Best ACES workflow with C300 MkII

Hi. We are shooting on couple Canon C300 Mark II in this specifications:
3840x2160 23.976
XF-AVC Intra (.MXF)
Canon CLog3 in Cinema Gamut (Rec.709)

In DaVinci Resolve — Color Science is ACES and ACES Input Device Transform is set to Canon C300MkII Clog Daylight v1.0 Cinema. While Output Device Transform is set to Rec.709.

If anyone have any experience in working with ACES, I would appreciate your ideas in finding best workflow in this case.

First of all, since there is no CLog3 IDT in Resolve, I'm not sure if color science is even transforming properly when I'm using CLog IDT. Currently, the footage Saturation is changing drastically from location to location. But if I use Color Space Transform in the Color tab on the same clip — the saturation will go to normal level.

Of course it will be color corrected soon, but I just want to make sure the starting point is correct.

Is it better to just use IDT Rec.709 and while footage will look like Log straight from the camera, and just go from there? Or maybe C200 CLog3 (which is included in DaVinci) is a better one to use).

If someone else will be looking for this, here is a solution I find so far:

Download official IDT (.ctl) from Canon website.
Open CLog3 file and grab color matrix values from there. Both for CLog3 and for ACES.
Make your own DCTL and import into DaVinci.
Make ACES Input Device Transform as No Input Transform.
Select you new DCTL in Input DaVinci CTL setting.

.dctl for that looks like this:

Code: Select all

// IDT CLog3 to ACES.
__DEVICE__ float CanonLog3_to_Linear( float clog3 ){
   float out;

   if(clog3 < 0.097465473)
      {out = -( pow( 10, ( 0.12783901 - clog3 ) / 0.36726845 ) - 1 ) / 14.98325;}
   else if(clog3 <= 0.15277891)
      {out = ( clog3 - 0.12512219 ) / 1.9754798;}
   else
      {out = ( pow( 10, ( clog3 - 0.12240537 ) / 0.36726845 ) - 1 ) / 14.98325;}

   return out;
}
__DEVICE__ float3 transform(int p_Width, int p_Height, int p_X, int p_Y, float p_R, float p_G, float p_B)
{
   
   float r_Lin = 0.9 * CanonLog3_to_Linear( p_R );
   float g_Lin = 0.9 * CanonLog3_to_Linear( p_G );
   float b_Lin = 0.9 * CanonLog3_to_Linear( p_B );

    float r_Out = (r_Lin * 0.817416293) + (g_Lin * 0.090755698) + (b_Lin * 0.091828009);
    float g_Out = (r_Lin * -0.035361374) + (g_Lin * 1.065690585) - (b_Lin * 0.030329211);
    float b_Out = (r_Lin * 0.010390366) - (g_Lin * 0.299271107) + (b_Lin * 1.288880741);
 
    return make_float3(r_Out, g_Out, b_Out);
}


The color in the footage looks exactly how it should. It's still not a perfect solution, because using when using other cameras on the timeline, their colorspace is completely off. Before I was able just to select a different ACES IDT for them, but now looks like the only solution is to apply DCTL only for C300 footage manually.

If someone wants more info about this - let me know, I'm happy to help with what I figured out so far.

Are you using R15? There are CLog3 IDTs included.

Are you using R15? There are CLog3 IDTs included.

Yes, version 15. In the list of IDTs there are only CLog3 for Canon C200 cameras, not C300 MkII (as we use).

Is the response going to be different for the IDTs? It shouldn't be. CLog3 defines a specific gamma and the various color space options (cinema, REC709, REC2020) all define specific color primaries.

Are you using R15? There are CLog3 IDTs included.

Yes, version 15. In the list of IDTs there are only CLog3 for Canon C200 cameras, not C300 MkII (as we use).

The CLog3 transfer function is not camera dependent, just use the CLog3.

You are right, switching between C200 Log2 and C300 Log2 did't have any effect on the color, so I just went with C200 CLog3 and will be using that from now on. Having a variety of camera models in the list confused me in the first place.

Thanks for your help!

P.S. At least I know how to make DCTL from CTLs now, haha.

P.S. At least I know how to make DCTL from CTLs now, haha.

Working it out for yourself is always the best method, but there are other examples online that may come in useful.

https://github.com/baldavenger/ACES_DCTL

P.S. At least I know how to make DCTL from CTLs now, haha.

Working it out for yourself is always the best method, but there are other examples online that may come in useful.

https://github.com/baldavenger/ACES_DCTL

Hi Paul,

I've tried to import those DTLS from Github several times inside DaVinci Resolve.The problem is that inside folders there are ".h" files, and no ".dctl". When I put the "ACES DCTL folder" inside "LUT" folder from DaVinci, nothing happens. DaVinci does not recognize these CTLs. I,ve read this forum and ACES Central too, but I didn´t achieve to find a solution. I´m just a colorist and this programation language is too difficult for me. Could you please take me a hand??

Thanks!

Hi Paul,

I've tried to import those DTLS from Github several times inside DaVinci Resolve.The problem is that inside folders there are ".h" files, and no ".dctl". When I put the "ACES DCTL folder" inside "LUT" folder from DaVinci, nothing happens. DaVinci does not recognize these CTLs. I,ve read this forum and ACES Central too, but I didn´t achieve to find a solution. I´m just a colorist and this programation language is too difficult for me . Could you please take me a hand??

Thanks!

Hi Paul,

I've tried to import those DTLS from Github several times inside DaVinci Resolve.The problem is that inside folders there are ".h" files, and no ".dctl". When I put the "ACES DCTL folder" inside "LUT" folder from DaVinci, nothing happens. DaVinci does not recognize these CTLs. I,ve read this forum and ACES Central too, but I didn´t achieve to find a solution. I´m just a colorist and this programation language is too difficult for me . Could you please take me a hand??

Thanks!

Elena,

If you are talking about IDT — try renaming .h to .dctl and putting it in /Library/Application Support/Blackmagic Design/DaVinci Resolve/LUT/DCTL/

Otherwise, you should be able to create your own .dctl files, using the values from that github link.