from aces-dev.

Adding @JGoldstone to the party! :)

from aces-dev.

Something like that fixes all the current ACEStransformID on my end and gives an idea of all the problems:

def patch_invalid_id(id_):
    if not id_.startswith(ACES_URN):
        id_ = f'{ACES_URN}:{id_}'

    if id_.endswith('a1.v1'):
        id_ = id_.replace('a1.v1', 'a1.1.0')

    if id_.endswith('a1.v2'):
        id_ =  id_.replace('a1.v2', 'a1.2.0')

    if 'Academy.P3D65_108nits_7.2nits_ST2084' in id_:
        return id_.replace('7.2', '7')
    elif 'ACEScsc' in id_ and not 'ACEScsc.Academy' in id_:
        return id_.replace('ACEScsc', 'ACEScsc.Academy')
    elif 'ACESlib' in id_ and not 'ACESlib.Academy' in id_:
        return id_.replace('ACESlib', 'ACESlib.Academy')
    elif 'ACESutil' in id_ and not 'ACESutil.Academy' in id_:
        return id_.replace('ACESutil', 'ACESutil.Academy')
    else:
        return id_

from aces-dev.

urn:ampas:aces:transformId:v1.5:IDT.ARRI.Alexa-v2-logC-EI160-CCT3900.a1.v1
urn:ampas:aces:transformId:v1.5:IDT.Sony.SLog1_SGamut_10i.a1.v1

These are correct according to 4.6.1

4.6.1 Input Transforms (IDTs)
URN:IDT.Namespace.DeviceName.aACESmajorVersionNumber.vIDTversionNumber

urn:ampas:aces:transformId:v1.5:ACESutil.Lin_to_DolbyPQ.a1.0.3 (No Academy namespace)

Also correct according to 4.5.3. (whether or not the way 4.5.3 is written makes sense is a separate issue)

4.5.3 ACES Core Libraries and Utilities
Academy-supplied core libraries and utilities’ major version number shall match the ACES System Version Major Version number. Core library and utilities Transform Identifiers shall use the following versioning convention:
URN:ACESlib.Name.aACESmajorVersionNumber.ACESlibMinorVersionNumber.ACESlib PatchVersionNumber
URN:ACESutil.Name.aACESmajorVersionNumber.ACESutilMinorVersionNumber.ACES utilPatchVersionNumber

from aces-dev.

Thanks @scottdyer! I did not read further that section 4.1 because I was naive in assuming that the specification was consistent which it is not!

A few things then:

• I reckon that a note should be added in Section 4.1 that the convention is not consistent across transforms, i.e. where applicable is not enough.
• ACESutil.HLG_to_DolbyPQ_1000nits.a1.1.0 is missing the URN
• urn:ampas:aces:transformId:v1.5:RRTODT.Academy.P3D65_108nits_7.2nits_ST2084.a1.1.0 is invalid because the name is split in two with the dot in 7.2nits.

I will update my code tonight and see what is left!

from aces-dev.

Yeah, it's extremely confusing. There are almost certainly transforms that were missed when the transform IDs were last updated to add the urn prefix. I used a multi-file find & replace and so if there were any deviations from the match pattern I used then I probably missed one or two. We can fix them all once we think we've found them all.

from aces-dev.

Some of the ACEScsc are not correct:

// ACEScsc.CLog2_CGamut_to_ACES.a1.v1

No URN, no namespace and no patch version number!

from aces-dev.

Kind of validates the point that there should only be a single naming convention, the first one described in the spec :)

from aces-dev.

Here is the current list of ugly ducklings if we are sticking to the current spec:

Missing URN

• ACESutil.HLG_to_DolbyPQ_1000nits.a1.1.0
• ACEScsc.ACES_to_CLog3_CGamut.a1.v1
• ACEScsc.CLog3_CGamut_to_ACES.a1.v1
• ACEScsc.ACES_to_CLog2_CGamut.a1.v1
• ACEScsc.CLog2_CGamut_to_ACES.a1.v1
• ACEScsc.ACES_to_Log3G10_RWG.a1.v1
• ACEScsc.Log3G10_RWG_to_ACES.a1.v1
• ACEScsc.ACES_to_LogC_EI800_AWG.a1.v1
• ACEScsc.LogC_EI800_AWG_to_ACES.a1.v1
• ACEScsc.ACES_to_VLog_VGamut.a1.v1
• ACEScsc.VLog_VGamut_to_ACES.a1.v1
• ACEScsc.ACES_to_SLog3_SGamut3Cine.a1.v1
• ACEScsc.SLog3_SGamut3Cine_to_ACES.a1.v1
• ACEScsc.ACES_to_SLog3_SGamut3.a1.v1
• ACEScsc.SLog3_SGamut3_to_ACES.a1.v1

