LUT's that emulate the ACES workflow

Tags: #<Tag:0x00007fd92c8ac368>

This is a new page at CML reproduced in its entirety, hopefully the links will work!

ACES LUT’s for use in a non-ACES environment
The LUT’s attached below were developed by Geoff Boyle of CML and Nick Shaw of Antler Post with the cooperation of the NSC and Netherlands Film Academy.

The reason for these LUT’s is to enable low budget and very fast paced productions to have the advantages of an ACES workflow whilst shooting with their "normal"kit.

The LUT’s are designed to take the log output of many commonly used cameras and enable you to use a standard rec 709 monitor and reproduce the "look"of having gone through a complete ACES workflow. IE they contain the effect of the relevant IDT, RRT & ODT.

If you use these LUT’s when shooting you should be able to use "standard"IDT’s & ODT’s in post with no LUT and get exactly what you saw when you were shooting.

We are currently working on versions of these LUT’s that will load directly into the monitoring of the cameras that are capable of this. Currently you will need to use an external LUT box between the camera and the monitor. I have personally done this with an Odyssey and a Terradeck Bolt, we are testing and talking to manufacturers to establish what their systems need.

Working this way enables you to use a conventional film style workflow, ie no grading onset, no DIT, just a very simple data transfer and backup :slight_smile:

It is easy using Prelight to generate this type of file incorporating a “look” if you follow this route you will need to generate 2 LUT’s in Prelight, one incorporating the ACES "look"and one without. This is simply clicking 2 buttons withing Prelight.

There are two versions of these LUT’s, one with normal contrast and one with contrast reduced to 0.85, this lower contrast is my personal choice.

ACES 1.0.3 LUTs for On-Set Preview normal contrast

ACES 1.0.3 LUTs for On-Set Preview lower contrast

from CML in association with Antler Post

These LUTs are designed only for use in LUT boxes or LUT capable monitors to preview the “base look” of ACES when shooting. They should not be used in post production. Use software which supports ACES directly.

Each LUT comes in two versions, designated EE or LL in the file name, folowing the convention used by the ARRI LUT Generator. Both versions are designed to achieve the same result, but you need to choose the appropriate one, depending on the way your LUT box works, or has been set up. EE means Extended range in and Extended range out, and is designed for LUT boxes which take the 64-940 SDI range (0-100%) and scale it to 0-1 before applying the LUT, then scale the result back to 64-940. LL means Legal range in and Legal range out, and is designed for LUT boxes which apply the LUT directly to the un-scaled SDI code values. A good double-check of whether you are using the correct version of the LUT is to overexpose an image and look at it through the LUT on a waveform. ACES ODTs roll off to 100%, so if you see “super-whites” in the image, or if the whites peak below 100%, it is likely that you are using the wrong version of the LUT.

The LUTs included in the set are:

















PANASONIC_V-Log_V-Gamut_ACES_709_LL.vlt (for in camera use)









SONY_S-Log3_S-Gamut3Cine_ACES_709_MLUT.cube (for in camera use)

These LUTs are built from the CTL IDTs provided by the various manufacturers, as well as CTL transforms from A.M.P.A.S. This should not be taken as an endorsement by either the camera manufacturers or A.M.P.A.S. The LUTs are provided “as is” and you should test that they are suitable for your purposes. Neither CML nor Antler Post can be held responsible, legally or otherwise, for any damages or losses which may arise from their use.

The LUT’s were created with both Prelight and Lattice. Prelight being a WISWYG system with Lattice being a more technical but contriollable approach.

These CTLs are some custom ones created when building the CML / Antler ACES LUTs. They are provided to help you if you wish to make your own ACES LUTs within Lattice, they are not needed for Prelight.

The CTLs included in the set are:








They can be used in combination with manufacturer supplied ACES IDT CTLs, and the transform CTLs published by A.M.P.A.S. This should not be taken as an endorsement by A.M.P.A.S. of the CML / Antler CTL files.

