UHD BD to BD?

[Beber]

It's the orange that bothers me, it just looks wrong to my eyes.

That's also a thing I don't like with madVR. It tends to turn skintones into a weird shade of orange/pink, like salmon, lobster, you get the idea.

[Doctor M]

Looking through ripbot's plugins it looks like it's using something called DGTonemap. It's an avisynth+ plugin

That's a huge piece of information.  DG is Donald Graft AKA Neuron2, the maker of DGIndex, DGDecNV, DGPulldown, etc.

So you can all now delete RipBot and fire up your AVISynth scripts, because here is a link for the DGTonemap filter: http://rationalqm.us/DGTonemap.rar

Code:

DGTonemap 1.2
-------------

DGTonemap provides filters for Reinhard and Hable tonemapping operators. This
is a starting point to make available a full HDR->SDR process for Avisynth+.
In the future, I will likely add other operators. Suggestions in this
regard will be appreciated. I will also port it to CUDA.

Requirements
------------

1. Avisynth+ MT recent version
2. avsresize (https://forum.doom9.org/showthread.php?t=173986)
(be sure to get the latest high-performance version r1d)

Reinhard Usage
--------------

DGReinhard(clip, float contrast, float bright)
contrast default: 0.3
bright default: 5.0

Input must be CS_RGBPS, see z_ConvertFormat() in the script below.
If 'clip' is omitted, then as usual implicit last is used.

Higher numbers for bright reduce brightness, you can even go to bright=1000 and beyond,
but it stops having an effect up there.

Hable Usage
--------------

DGHable(clip, float exposure, float a, float b, float c, float d, float e, float f, float w)
exposure default: 2.0
a default: 0.15
b default: 0.50
c default: 0.10
d default: 0.20
e default: 0.02
f default: 0.30
w default: 11.2

Input must be CS_RGBPS, see z_ConvertFormat() in the script below.
If 'clip' is omitted, then as usual implicit last is used.

Meaning of Hable coefficients:

exposure: Gain to apply.
a: Shoulder strength
b: Linear strength
c: Linear angle
d: Toe strength
e: Toe numerator
f: Toe denominator
w: Linear white point

hable(x) = ((x*(a*x+c*b)+d*e) / (x*(a*x+b)+d*f)) - e/f
output_pixel = hable(exposure * input_pixel) / hable(w)

Sample Script
-------------

Following is a sample script for the full HDR->SDR process. The source
UHD video is tonemapped, converted to rec709 colorspace, and downsampled
to 8 bits. To leave it in 16 bits, set pixel_type=YUV420P16 in the second
z_ConvertFormat() call.

loadplugin("DGDecodeNV.dll")
loadplugin("DGTonemap.dll")
loadplugin("avsresize.dll")
SetFilterMTMode("z_ConvertFormat", MT_MULTI_INSTANCE) # May not be needed.
DGSource("THE GREAT WALL.dgi",fulldepth=true)
ConvertBits(10) # Use 12 for 12-bit
z_ConvertFormat(pixel_type="RGBPS",colorspace_op="2020ncl:st2084:2020:l=>rgb:linear:2020:l", dither_type="none")
# Choose one of these:
DGReinhard()
#DGHable()
z_ConvertFormat(pixel_type="YV12",colorspace_op="rgb:linear:2020:l=>709:709:709:l",dither_type="ordered")
prefetch(4)

Copyright (c) 2018 Donald A. Graft, All rights reserved

Edit: Quote to code tag.

[Chewtobacca]

^ Fascinating!  It's interesting that the author says that it's a starting point.

Putting the information in code tags should get rid of the emoji, which doesn't look like it knows what it's doing there. :-)

[Beber]

So you can all now delete RipBot and fire up your AVISynth scripts

"All"? No. Non-tech guys like me need a GUI as simple as RipBot264. I'm clueless about scripts.

[X5gb]

Nothing wrong with Ripbot, why the need to get rid of it. Reminds me alot of Simple x264 encoder only with the added convenience of converting as well.

[Doctor M]

So you can all now delete RipBot and fire up your AVISynth scripts

"All"? No. Non-tech guys like me need a GUI as simple as RipBot264. I'm clueless about scripts.

It just seems that with access directly to the filter, your output would be much more tune-able.
It would also give better flexibility with the encoder settings, something else people have been looking for.

[zoidberg]

It just seems that with access directly to the filter, your output would be much more tune-able.
It would also give better flexibility with the encoder settings, something else people have been looking for.

Agreed, especially when it comes to edits or regrade projects. Create a lossless intermediate of the downconversion and proceed from there. Also things like tweaking encodes would be far easier.

[spoRv]

What's about normal Avisynth?

[DoomBot]

Has anyone tried this yet http://rationalqm.us/DGTonemap.rar with Avisynth

[Chewtobacca]

What's about normal Avisynth?

I don't think it's got the support for high bit-depth needed.

Has anyone tried this yet http://rationalqm.us/DGTonemap.rar with Avisynth

That's the one mentioned in post#102. People have been using it via RipBot264 (without realizing it was what was being used).