and wait for v1.0.2 which I'm working on currently, it fixes some crash things.
BTW you should provide a WB colorpickerfor test images when testing a profile since the embedded WB in DNG or RAW might be wrong.
Logged

torger

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 3243
Re: New profile making software: Lumariver Profile Designer
« Reply #54 on: May 29, 2017, 03:50:27 PM »

Using a DataColor chekr24 the skin tone is too redas in common profiling. Could you be so kind and quote your reference data for SpderCC24 or point to the source of it? I might send you mine if you need it.

It's data coming from scyth. You can manually adjust the skintone patch in the customize colors dialog if you think the reference is wrong. Otherwise if it's just very slightly too red it can be due to approximations in the colorimetric step, which you can control in the optimization tab. Or if that's correct it may simply be that "too red" is caused by a mismatch between the tone reproduction operator and your taste, if so you could try for example the RGB-Lum or a mixed variant which is less hue accurate and makes skintones more yellow.
Logged

daicehawk

  • Newbie
  • *
  • Offline Offline
  • Posts: 44
Re: New profile making software: Lumariver Profile Designer
« Reply #55 on: May 29, 2017, 03:51:12 PM »

So for ICC profiles there is only TIFF shots support?
Logged

torger

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 3243
Re: New profile making software: Lumariver Profile Designer
« Reply #56 on: May 29, 2017, 03:51:29 PM »

BTW you should provide a WB colorpickerfor test images when testing a profile since the embedded WB in DNG or RAW might be wrong.

There is one...
Logged

daicehawk

  • Newbie
  • *
  • Offline Offline
  • Posts: 44
Re: New profile making software: Lumariver Profile Designer
« Reply #57 on: May 29, 2017, 03:55:40 PM »

It's data coming from scyth. You can manually adjust the skintone patch in the customize colors dialog if you think the reference is wrong. Otherwise if it's just very slightly too red it can be due to approximations in the colorimetric step, which you can control in the optimization tab. Or if that's correct it may simply be that "too red" is caused by a mismatch between the tone reproduction operator and your taste, if so you could try for example the RGB-Lum or a mixed variant which is less hue accurate and makes skintones more yellow.
The customize color tab's contorl Hue provides only in effect a Tint control - a magenta\pink axis/
Logged

torger

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 3243
Re: New profile making software: Lumariver Profile Designer
« Reply #58 on: May 29, 2017, 03:56:01 PM »

So for ICC profiles there is only TIFF shots support?

Yes, you export to TIFF for profiling from the raw converter. That TIFF can be pretty much a straight dump of the raw data, it depends on how the raw converter does it. In any case you need some sort of profiling export from the raw converter, as to make an ICC profile the profile maker must know which pre-processing is applied before the profile, it's not standardized as for ICC.
Logged

torger

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 3243
Re: New profile making software: Lumariver Profile Designer
« Reply #59 on: May 29, 2017, 03:59:47 PM »

The customize color tab's contorl Hue provides only in effect a Tint control - a magenta\pink axis/

It's all three axes (the "tint" is the hue axis), you can make modification in a gamut sphere around the anchor color, just not that many DE. If the patch is several DEs off then there is something really wrong though.
Logged
Pages: 1 2 [3] 4 5 ... 11   Go Up