Last call for color-target shots for inclusion in RawTherapee 5


(Morgan Hardwood) #1

Last call for D65+tungsten color-target shots for inclusion in RawTherapee 5. If you can supply raw shots of a color target for cameras other than the ones listed below, please do so ASAP. By doing so you confirm that you took the photos and that you can and do release them into the public domain (CC0 licence https://creativecommons.org/about/cc0 ).

Cameras we already have covered:

  • Canon EOS 6D
  • Canon PowerShot G7 X
  • Canon PowerShot S110
  • FUJIFILM X100T
  • FUJIFILM X-T1
  • NIKON D7200
  • PENTAX K10D
  • PENTAX K-5
  • PENTAX K-5 II
  • RICOH PENTAX K-3
  • SONY ILCE-7M2
  • SONY ILCE-6000

Shooting instructions:
http://rawpedia.rawtherapee.com/How_to_create_DCP_color_profiles#Shooting_the_color_target


(Mica) #2

Do you want the DCP as well?

I have Wolf Faust targets, if that’s OK.

The page also mentions google code, you may want to update that.


(srowed) #3

I don’t have an x-rite, but do have a Macbeth ColorChecker. It’s old but in good condition and has the same swatches as the x-rite. Will this work for you?

The sun will not be “high in the sky” even at noon here, given the 51 degrees N latitude. It’s also supposed to be cloudy for the next few days so I’m wondering how quickly you need the raw files.

I have a Nikon D800, D7100 and D3.


(Alex Mozheiko) #4

Have ready made color profile for Nikon J1, do you want it? :slight_smile: )
and about to make the same for sony a5000…are interested in those?


(Morgan Hardwood) #5

Macbeth is fine.

No, I only want the raw shots, not ready-made DCPs.

Yes, I’m interested in all camera models not already covered. Both daylight and tungsten shots.


(Morgan Hardwood) #6

Thank you, fixed.

I updated the whole article, please see it again - I added missing info and clarified some things.


(Mica) #7

Awesome, about to shoot my target with a D750 and Yi M1.


(Assaf Toledo) #8

I can add a profile of Olympus E-M10, but three questions:

  1. Regarding the StdA sample, I have the required tungsten bulb, but it is in a corridor lamp, attached to the wall, not to the ceiling. The width is only 60cm and the height is 120cm. Please advice if this is a workable setup and where it’s better to hold the color checker. .There will be a lot of reflection from the walls, but they are white.

  2. Regarding the D65 shot, unfortunately the days aren’t perfectly clear where I live. There is a lot of sun, but not like a perfectly clear sky. Is this good enough?

  3. My card is a CC passport, and I know that DCamProf has a special treatment for that. So should I take a photo with the two sides open (the 24 patches + additions), or only the 24 patches?

Thanks.


(Morgan Hardwood) #9

I had no issues with using my white-tiled bathroom or white-painted basement, so I guess your white wall should be fine too.

Should be fine. In writing the guide I stuck to the practice of shooting in the sun, but in my own experience shooting on an overcast day presented no problems. I want to confirm this with new measurements before I amend the guide.

Just the part which interests us, so that this part is the brightest part of your frame.

Send in the shots, we’ll measure and get back to you if there are issues.


(Assaf Toledo) #10

OK. I’m attaching a tungsten shot. Tomorrow I’ll try to capture some sun.

P1171562.ORF (14.2 MB)


(Morgan Hardwood) #11

@assaft I took a quick look generated a DCP - your shot looks fine, your chart is in great condition and the result works well:

There was room for a +0.9EV adjustment to get the D4 patch to L*=50 (patch 22 in this image http://xritephoto.com/ph_product_overview.aspx?ID=824&Action=Support&SupportID=5159 ).

Looking forward to the D65 shot.

Also, there is a cat lying on my mousepad.


(Mica) #12

Here are the Nikon D750 and Yi Mi. I’m not too smart and left my Nikon set at f10. Sorry. Both images are released under the Creative Commons Zero license, as requested.

https://filebin.net/07taq343j0sxsy3m


(Morgan Hardwood) #13

Thanks @paperdigits, I never processed one of these before. The YI M1 shot is just on the verge of clipping (green channel over A15). Could you make tungsten shots too?


