3D LUT Standardisation

Published : 13th August 2009


I heard there was some discussion on a 3DLUT standard format, so here is an update on the Common LUT Format (also sometimes known as the Academy/ASC LUT format, or Academy LUT or ASC LUT, really all of the same thing, not ASC LUT II.
The ASC-CDL was a separate development although it included many of the same people )
A main goal of the committee was to provide a LUT format that could handle metadata in a fashion consistent with current practices. XML was chosen as the format to carry this metadata. This allows further information to be attached to LUTs in a consistent fashion.
Example:
<ProcessList xmlns="urn:NATAS:ASC:LUT:v1.0" id="LUT12Jul09" name="Prod
Lut A">
<Description> Convert to rec709 from Slog with 3D LUT </
Description>
<Description> made for Prod XYZ </Description>
<InputDescriptor> Sony F23 Slog </InputDescriptor>
<OutputDescriptor> Sony BVM CRT </OutputDescriptor>
<LUT3D id="LUT120709" name="Slog convert" interpolation="trilinear"
inBitDepth="10i" outBitDepth="10i" >
<Description> converted to rec709 primaries with D65 white </
Description>
<Array dim="17 17 17 3">
0 0 0
1 1 1
3 3 3
5 5 5
***lots of other lines*************
1023 1023 1023
</Array>
</LUT3D>
</ProcessList>
One of the goals of the committee to improve adoption was to have a reference implementation to speed adoption. There are several implementation efforts underway, and there has been progress on creating an XML reader for the format. But the effort is largely volunteer, so it is taking a while to get results.
The specification is contained in :
"A Common LUT Format for Look-Up Tables", version 1.01, May 11, 2008
Ed: Jim Houston, Common LUT Format Committee, Academy of Motion Pictures Arts and Sciences
I have put this up on a temporary basis at
http://web.me.com/starwatcher/
There have been some implementation notes that may require some revisions, but these have not yet been incorporated.
Thanks for the interest in this format. We still have regular problems with LUTs on a frequent basis many of which can be traced to lack of communication of what is in the LUT. This format was designed to help, but faces a bit of a chicken and egg problem as vendors are waiting for others to use it or request it, and users don't have it available in their most common tools so they can see how it might help.


Comments welcome


Jim Houston
chair, Common LUT committee


>>Just in time for this discussion of 3D LUT standardization comes the following message from David >>Reisner who is Secretary, ASC Technology Committee and DI subcommittees.


Thanks Bill.
Regards,
Adrian Hauser
All things Color for Film and Digital Cinema
www.iris-digital.org/wordpress



Hi everyone:


Just in time for this discussion of 3D LUT standardization comes the following message from David Reisner who is Secretary, ASC Technology Committee and DI subcommittees.
Please follow the instructions below:

Hi,


The ASC CDL v1.2 release is now available. The release contains the official definition for the Saturation operator, example code and test files for Saturation implementations, an example ASC CDL workflow diagram, and small clarification of the ALE format.

To get download instructions, send an e-mail to asc-cdl@theasc.com : Notice that this is a DIFFERENT ADDRESS than you may have used before.

Tell your friends.


David Reisner
Secretary, ASC Technology Committee and DI subcommittee
D-Cinema Consulting
www.d-cinema.us



>>Thanks Dan. I'd just never heard it referred to as the "ASC" LUT standard.


A lot of the same people were involved both the ASC CDL and the Academy LUT, but I just don't want to step on any toes by not crediting the proper overseeing organization.
The LUT standard used to be available on-line but doesn't seem to be there anymore. (Google returns references to "slut academy" which is a different branch of the film industry.) Maybe Jim Houston has a current link?
Does anyone know who has implemented the full standard for reading LUTs?


Writing is easy, but supporting the full standard on input is more work; it's a comprehensive and ambitious standard. There was talk of an open - source reference implementation, but I haven't heard anything since.

Bob Currier 3349 Cahuenga Blvd West, Suite 3
CTO/Partner Los Angeles, CA 90068
Gamma & Density Co. +1 323 436-7593
www.gammadensity.com



Note:


I was quoting Robert Monihan
I don't have it myself or I would have posted it :-)
Michael Bravin

I too have the LUT specification.
If anyone wants it, drop me an email.
Shouldn't we post it here on CML?


Michael Bravin
Chief Technology Officer
Band Pro Film & Digital



Dan Mulligan wrote:


>>Bob Currier : Dan, can you expand on this? I'm aware of the ASC CDL and the Academy LUT standard,


I too have the LUT specification.
If anyone wants it, drop me an email.
It should be available on different sites, as well.