Multiple CTL files can be combined in software such as Lattice, LightSpace or CTLrender to produce a transform which may then be saved as a LUT. The order of operations is important. The exposure adjustment operators should be used in a linear space, e.g. ACES 2065-1, and the contrast operator should be used in a logarithmic space, e.g. ACEScct.

The following example sequence of CTL files will produce a legal-to-legal LUT (i.e. one which operates on un-scaled SDI code values) for a Rec.709 / BT.1886 display. It is for an ALEXA LogC signal, reduces exposure by one stop, and contrast to 85%:










The legal to full conversion is applied at the start because the ALEXA IDT expects full range (i.e. with 0-1 representing 0-100% ‘IRE’) input. Some IDTs (e.g. those from Canon, Panasonic and Sony) expect legal range input (i.e. 0-1 represents the un-scaled 0-1023 SDI code values) in which case no input scaling operation is needed for a legal-to-legal LUT. They would however need a full to legal input scale for a full-to-full LUT.

These CTLs are provided “as is” and you should test that they are suitable for your purposes. Neither CML nor Antler Post can be held responsible, legally or otherwise, for any damages or losses which may arise from their use.


I am compiling a table of which version of the LUTs I have confirmed is appropriate for various LUT devices.

Here is my initial list:

I will update the file at the same link as I add more devices.

I’m happy to see these since the tipping point of whether a job uses ACES can easily be which LUT is used in the monitors on-set, and in the offline media that’s created there.

The EE Alexa and Red LUTs match the native ACES implementation in Flame 2018.3, comparing R3D and Arriraw files debayered to ACES directly and the same files going via Log3G10 or LogC.

I’m curious about the video levels scaling that’s needed for Panasonic and Sony - I don’t get a match unless I apply an extra full-to-legal scale as mentioned at the end of Geoff’s post. Is it documented somewhere that these IDTs expect legal range input? In the Slog2 IDT there is clearly 64-940 maths, but for Slog3 the Sony PDF seems to imply full range.

If I debayer an F65 MXF to ACES directly using the Sony RAW controls and then apply RRT+ODT, it does not match the same file debayered to Slog3/SGamutCine followed by SONY_S-Log3_S-Gamut3Cine_ACES_709_EE.cube. To match that .cube I have to debayer to SLog, apply a full-to-legal scale, then use the IDT, RRT+ODT. This makes me wonder whether the SLog3 IDT does in fact expect legal range input… I would expect Sony’s debayer-directly-to-ACES result to match if that was the case?

If working in post software, you would indeed need to add additional range scaling for Sony, Panasonic and Canon. These LUTs are designed to work in hardware LUT boxes, not post software, as the range handling for those manufacturers is different there.

Yes, indeed it does. It’s the same as S-Log2, but the range scaling is not explicit. In post software the raw decodes effectively to legal range, i.e. matches the unscaled S-Log3 SDI code values output by the camera.

I’ve added in camera LUTs for Sony & Varicam here


Geoff Boyle NSC

1 Like

I’ve just updated the links in the original post to include a package of LUTs with a lower contrast than normal. 0.85 which is my personal favourite.

I’ve also updated the original package to correct some errors in the Canon 2020 varieties, so many possibilities, so little time :slight_smile:


Geoff Boyle NSC

1 Like

The 0.85 contrast versions can be matched in Resolve or Baselight by applying 0.85 contrast, with the default pivot setting of 445 (or 0.435) in ACEScct. It can also be replicated with an ASC-CDL adjustment with Slope: 0.85 and Offset: 0.06525.


I have tested a Teradek Bolt 3000 and an AJA LUT box today and added them to the list:

1 Like

And as a bonus freebie, here is a simple drag-and-drop Automator app for OS X which will convert a 33^3 LUT to 17^3 for LUT devices which require that:

It’s not a robust universal converter, so no guarantees it will work with other LUTs, but it works with my ACES sets.

