Baselight to Nuke BLG tutorial


(Andreas Brückl) #1

Hi all,

I made a small roundtrip tutorial for BLGs from Baselight to Nuke.
I am working on all my films in ACES or an ACES flavour and usually with BLG files for the VFX departments. In the past I have invited the VFX sups of the different facilities and explained them how ACES and the BLGs work (If it was new to them). So this video hopefully helps some of you guys to implement that awesome plugin from Baselight in your pipeline.
Thanks. Have fun.

Andreas


(Andreas Brückl) #2

The tutorial is kept very simple. Here are some additional details on ACES:

The Academy has defined ACES Linear AP0 as ACES exchange format and not ACEScg. However it may be simpler in practice to deliver VFX shots in ACEScg because it reduces complexity on the VFX vendors side.

The main difference between VFX grading on a projector and VFX pregrading on a monitor is that on a monitor you are closer to the viewing condition of the compositing artist (dim surround and active display).

If other DRTs than the ACES RRT are used, it is not 100% ACES complient but still workflow parts of the ACES Framework can be used.


(Walter Arrighetti) #3

Hi Andreas.
Thanks for the tutorial. I also think BLG are very powerful, particularly for workflows roundtripping across several products (including Baselight Plugins for NUKE and Avid). I used BLGs in a couple of major theatrical projects and many people I worked with can’t stop using it now.

If you allow me, a few minor corrections to your last post:

  • The space you are referring on top is really called ACES2065-1 (because it is defined in SMPTE standard ST2065-1), not at all “ACES Linear AP0” which is how it’s referred to in FilmLight products. Since interoperability is a key purpose of ACES, let’s all just use consistent names across the ecosystem.
  • ACEScg should be used in CG/VFX workflows and, in fact, it very well is. I think you’re referring here to the fact that, by specs, footage should always be rendered in ACES2065-1 rather then in ACEScg. What’s wrong if the color-encoding within the files is ACES2065-1 rather than ACEScg? All professional CG tools I know of can be easily configured to apply encodings before reading/writing files. If all stick with OpenEXRs in ACES2065-1 colorimetry (as a matter of fact, described in another SMPTE standard: ST2065-4), that helps any workflow that “touches” at least another department (e.g. color-grading?) that is not doing CG. So, if anyone has got ACES-compliant OpenEXR footage, they don’t need to interpret or figure out anything else about the color-space footage is in: whether ACEScc, ACEScg, ACES2065-1, or any other sort of colorimetry.
  • If other DRTs different from the ACES RRT are used, you’re not using ACES color-management any longer. Period. This is, by the way, not because “the RRT” is no longer used, but because all the ODTs that any product is using ACES with are invalidated.

In principle, in fact, you can use another DRT as an alternative RRT in an ACES wotkflow, but then your output colorimetry is no longer interoperable with that of any other ACES product (unless you change all the ODTs and communicate them to the other facilities and departments in your workflow) so, unless this is done very efficiently to still justify an ACES workflow overall, what’ the point of saying “it’s still ACES”, if no ACES component is used any longer (apart, maybe, from the IDTs) ?


(Andreas Brückl) #4

Hi Walter,

thank you very much for your input. And yes, all your points are 100% right.
My issue was “how do I make it easy for users” and how can I make them use ACES.

I am dealing with post facilities in Asia that told me that they don’t want to work in ACES. After a while I found out that they are scared of the size of uncompressed EXRs and not ACES.
Fact is, they didn’t really know in-deep what ACES is. And I am talking about companies with 200 and more people. I can tell you it was a fight to get them to use ACES.
So I started using BLGs which made it quite convenient for both sides.
Additional I invited the supervisors, told them about ACES and went through the process with them. Step by step, from grading to VFX. And the moment I tell them about ACES2065-1 they are looking at me like I am a big fat cow.
And I seem to become an even bigger fat cow when I talk to colorists about ACES. Sadly a lot of guys just want to know “the magic setting” which works for all projects.
The majority of colorists still works display referred, rather than scene referred. Sadly.

For the DRTs. You are 100% right again, of course. :slight_smile:
I could call it ACES flavoured or “kind of ACES” or “based on the priciples of ACES”. Colorists (including me) have started mixing DRT’s to get different looks and be more flexible. Or some change the colorspace at the end of their process (before the ODT) to use their old REC.709 Luts.
That’s not ACES anymore and causes other problems like false colors or loosing range later when it comes to HDR. As a colorist I am quite happy that baselight gives me a few more possibilities and prevents me from shooting myself in the knee. Especially with TCAM and the use of look operators.
I think the issue of understanding ACES is more on the colorists side than the VFX/CG side. Especially if the colorist is not a technical guy at all. Which probably most of them are not.
Not every facility has engineers or workflow specialists.

My purpose for the video was to get people to use ACES, and make it easier for them with BLGs.
My issue was where do I start and stop or how complicated should it be? So I compromised a bit to make it easier. Sorry for that.
In fact I am very thankful that you gave me your feedback and corrections.

Thanks again.
Best

Andy


(Daniel Brylka) #5

Hi Andreas,

thanks for sharing your video tutorial.

I have a question on the grading that you apply before handing out the files to VFX.
I suppose you only use (in Nuke terms) gain (multiplication) and offset (plus) of the RGB values.

Otherwise your grade would not be scene linear anymore and won’t match easily with rendered 3D elements that are lit with a good HDRI image from the actual scene where the shot was filmed.

We don’t have a Baselight here, but I would be interested to get a BLG file for a shot.
Would it be possible that I could provide you a demo camera file and you send me back the BLG file with a grade?

Thanks in advance.

Daniel Brylka


(Andreas Brückl) #6

Hi Daniel,

sure, please send me the footage. I will send you the BLG back.
We are working on a project with a lot of 3D creatures at the moment. Didn’t hear any complaints from our Nuke artists. As I see it you stay scene linear all the way and the BLG file manages the ODT part smoothly.
The ACES DRT feels always a bit contrasty and saturated so a less contrasty grade (which I usually give to Nuke) might even support your 3D render as your highlights come in softer.

Best
Andy


(Daniel Brylka) #7

The moment you change anything else apart from offset & gain in base light or anywhere else after the IDT you are still working in scene linear but your result is not scene linear anymore.

Means if you light up the plate with a gamma operation and try to comp the 3D renders they also needs a gamma operation before merging the two elements. Because unless you cheat something in 3D the result will be scene linear.

About the footage:
Do you use the ARRI Sample Footage from their FTP server?
For example use a frame like
ftp://ftp-footage.arri.de/ALEXA%20LF/ARRIRAW/B003R1ZA/B003C002_180327_R1ZA/4448x1856/B003C002_180327_R1ZA.1196200.ari

Thanks a lot for your help.

If you prefer a frame that I send to you, I will provide you one.