Robert Monaghan
Glue Tools LLC
629 State St. Suite 220
Santa Barbara, CA, 93101
United States
tel: +1 805 456 7997
fax: +1 805 456 7998
www.gluetools.com



Bob Currier : Dan, can you expand on this? I'm aware of the ASC CDL and the Academy LUT standard,
April 26 2007 : This process needs to be automated all through the pipeline,” Stump said. “I propose the ASC LUT — a standardization of the syntax in which LUTs are shipped from one machine to another. [Currently,] none of them have the same header information and syntax.”
The ASC LUT standard quite rightly evolved into the ASC CDL and Academy LUT we now have, the LUT came out of the same early discussions and was used, as I saw it, to show an evolving idea.
A Common File Format for Look-Up Tables by Jim Houston from April 2008 with AMPAS and the ASC technical committee also addresses this common standard for LUTs.
The goal is to improve the pipeline by enabling LUTs to be used in different technologies while maintaining a consistent meaning
Director of photography David Stump, co-chair of the ASC's metadata subcommittee, said that he has started to introduce this topic to various manufacturers, "It's all in the name of open source," Stump said. "The more people embrace open-source technology, the easier it is to do the work."


This was in 2007 and I've posted the articles and white paper through the links below. Do have a read, I'm just pointing out that the 3D LUT standard has already been looked quite deeply already for a long while, and certainly Dave and myself have had many discussions on LUTs, especially with Bronx is Burning on the Viper and the Features we were all shooting at the time and using Luther, Truelight and more for all sorts of work. It’s already out there Bob, as G&D have adopted as well.


Link to three articles on the ASC CDL and LUT standard :


http://www.rogueelementfilms.co.uk/forum/index.php?topic=221.msg299#msg299


Regards
Dan Mulligan
www.rogueelementfilms.com
Digital Film Rental, F35


Of course, you can use the Light Illusion 3D LUT Builder to convert between different LUT formats, including gamma and colour space ...


This means Geoff can stick to Iridas as a working environment, and then simply convert the LUT to the post house's required format.


Yes, yes, I obviously have an interest in this software ...


Steve Shaw
Light Illusion
UK: +44 (0)7765 400 908
India: +91 917 641 9810
www.lightillusion.com
Skype: shaw.clan
++++++++++++++++++++++++++
The Art & Science of Digital Images
++++++++++++++++++++++++++



>>I am sure you must have run into some of them. I know Wolfgang and studied at Derby ?

Geoff Boyle FBKS
Cinematographer
EU Based
Skype geoff.boyle
US +1 818 574-6134
UK +44 (0) 20 7193 3546
mobile: +44 (0)7831 562877
www.cinematography.net/geoff

**************************
>ASC LUT II standard


Dan, can you expand on this?
I'm aware of the ASC CDL and the Academy LUT standard, having sat in meetings on both, but not sure what you're referring to here.
I know Dave uses LUTs extensively, but I haven't seen anything published on a unique LUT format.

Bob Currier 3349 Cahuenga Blvd West, Suite 3
CTO/Partner Los Angeles, CA 90068
Gamma & Density Co. +1 323 436-7593
www.gammadensity.com



This subject has been a thorny issue now for quite a while. As I’m sure you are aware Dave Stump ASC developed the ASC LUT II standard that has been adopted now for a lot of post workflows to create that standard, in fact it’s something that’s been developed for a long time and not such a new idea. Certainly white papers have been written on the subject, which I'll gladly post (with permission) on our client forum.

In fact in 2007 I had written an article on this subject for HD Magazine that’s still top of their downloads list, so this is not a new subject at all. The 3D LUT standard has been here for at least two years now the next step is to try and get a standard for digital itself and prevent all the digital banter that’s floating around.

3D LUTs, workflows, standard practices etc need to be adopted, I feel, and soon. Film language is part of everyone’s DNA it’s now time to understand a new language, and I can't believe it’s taking us so long to do it?
Best
Dan Mulligan's two pence worth!
www.rogueelementfilms.com/news_press.php
www.rogueelementfilms.com/forum/index.php
Elstree Studios, UK



Geoff Boyle wrote:


>Yeah well you know, I’m just a simple bear.
>I deal with real world situations and real world problems.


I am sorry Geoff, my intent is not to offend you. Lets bury the hatchet and as I have said I am already done with this topic. BTW, just as a last comment the problem is important and even in UK several joint academic-industrial studies have been done on similar topics, such as the transfer function from film negative/positive to color spaces like XYZ, which has some bearing on this discussion.
For e.g., studies by Wolfgang Lempp at Filmlight (well at CFC at that time frame), and several researchers from the University of Derby (Noreiga, Morovic, McDonald). I am sure you must have run into some of them.