The table is updated again to add IS Mini and ATOMOS. I am working on getting other devices to test. There is also a correction to the entry for FSI monitors, due to an earlier misunderstanding over terminology.

I don’t know if this has been mentioned here before but if not here it goes:

There is now a really easy way to emulate some ACES workflows using 3D LUTs exported from DaVinci Resolve.

As of the latest versions of Resolve (15.2.2 I think but could be earlier), DaVinci’s own ACES Transform OFX Plugins as well as their Colour Space Transform Plugins can now be exported into 3D LUTs (previously they couldn’t).

If you set your Colour Management to DaVinci YRGB and on any given clip create two serial nodes with the new ACES Transform OFX plugin on each you can emulate most ACES IDT+RRT+ODT combinations and then export a 3D LUT out of them using the “Generate 3D LUT (CUBE)” command.

Basically the first node works to select the IDT and the ACES working Space and the second node works for coming from the ACES working space to the selected ODT.

If you wanted to create an ACES emulation LUT for the Alexa IDT using ACES Rec709 ODT to monitor on set for example you could easily:

Set an ACES Transform OFX plugin on the first serial node, set the ACES version to say 1.1, then on Input Transform select Alexa, on Output Transform Select say ACEScct. On the second serial node you would then chose ACEScct as the Input Transform and the “Rec709 as the output transform.

All that is left to do is right click on the clip thumbnail and select “Generate 3D LUT (CUBE)” and voila, you have a 3D LUT that combines the IDT, RRT and ODT.

As far as I know, this method will only export LUTs in Extended Range levels (Full Range). There may be a workaround to export Legal Range level LUTs in Resolve but I’m not aware of it. You could always use a tool like Lattice if you need to convert from extended to Legal range but otherwise this method seems to work wonders and it’s accessible to anyone with a copy of Resolve Studio.

Just as a side note, you can obviously put however many nodes with primary corrections between the two ACES transform nodes to export a LUT with a baked in look that goes farther than the ACES IDT+RRT+ODT transforms.

1 Like

I’ve been experimenting with this approach and it seems to work well.

I’ve added a new ACES LUT page to CML that has LUT’s generated this way for Alexa, BMD, Canon, Varicam & Venice.

They come in 2 flavours, LowCon as before, Basic again as before and a new “Kodak” version where I have tweaked the image so that the black, grey, white patches on the Kodak Grey Scale Plus card match the levels printed on the back of the chart. It’s an increased contrast from 1 to roughly 1.12.

Can these LUTs be used on the camera to emulate ACES

Can these LUTs be used on the camera (arri alexa) to emulate ACES ie. without using a LUT box?

You’ll need to use ARRI software that is downloadable from to convert them to the arri look format but then yes, you can :slight_smile:

I’ve converted the Sony and Panasonic already, I guess I’ll have to add the ARRI ones :slight_smile:
These are the original ones’s-for-use-in-a-non-ACES-environment.html
I added these ones later
I’ll try converting some of the ARRI ones to arri look format by this time next week.

Great thanks for sending through. All looking positive!

Another query I had was on the livegrade system using BoxIO for the IDT and ODT from the arri log c, the frame overlays like the red REC turns orange. Is this to do with the IDT and ODT?


That’s right. The saturated red of the REC overlay is not a colour that would ever appear in a normal LogC / AWG image. And passing it through a process intended for LogC /AWG gives that result.

Hmm, I don’t have an Alexa to check on and I’m not happy with what I’m seeing in the ARRI look tool.
There seems to be a legal/data issue.
I’ve checked the BMD ones loaded into a LUT box and they work fine (I have a G1 and a G2 here at the moment)
It’s just the conversion to ARRI camera format that seems an issue, I’m doing that with Lattice.
I’ll have another go later but it’s a pain that it only works on a Mac.

Once again Nick Shaw has come to the rescue :slight_smile:
I’ve uploaded a revised page that now include .aml look files for Alexa

1 Like