Welcome to our community

Be a part of something great, join today!

Log in
  • Hey all, just changed over the backend after 15 years I figured time to give it a bit of an update, its probably gonna be a bit weird for most of you and i am sure there is a few bugs to work out but it should kinda work the same as before... hopefully :)

IIF discussion thread

JonathanF said:
The one thing that this ignores is that debayering algorithms are continuing to improve and just as we have seen Graham's 'color science' evolve, we are certain to see this continue as vendors develop more and more expertise in working with raw sensor data.

If you debayer the data will you be able to take advantage of subsequent algorithmic advances in debayering technology?

Food for thought,

J
Click to expand...

I wouldn't underestimate the power of this. However let's talk practically for a second. How do you update 1,000 applications with this new debayer algorithm? Is it acceptable as a user to potentially work in two different applications which might produce two different color and image outputs?

If I am keying footage or grading secondaries and the source material changes because of a software update that's going to screw everything up.

Do you update the standard and get everyone to agree on every single new debayer and color improvement? Can color improvement made for the RED sensor apply to an Arri sensor?

I personally wouldn't feel it's the place of a standards committee to constantly upgrade the standard and pretty fundamentally change the output of the file every few months. Can you imagine if an OpenEXR's output changed every few months based on what version of the standard was last patched into the host application? You've sacrificed all claims to interchangeability and consistency.
 
Gavin Greenwalt said:
I wouldn't underestimate the power of this. However let's talk practically for a second. How do you update 1,000 applications with this new debayer algorithm? Is it acceptable as a user to potentially work in two different applications which might produce two different color and image outputs?
Click to expand...

Welcome to post production. I can't tell you how many times on major features traffic has called up a producer and said new scans were available or that a shot was omitted or added.

If the DoP convinces the Director that it's a worthwhile improvement and the studio agrees who is the post facility to argue? This happens often presently, albeit in other forms, and I have actually heard artists bitch about it, but who are they to bitch?

This is the kind of complaint a junior compositor has when they change the plates out on him. Like it's his film? Please.

Gavin Greenwalt said:
If I am keying footage or grading secondaries and the source material changes because of a software update that's going to screw everything up.
Click to expand...

Yes, this happens. This happens when you restore a film as well, ask Coppolla. The print, IN and IP stocks had changed so much when the did the Redux that the whole film needed re-timing. The same will be true as display technologies are able to use more and more of the XYZ color gamut in time. I don't see how this argument holds any water. Technology changes and if you can make the source footage less noisy, have better color fidelity, etc., why would you not do it if you have the budget?

Gavin Greenwalt said:
Do you update the standard and get everyone to agree on every single new debayer and color improvement? Can color improvement made for the RED sensor apply to an Arri sensor?
Click to expand...

No you can't do this. That's why there are so many raw formats, they are unique to sensor types.

Gavin Greenwalt said:
I personally wouldn't feel it's the place of a standards committee to constantly upgrade the standard and pretty fundamentally change the output of the file every few months. Can you imagine if an OpenEXR's output changed every few months based on what version of the standard was last patched into the host application? You've sacrificed all claims to interchangeability and consistency.
Click to expand...

That's your prerogative but I know a lot of filmmakers who would disagree with you.

Now I'm not saying that the IIF methodology is flawed, but, there is the distinction between an archival of the camera source, and an exchange of data, and mastering.

It is completely true that once a film is finished and delivered, few will ever revisit the entire post process. This is reserved for a very select number of very special films. Lawrence and Apocalypse being two off the top of my head.

Both workflows have their use which means that the goal of the Academy's technical committee is a laudable one.
 
16bit float??? might as well use a faster fixed-point system then that pre-multiplies values during input in that case..
 
JonathanF said:
Welcome to post production. I can't tell you how many times on major features traffic has called up a producer and said new scans were available or that a shot was omitted or added.
Click to expand...

I'm not saying plates shouldn't and couldn't be updated with the latest/greatest. I'm just saying I don't think there should be uncertainty at the post house as to whether or not they're actually using the same version as the DP. And if a plate changes it should be a deliberate and intentional process. If the compositor updates to the latest version of Nuke for instance you don't want an approved composite to change.