(Mica) #14

I can look for a tungsten bulb, but I don’t own one at the moment.


(Assaf Toledo) #15

Here are two D65 shots, I’m not sure which one is better. The sun was coming through some thin clouds, in the second one more than the first.

P1171581.ORF (15.0 MB)

  • has a straighter angle but slightly harsher light

P1171595.ORF (14.8 MB)

  • tilted but slightly more diffused light

(Pata Kreyp) #16

Hello,

I have shot my X-Rite passport with my Nikon D800 this noon (as bright sun as possible in winter) but I have no more tungsten bulb. Is it useful that I provide my sunlight DCP?

P.S.
And a big “thank you!” for the astounding work on RawTherapee :smile:


(Morgan Hardwood) #17

@Pata_Kreyp yes please, the raw file not the DCP.


(Morgan Hardwood) #18

Making a DCP for the IT8 (IT 8.7/2-1993) using DCamProf turned out to be easy.

I got the reference values by downloading the R151101 file from http://www.colorreference.de/targets/

argyll-scanin -v -G 0.454545 -p -digIcrpn daylight.tif /usr/share/argyllcms/ref/it8.cht /tmp/R151101/R151101.txt daylight-diag.tif
./dcamprof make-profile -i D50 -C daylight.ti3 daylight.json
./dcamprof make-dcp -n "YI TECHNOLOGY M1" -d "YI TECHNOLOGY M1" -c "RawTherapee CC0" -t acr -o neutral -g none daylight.json "YI TECHNOLOGY M1.dcp"

Before:

After:

After with embedded tone curve:

Delta E was quite high, not surprisingly as these IT8 targets are known to be difficult to use outside of the studio, though as this is my first DCP made from one of them I don’t know what is normal for them.