Thanks a lot.


DJ Joofa
Imaging Solutions, HD Video, and Software Architecture
Austin, Texas


>>Geoff Boyle just repeated ("we know 1D LUTs can't do saturation"), which result from a simplistic >>application of 1D LUTs and drawing conclusions from that


Yeah well you know, I’m just a simple bear.
I deal with real world situations and real world problems.

Geoff Boyle FBKS
Cinematographer
EU Based
Skype geoff.boyle
US +1 818 574-6134
UK +44 (0) 20 7193 3546
mobile: +44 (0)7831 562877
www.cinematography.net/geoff



>>Sorry to say, not something I'm really interested in wasting time on.


Steve, no where have I suggested that you spend/waste more time or change your business around making 3D LUTs. I have just pointed out some inconsistencies and misconceptions regarding 3D LUTs and 1D LUTs held by many, that for e.g., Geoff Boyle just repeated ("we know 1D LUTs can't do saturation"), which result from a simplistic application of 1D LUTs and drawing conclusions from that, while not exploring various input/output mixtures in which 1D LUTs can be configured to approximate the response of a 3D LUT.
I hope you see this viewpoint. I think I have said enough on this topic, and we all can move on.


DJ Joofa
Imaging Solutions, HD Video, and Software Architecture
Austin, Texas



Geoff Boyle wrote:


>>Well Bill when I grade something, produce a "look" I have a choice of a huge variety of formats to store it in >>depending on which system I'm going to finish in.


I was on a committee with a bunch, to create a standard file format for 3DLUTs. It was pretty good!
Jim Houston was the lead on this.


Robert Monaghan
Glue Tools LLC
629 State St. Suite 220
Santa Barbara, CA, 93101
United States
tel: +1 805 456 7997
fax: +1 805 456 7998
www.gluetools.com




Rupert I'll call them Smurfs if someone will standardise them.
Hell! I'll call them any bloody thing that anyone who produces a standard wants to call them!


Oh and please don't quote the whole message.


Geoff Boyle FBKS
Cinematographer
EU Based
Skype geoff.boyle
US +1 818 574-6134
UK +44 (0) 20 7193 3546
mobile: +44 (0)7831 562877
www.cinematography.net/geoff


>> ,,,,,,I know that I'm a voice in the wilderness, after at least 10 years of asking for this I'm no closer.,,,,,


Standardization and coherence are not in the best interest of the post-production and equipment-manufacturer profit model

.
David Perrault, CSC



Geoff


Would it not be an idea to keep LUTs for engineering purposes and
"looks" for creative ones?

Rupert Watson
www.root6.com
+44 7787 554801



>>Could you explain what you mean by "standardizing" 3D luts?


Well Bill when I grade something, produce a "look" I have a choice of a huge variety of formats to store it in depending on which system I'm going to finish in.
I'll start with .look from Iridas, then add flavours of Truelight, Luther, Black Magic, G&D, Hell! I can't remember them all!
Of course they're all incompatible with each other.
This is where I think the ASC have been taking perhaps the wrong approach, I understand the CDL approach but it's hugely limited.

If we could just get a standardised LUT format, this is for creative LUT's not engineering ones, then life would be a lot simpler.
Then instead of me getting log stills files to grade and then email back reference pics that the colourist tries to get close to I can email him a "look" file that he can load and know the he's at least staring close to the original idea.
Of course I know that I'm a voice in the wilderness, after at least 10 years of asking for this I'm no closer.
I know the monitors won't be spot on, I know that there will be variations due to room conditions, I don't care! These are creative LUT's, guides as to the look and not precise engineering ones.


The problem is that all of this is dealt with by engineers who can't cope with the imprecision we want!


Geoff Boyle FBKS
Cinematographer
EU Based
Skype geoff.boyle
US +1 818 574-6134
UK +44 (0) 20 7193 3546
mobile: +44 (0)7831 562877
www.cinematography.net/geoff


Hi Bill,
Ideally we should all use the ICC color profile format (www.color.org)
It can handle 1D, 3D LUTs and much more (metadata, etc)


Dan Tatut



 

Recent CML Tests


9 cameras RAW

RAW V Compressed F55 Alexa

Contrast handling in a real world test

Alexa 4 stops over and back


Aces Rushes One Light

Alexa - 435/5219

Alexa Epic F55 F65 single candle

Arri LUT's

Alexa BMCC C500 Epic F55

Camera Matching
Alexa BMCC C500 Epic F55 F65

Alexa-Epic-F65

Effect of noise & compression on Green Screen Alexa C500

Initial ACES tests

 

Sponsored by:-










Copyright 1996-2014 CML-Cinematography Mailing List