Invalid Separator

• urn:ampas:aces:transformId:v1.5:RRTODT.Academy.P3D65_108nits_7.2nits_ST2084.a1.1.0
• urn:ampas:aces:transformId:v1.5:InvRRTODT.Academy.P3D65_108nits_7.2nits_ST2084.a1.1.0

Missing Namespace

• ACEScsc.ACES_to_CLog3_CGamut.a1.v1
• ACEScsc.CLog3_CGamut_to_ACES.a1.v1
• ACEScsc.ACES_to_CLog2_CGamut.a1.v1
• ACEScsc.CLog2_CGamut_to_ACES.a1.v1
• ACEScsc.ACES_to_Log3G10_RWG.a1.v1
• ACEScsc.Log3G10_RWG_to_ACES.a1.v1
• ACEScsc.ACES_to_LogC_EI800_AWG.a1.v1
• ACEScsc.LogC_EI800_AWG_to_ACES.a1.v1
• ACEScsc.ACES_to_VLog_VGamut.a1.v1
• ACEScsc.VLog_VGamut_to_ACES.a1.v1
• ACEScsc.ACES_to_SLog3_SGamut3Cine.a1.v1
• ACEScsc.SLog3_SGamut3Cine_to_ACES.a1.v1
• ACEScsc.ACES_to_SLog3_SGamut3.a1.v1
• ACEScsc.SLog3_SGamut3_to_ACES.a1.v1

Invalid Versioning

• ACEScsc.ACES_to_CLog3_CGamut.a1.v1
• ACEScsc.CLog3_CGamut_to_ACES.a1.v1
• ACEScsc.ACES_to_CLog2_CGamut.a1.v1
• ACEScsc.CLog2_CGamut_to_ACES.a1.v1
• ACEScsc.ACES_to_Log3G10_RWG.a1.v1
• ACEScsc.Log3G10_RWG_to_ACES.a1.v1
• ACEScsc.ACES_to_LogC_EI800_AWG.a1.v1
• ACEScsc.LogC_EI800_AWG_to_ACES.a1.v1
• ACEScsc.ACES_to_VLog_VGamut.a1.v1
• ACEScsc.VLog_VGamut_to_ACES.a1.v1
• ACEScsc.ACES_to_SLog3_SGamut3Cine.a1.v1
• ACEScsc.SLog3_SGamut3Cine_to_ACES.a1.v1
• ACEScsc.ACES_to_SLog3_SGamut3.a1.v1
• ACEScsc.SLog3_SGamut3_to_ACES.a1.v1

from aces-dev.

Can we get this cleaned up by then?

I hope so!

@scottdyer, @aforsythe : Here is another one whose ACESTransformID is incorrect:

s/Rec2020_100nits/Rec2020_P3D65limited_100nits_dim

from aces-dev.

I just found out two other problematic ones that are fine from a specification standpoint but are not consistent:

odt/p3/ODT.Academy.P3D65_Rec709limited_48nits.ctl

https://github.com/ampas/aces-dev/blob/master/transforms/ctl/odt/p3/ODT.Academy.P3D65_Rec709limited_48nits.ctl

// <ACEStransformID>urn:ampas:aces:transformId:v1.5:ODT.Academy.P3D65_709limit_48nits.a1.1.0</ACEStransformID>

709limit where all the other ones will consistently use Rec and limited:

• ODT.Academy.Rec2020_Rec709limited_100nits.a1.1.0
• InvODT.Academy.DCDM_P3D65limited.a1.1.0
• ODT.Academy.Rec2020_Rec709limited_100nits.a1.1.0

from aces-dev.

This one should be fine to close now!

from aces-dev.