$ argyll-scanin -v -G 0.454545 -p -digIcrpn daylight.tif /usr/share/argyllcms/ref/it8.cht /tmp/R151101/R151101.txt daylight-diag.tif
Input file 'daylight.tif': w=1560, h=1040, d = 3, bpp = 16
Data input file '/tmp/R151101/R151101.txt'
Data output file 'daylight.ti3'
Chart reference file '/usr/share/argyllcms/ref/it8.cht'
Creating diagnostic tiff file 'daylight-diag.tif'
About to allocate scanrd_ object
Verbosity = 2, flags = 0x62af9
About to read input tiff file and discover groups
adivval = 1.000000
About to calculate edge lines
442 useful edges out of 1143
About to calculate perspective correction
Perspective correction factors = 0.000004 -0.000016 780.000000 520.000000
About to calculate rotation
Mean angle = -0.304448
Standard deviation = 3.864164
Robust mean angle = 0.087368 from 428 lines
About to calculate feature information
About to read reference feature information
Read of chart reference file succeeded
About to match features
Checking xx
Checking yy
Checking xy
Checking yx
Checking xix
Checking yiy
Checking xiy
Checking yix
Axis matches for each possible orientation:
  0: xx  = 0.714416, yy  = 0.631872, xx.sc  = 0.409207, yy.sc  = 0.409131
 90: xiy = 0.283689, yx  = 0.524188, xiy.sc = 0.822574, yx.sc  = 0.406687
180: xix = 0.565962, yiy = 0.511631, xix.sc = 0.409500, yiy.sc = 0.408823
270: xy  = 0.310177, yix = 0.439876, xy.sc  = 0.821863, yix.sc = 0.407303
r0 = 0.953579, r90 = 0.294682, r180 = 0.761680, r270 = 0.266743
There are 2 candidate rotations:
cc = 0.953579, irot = 0.087368, xoff = 22.089452, yoff = 12.751905, xscale = 2.443749, yscale = 2.444205
cc = 0.761680, irot = 180.087368, xoff = -1531.860554, yoff = -1018.097609, xscale = 2.442004, yscale = 2.446049
About to compute match transform for rotation 0.087368 deg.
About to setup value scanrdg boxes
About to read raster values
About to compute expected value correlation
About to compute match transform for rotation 180.087368 deg.
About to setup value scanrdg boxes
About to read raster values
About to compute expected value correlation
Expected value distance values are:
0, rot 0.087368: 2216.488313
1, rot 180.087368: 3539.747304
Chosen rotation 0.087368 deg. as best
About to compute final match transform
Improve match
About to setup value scanrdg boxes
About to read raster values
About to write diag file
Writing output values to file 'daylight.ti3'
dcamprof-1.0.4 $ ./dcamprof make-profile -i D50 -C daylight.ti3 daylight.json                                                                        
Reading target...
Generating values for the calibration illuminant D50...
Warning: large dynamic range difference detected. Likely glare issue.
Camera G on darkest patch(es) is 296.0% lighter compared to observer Y.
  Y dynamic range is 7.52 stops, G dynamic range is 5.54 stops, difference
  1.99 stops. A small difference is normal, while a large indicates that there
  is glare.
Warning: auto-selected neutral patch (A16) is not the lightest, as the
  lightest patch is considerably off-white. That is if you later use the target
  for white balancing you should use the indicated patch instead of the
  lightest.
The most neutral patch (A16) differs 1.04 DE from actual neutral,
  transforming target reference XYZ values to match, using CAT02.
Automatic LUT relaxation weights assigned.
Making camera profile...
Finding a camera raw RGB to CIE XYZ matrix for calibration illuminant D50...
Whitest patch in target (A16) differs DE 1.04 from calibration illuminant,
  close enough to calculate whitepoint preservation.
Inverting to get ColorMatrix:
  {
    "ColorMatrix1": [
      [  0.537094, -0.095588, -0.022356 ],
      [ -0.374916,  1.103322,  0.316279 ],