Consistency above all else in an interchange format imo. I'm arguing that a shifting debayer standard would result not in the latest/greatest being used but perhaps something different from what the DP/Director decided on in another room.

Everyone should be operating from the same script. Shifting debayer standards/color science in a 'standard interchange' format would strike me as a recipe for disaster. I would rather replace the plate then have an existing plate change without me realizing it. And that's not a theoretical problem, I've had this happen with updates or even worse this scenario which has also happened: several computers on the farm somehow don't get the latest update. Now every 16th frame is different as its rendered on the black-sheep computer and you have to track it down.

I just find a deliberate debayer process seperate from the file format to be the safest workflow for consistency and repeatability. Lest you send out a single plate to 8 different post houses and get back 8 different results. That's the exact opposite desired goal of an interchange format!
 
Gavin Greenwalt said:
Consistency above all else in an interchange format imo. I'm arguing that a shifting debayer standard would result not in the latest/greatest being used but perhaps something different from what the DP/Director decided on in another room.
Click to expand...

I think you need to read a bit more about the spec. The creation of the ACES interpretation of an image is done by putting the original image information through an input device transform, supplied (hopefully) by the manufacturer. If a manufacturer is changing their debayer routines in a way that affects the image's colorimetry, they would also supply an updated image device transform for importation into ACES. So the only question would be whether the RAW file and the IDT are in lock step. The ACES representation would always be consistent.
 
M Most said:
I think you need to read a bit more about the spec. The creation of the ACES interpretation of an image is done by putting the original image information through an input device transform, supplied (hopefully) by the manufacturer. If a manufacturer is changing their debayer routines in a way that affects the image's colorimetry, they would also supply an updated image device transform for importation into ACES. So the only question would be whether the RAW file and the IDT are in lock step. The ACES representation would always be consistent.
Click to expand...

Indeed I'm in the same boat as Mark and am largely unfamiliar with the current proposal.

Does the transform become embedded with the file? Does 'color science build 21' get embedded for instance? I'm thinking of an instance where one project is wanting to run theoretical redcode v21 and another DP is wanting to stick to the v16 look. I imagine they embed the transform in every frame so I can run both projects simultaneously and it all happens automatically behind the scenes? If that's the case then you also need a means of changing the embedded profile or overriding it. Probably similar to loading a truelight profile.

Seems like that would add a lot of overhead, although I guess PSDs do it already. I was picturing the conversion/import into the iff defined color space would ONLY happen once at file creation and forever be assumed as a static colorspace.
 
All of us come at this with our own experiences and understandings which inevitably shape our perspectives. What I believe needs to be kept in mind is that the central point of the IIF paradigm is the ACES benchmark in the middle. Whatever format comes in from whatever source is aimed at a common known target - this is a critical improvement from the wild west show out there today.

While there may be payload issues vs compressed RAW and other negatives of the IIF/ACES topology it sure seems like a worthwhile tradeoff. I also like the advantages of creating a neutral target that allows the use of output transforms to target output device spaces. With the rapid expansion of viewing technologies the days of simply making the best film print and living with the lost information when creating secondary deliverables is, if nothing else, bad economics since so much revenue comes outside the theatrical window (a situation that stands to get even more pronounced if current trends continue).

I realize the devil often lurks in the details, but, on balance, its hard for me to see how the negatives introduced by the IIF can possibly outweigh the positives.

Cheers - #19
 
Lucas Wilson said:
For anyone that wants to get involved at a deeper level...

http://www.stcatp.org/

The licensing form and sign-up instructions are on that page.

Lucas
Click to expand...
Thanks for the link Lucas.
However it does mentioned that "The Academy will make the IIF components available to Program participants for evaluation purposes until December 31, 2009."
Does that mean interested parties are now locked out of any further evaluation until the committee comes up with a recommendation?

regards
Chris
 
Chris Swinbanks said:
Thanks for the link Lucas.
However it does mentioned that "The Academy will make the IIF components available to Program participants for evaluation purposes until December 31, 2009."
Does that mean interested parties are now locked out of any further evaluation until the committee comes up with a recommendation?

regards
Chris
Click to expand...

Sign up and find out! : )

Lucas
 
You must log in or register to reply here.
Back
Top