View Full Version : Workflow question.


Bob Grant
April 23rd, 2007, 11:35 PM
First off I'd like to thank Arri for the time he devoted to my endless questions at NAB but I'm still missing a piece in this jigsaw puzzle. I've gone through the ppt on workflow and well, I'm again convinced how great the camera is and the advantages of the RAW Cineform codec. But then I'm left hanging!

I know I can modfy the look in post, but with what, Speedgrade? No problem if that's the best option, I assume we'll need Speedgrade DI. Are there any other options, AE etc?

And what about editing, compositing and the CGI team. How to keep the metadata that holds the Look through the whole flow, right to the last possible moment which I assume is grading. How to ensure the CGI elements and the composites will track any changes made in the final grade?

I realise there's probably going to be several possible workflow solutions but even one complete end to end flowchart would be helpfull. There's one on the SI-SK site but it's pretty scant on details, the Editing / Post-Production box leaves out the question of grading, unless AE or PPro are upto the task. It's been a long time since I've looked at either of those so maybe I'm missing the obvious.

David Newman
April 24th, 2007, 06:37 PM
The workflow depends mostly on the your own post production needs. Currently both Adobe Premiere Pro and After Effects can control whether the look is enabled, so they offer the most flexibility (this is being explanded to other apps and going across platform.) If you grade is a tool that support RAW then you have no issue, if not the workflow will depend on how you shot through a look. If you shoot with a subtle look, not bending the data to the extreme, your post options are more flexible, as you use the look data for just post "camera tweaks". You would perfrom these tweaks while in your NLE, then flatten to CineForm 444 or DPX or whatever you grading tools prefers. In this workflow the look files are for no destructive film stock preferences. You don't need to carry look data from the edit to your grading system, as you base look will be locked upon the flatten. All the advantages of the camera and post flexibility still hold -- just check your highlights before flattening.

Now it you apply an extreme look in camera, you have the danger of clipping highlights if you flatten to any non-RAW format. If you need to grade in a system that can't handle RAW bayer (most will not -- Iridas is working on it to be the first), export (flatten) with the look data switched off to give you a RAW 444 image (you may choose to leave white balance and matrix on for reduce grading work.) The .look file can be given to the grading system and reapplied as a starting point for the finished color.

In either workflow, you shouldn't consider the .look file generated in camera to be the final grading 3D LUT, even though that is what the 3D LUT where intended for. These LUTs are used for creative intent, so they should only enhance not burden your workflow.

Jason Rodriguez
April 24th, 2007, 07:22 PM
Hi Bob,

Thanks for all the feedback . . . I'm working on a comprehensive workflow diagram from start to finish using .look files and the color management features that .look files can help you with when going from the shoot through post.

As David mentioned, .look files are meant to help, not burden you. Getting all these questions is great because it helps to taylor what issues we need to address. There are many ways to skin the cat, but I do realize that a good starting point is needed.

Expect to see more comprehensive workflow diagrams in the coming weeks.

Thanks,

Jason

Bob Grant
April 25th, 2007, 03:42 AM
David and Jason,
thanks for the replies. After I'd made that post I did realise I was making what is a feature sound more like a burden. I'm just trying to workout how to get the most of out it.
There seems to be so much overlap between various products that it's hard to know which way to jump. For example Cinetal seem to have some great things happening with their monitors but I was still left wondering quite why I'd want all those features in a monitor as they seemed to replicate what was in the camera.
So yes, an expanded workflow diagram would be mighty useful. I think anyone will appreciate that it will not be the only way to get the job done but rather serve as a reference framework for the possible workflow variations.

The other thing that's got my interest is emulating Kodak's look management system. Will this be possible, if so I think you'll have a big marketing advantage. The cost of Kodak's system to a production would buy a few SI-2Ks.

Jason Rodriguez
April 25th, 2007, 05:36 AM
Overlap IMHO is a good thing . . . it means that you don't have to rely on just one product from one manufacturer, there is a professional community delivering products that you can pick and choose your poison from for the task at hand.

Closed and/or "turnkey" systems from the outset seem very easy to use . . . the choices are clear, and there's a fixed and very clear path to the end product . . . but the minute your imagination grows beyond the bounds of the fixed box, that "easy-to-use" box that was enabling before is now stifling. So while the myriad of choices that an open system presents the first-time user with might seem a bit daunting, there is always a "path of least resistance", and then other alternative choices that let your creativity fly in whatever direction you choose. As you mature, the system matures with you.

Bob Grant
April 25th, 2007, 07:15 AM
I'm no fan of anything closed and I'm all for open systems, I'm a Vegas user and it doesn't get much more open than that.
Where my concerns sort of lie is this. If system 1 can only do A and system 2 can do A and B and so long as it does A as well as system 1 then no sweat. But it can be hard to know just how well it does the task, I don't just mean functionality, I mean quality of the pipelines etc it's using. If we're just talking about adjusting metadata then it's easy to just choose your poison. When we start talking about rendering images or mixing audio that's when it gets more complex to choose your poison.

On the other hand when I compare the SI-2K at NAB this year to where it was at last year every development dollar spent and feature that's been added has enhanced the core purpose of the tool, to record the best possible image with the highest degree of user control and ease of use. Simply brilliant. That's why this camera has our attention.