      [ -0.025329,  0.115141,  0.663253 ]
    ]
  }
Matrix patch match average DE 2.95, DE LCh 1.19 1.58 1.74
                    median DE 1.99, DE LCh 0.93 0.87 1.18
                       p90 DE 6.76, DE LCh 2.30 4.75 4.61
                       max DE 10.15, DE LCh 6.42 8.50 7.57
ColorMatrix optimal white balance for target: 0.402706,1,0.636244 (m2.4832,1,1.57172)
Finding a camera raw RGB to CIE XYZ matrix for connection space illuminant D50...
  {
    "LUTMatrix1": [
      [  0.664613,  0.137690,  0.161915 ],
      [  0.274660,  0.677869,  0.047472 ],
      [  0.000096,  0.017238,  0.807867 ]
    ]
  }
LUTMatrix optimal white balance for target: 0.402974,1,0.630545 (m2.48155,1,1.58593)
Matrix patch match average DE 6.42, DE LCh 2.33 5.05 2.85
                    median DE 6.32, DE LCh 1.70 4.68 2.16
                       p90 DE 10.62, DE LCh 5.06 8.87 5.99
                       max DE 17.52, DE LCh 11.34 13.44 16.19
Finding a camera raw RGB to CIE XYZ matrix for connection space illuminant D50...
  Y row limit set to -0.2.
Applying white-balance to get ForwardMatrix:
  {
    "ForwardMatrix1": [
      [  0.822741,  0.000760,  0.140718 ],
      [  0.307258,  0.806480, -0.113738 ],
      [  0.029543, -0.258563,  1.054222 ]
    ]
  }
ForwardMatrix optimal white balance for target: 0.402974,1,0.630545 (m2.48155,1,1.58593)
Matrix patch match average DE 3.72, DE LCh 1.70 2.09 1.95
                    median DE 3.25, DE LCh 1.27 1.60 1.44
                       p90 DE 7.38, DE LCh 3.75 4.93 4.59
                       max DE 10.26, DE LCh 6.71 8.59 7.43
Making 2.5D chromaticity-addressed lookup table for XYZ correction...
53.47% of the patches was put in a chromaticity group due to nearby neighbor.
  26.39% of the patches was removed due to being nearby the whitepoint.
  Largest chromaticity group contains 16 patches. Patch count reduced from
  288 to 58. Note that patch matching cannot reach 100% when chromaticity
  groups are formed, as the LUT matches the average within a group.
Lightness axis is disabled. Since lightness affects chroma, the LUT chroma
  control points are recalculated to better match the uncorrected lightness.
  A residual error of up to about 0.2 DE is expected.
Relaxing LUT stretch with up to 4.00 DE. Iterating over 58 patches...
  Lightness correction is disabled.
.............Average DE for the 58 tested patches increased to 3.31 after LUT relax.
  18.97% could do without LUT correction.
Native LUT patch match average DE 3.32, DE LCh 1.72 1.81 1.49
                        median DE 2.77, DE LCh 1.22 1.26 0.98
                           p90 DE 6.66, DE LCh 3.79 4.35 3.84
                           max DE 10.06, DE LCh 7.49 8.47 7.27
5 worst patches for Overall DE:
  J22 RGB 0.012 0.039 0.023 XYZref 0.015 0.024 0.011 XYZcam 0.030 0.037 0.029 sRGB #09321B #253934 DE 10.06 DE LCh +3.46 -8.47 +4.17 (dark green)
  F01 RGB 0.011 0.031 0.021 XYZref 0.017 0.021 0.016 XYZcam 0.027 0.029 0.029 sRGB #192B25 #2A3035 DE 9.88 DE LCh +2.52 -6.65 +6.87 (dark grayish green)
  K21 RGB 0.012 0.036 0.026 XYZref 0.015 0.021 0.018 XYZcam 0.030 0.034 0.035 sRGB #0F2C28 #27353B DE 9.75 DE LCh +3.75 -6.23 +6.50 (dark grayish cyan)
  F02 RGB 0.009 0.032 0.020 XYZref 0.013 0.021 0.013 XYZcam 0.023 0.029 0.025 sRGB #012F1F #1E3331 DE 9.49 DE LCh +2.28 -8.11 +4.36 (dark grayish green)
 GS23 RGB 0.008 0.021 0.016 XYZref 0.004 0.005 0.006 XYZcam 0.020 0.020 0.023 sRGB #061215 #24262F DE 9.42 DE LCh +6.71 +0.65 +6.57 (black)
5 worst patches for Lightness DE:
  L19 RGB 0.017 0.066 0.136 XYZref 0.048 0.019 0.233 XYZcam 0.083 0.047 0.295 sRGB #0032A0 #002EAB DE 8.04 DE LCh +7.49 -2.41 -2.70 (very dark vivid purple-blue)
 GS23 RGB 0.008 0.021 0.016 XYZref 0.004 0.005 0.006 XYZcam 0.020 0.020 0.023 sRGB #061215 #24262F DE 9.42 DE LCh +6.71 +0.65 +6.57 (black)
  K19 RGB 0.025 0.093 0.173 XYZref 0.067 0.036 0.287 XYZcam 0.102 0.067 0.329 sRGB #003DAE #133CB2 DE 7.04 DE LCh +6.63 -2.51 -1.82 (dark vivid purple-blue)
  J04 RGB 0.017 0.073 0.145 XYZref 0.048 0.023 0.241 XYZcam 0.084 0.048 0.331 sRGB #003EA4 #003FB7 DE 6.60 DE LCh +6.54 -0.49 -1.00 (dark vivid purple-blue)
  J19 RGB 0.036 0.128 0.211 XYZref 0.092 0.060 0.337 XYZcam 0.125 0.094 0.342 sRGB #0045B8 #2F4EB5 DE 6.70 DE LCh +6.03 -3.48 -1.63 (dark strong purple-blue)
5 worst patches for Chroma DE:
  J22 RGB 0.012 0.039 0.023 XYZref 0.015 0.024 0.011 XYZcam 0.030 0.037 0.029 sRGB #09321B #253934 DE 10.06 DE LCh +3.46 -8.47 +4.17 (dark green)
  F02 RGB 0.009 0.032 0.020 XYZref 0.013 0.021 0.013 XYZcam 0.023 0.029 0.025 sRGB #012F1F #1E3331 DE 9.49 DE LCh +2.28 -8.11 +4.36 (dark grayish green)
  K20 RGB 0.017 0.050 0.030 XYZref 0.027 0.036 0.022 XYZcam 0.041 0.048 0.038 sRGB #223A2C #33403C DE 7.68 DE LCh +2.70 -6.76 +2.47 (dark grayish green)
  F01 RGB 0.011 0.031 0.021 XYZref 0.017 0.021 0.016 XYZcam 0.027 0.029 0.029 sRGB #192B25 #2A3035 DE 9.88 DE LCh +2.52 -6.65 +6.87 (dark grayish green)
  E01 RGB 0.024 0.057 0.032 XYZref 0.046 0.052 0.028 XYZcam 0.055 0.058 0.041 sRGB #3D4231 #45433D DE 7.51 DE LCh +1.19 -6.52 -3.52 (dark grayish yellow-green)
5 worst patches for Hue DE:
  D01 RGB 0.025 0.054 0.031 XYZref 0.049 0.051 0.027 XYZcam 0.059 0.057 0.040 sRGB #443F30 #4C413C DE 7.67 DE LCh +1.29 -2.05 -7.27 (dark grayish yellow)
  C20 RGB 0.016 0.028 0.019 XYZref 0.020 0.017 0.008 XYZcam 0.037 0.031 0.027 sRGB #321D18 #402B32 DE 8.30 DE LCh +4.54 -0.08 +6.95 (very dark grayish red)
  F01 RGB 0.011 0.031 0.021 XYZref 0.017 0.021 0.016 XYZcam 0.027 0.029 0.029 sRGB #192B25 #2A3035 DE 9.88 DE LCh +2.52 -6.65 +6.87 (dark grayish green)
 GS23 RGB 0.008 0.021 0.016 XYZref 0.004 0.005 0.006 XYZcam 0.020 0.020 0.023 sRGB #061215 #24262F DE 9.42 DE LCh +6.71 +0.65 +6.57 (black)
  K21 RGB 0.012 0.036 0.026 XYZref 0.015 0.021 0.018 XYZcam 0.030 0.034 0.035 sRGB #0F2C28 #27353B DE 9.75 DE LCh +3.75 -6.23 +6.50 (dark grayish cyan)
5 best patches for Overall DE:
  A16 RGB 0.386 0.959 0.605 XYZref 0.925 0.959 0.791 XYZcam 0.925 0.959 0.791 sRGB #FAFAFA #FAFAFA DE 0.00 DE LCh +0.00 +0.00 +0.00 (white)
  C06 RGB 0.146 0.216 0.077 XYZref 0.313 0.273 0.069 XYZcam 0.313 0.273 0.069 sRGB #C18043 #C18043 DE 0.09 DE LCh +0.00 +0.03 +0.08 (orange)
  G20 RGB 0.199 0.371 0.166 XYZref 0.435 0.418 0.176 XYZcam 0.438 0.421 0.180 sRGB #D0A577 #D0A578 DE 0.37 DE LCh +0.13 -0.22 -0.27 (orange)
  B17 RGB 0.359 0.800 0.490 XYZref 0.835 0.822 0.630 XYZcam 0.841 0.830 0.635 sRGB #FDE4E1 #FEE5E2 DE 0.39 DE LCh +0.22 -0.21 +0.24 (whitish pink)
  F22 RGB 0.210 0.374 0.172 XYZref 0.465 0.433 0.189 XYZcam 0.464 0.431 0.191 sRGB #D9A57B #D8A57C DE 0.42 DE LCh -0.11 -0.23 -0.33 (light orange)
Writing output to "daylight.json"...
Complete!
dcamprof-1.0.4 $ ./dcamprof make-dcp -n "YI TECHNOLOGY M1" -d "YI TECHNOLOGY M1" -c "RawTherapee CC0" -t acr -o neutral -g none daylight.json "YI TECHNOLOGY M1.dcp"
Generating 2.5D HueSatMap with 90x30 = 2700 entries...done!
The tone curve's contrast value is 1.30 (=> auto chroma scaling value 1.121)
Generating 3D LookTable with 90x30x30 = 81000 entries for the neutral tone reproduction operator...
  0%..10%..20%..30%..40%..50%..60%..70%..80%..90%..100%
Writing output to "YI TECHNOLOGY M1.dcp"...
Complete!

@paperdigits how about that tungsten shot?


(Philip B) #19

Ill post Nikon D600 tomorrow


(Pata Kreyp) #20

Here is the RAW file of my shot of X-Rite Passport with Nikon D800 ISO100 f/8.
_DSC8004.NEF (46.6 MB)