David Battistella
Well-known member
- Joined
- Apr 30, 2007
- Messages
- 10,003
- Reaction score
- 6
- Points
- 38
Bear with me, this is a long post
I’m writing this post in a public forum because I know the RED team reads this stuff and because I firmly believe that the improvement of REDCODE will have more of an impact than the further improvement of the cameras themselves. I also want to flush out what I don’t know or understand.
With all of this talk about new camera’s and the potential of new cameras, I wanted to take things in a different direction and back toward post production. The most attractive thing to me about the entire RED system is not the Camera. It’s not the RED One, it’s not the EPIC or the Scarlet, in fact it is not any of the camera offerings of RED. What I love about the RED system id REDCODE RAW.
The very idea of RED is fantastic. The whole system, as it is laid out is really intelligent and that is what attracted me and many of the other early adopters to it in the first place. The camera records REDCODE, then through a de-bayering process gives you access to the files through some RED software or you can watch your footage quickly and easily with QuickTime and you can (not really) edit those proxy files in Final Cut Pro, etc.
Shooting RAW gives RED the advantage over many systems, but the disadvantage has been a lack of REALTIME performance of REDCODE. This lack of performance has been the bane as well. The potential is still amazing.
I think that the single thing that will make RED a stronger force in all it’s marketplaces is the improvement and optimization of REDCODE on different platforms.
But REDCODE, the cornerstone of the program and the one that people rely on for the post production half of RED, is not real time. I know that there are those who will say that a scratch system plays back redcode real time but please hear me and read on before you jump all over me on this.
Imagine if the R3D debayering was realtime all the time?
Here is what I mean. I know that REDROCKET helps accomplish this but read on to see how great it could be if it was truly a realtime system.
The Scarlett -
If the fixed lens Scarlett was released with a realtime REDCODE codec I think that would make a much more attractive platform for people who want to go the DSLR route. You’d get RAW motion image in REALTIME instead of a baked in H264 codec that the current DSLR’s offer. Attractive and powerful tool.
RED ONE -
Jim has stated that with the release of EPIC they want to avoid the “beta” problems they faced with the release of the RED One program. I would ask RED to seriously not underestimate how much of those growing pains can be attributed directly toward workflow, post production, the lack of realtime performance and the fears producers had about shooting RED because the post tools were lagging so far behind the camera department. This forced people into complex workflows involving third party software did not match the elegance of the rest of the system.
Having a post background and owning a RED is a major advantage, but we can all agree that the early days of having to transcode footage using Apple compressor, creating DPX files, rendering out of REDCINE or REDALERT, all of which were clunky and unsophisticated ways of extracting images from RED and, dare I say, completely defeats the purpose of shooting RAW.
EPIC -
The high end future is bright, but I wonder about what bringing 5K into post will do and how that will tax existing storage and backup workflows. We all want the best image quality, more latitude and better images. I think there is a great need to push the optimization of REDCODE even further. Why? We are pretty capped out right now in terms of computer speeds, RAM and processors and even a REDROCKET card (a very great move forward) still taxes systems and puts things at their limit. I actually believe that this is exactly what the EPIC program and future camera offering should do.
What I’d like to see:
Release SCARLET with and out-of-the-box optimized REDCODE that is real time. It would be nice if the CORE of the Scarlett program was to give users all of the functionality of P2 or SRS data based cameras but with an optimized REDCODE that works the same way as these systems.
So I would shoot R3D files copy them to my system and edit in realtime in FCS, Premiere. Think of it as the high end DV codec of the future, offering RAW in Realtime would widen the use of the RED system.
For RED ONE and EPIC owners release a version of REDROCKET that works with all of our existing tools. This means that if I shoot on my REDONE and have a REDROCKET, I can edit R3D natively in FCP or premiere, etc. FCP has over a million legal licenses sold. The ability to interact directly with these applications in REAL TIME would be a giant leap forward.
Release CAMERA builds and post tool, QT updates less frequently and simultaneously.
Make transcoding OBSOLETE. I want to shoot the footage, back it up, edit the footage in R3D in FCP and finish it right there or send and XML and the r3d files to any high end system from discreet systems, Flame, Smoke, etc. to baslight, Apple Color, or any number of finishing solutions.
Transcoding to DPX is a major drawback. Do everything to encourace the use of R3D from shoot to finish. Make it the number one priority of the SDK program to ensure R3D implementation is as many systems as possible. Be on top of this as much as you are at releasing camera firmware builds.
Along with MAX mode, introduce a MIN mode that would allow for more realtime performance if I want it.
I know that these are things that many people have asked for but I think it is important to open up the discussion on what kinds of things would repair some of the myths and bad karma put out there about workflow.
As always I am sure this might spark debate. If you agree or want these features please include that too. I am hopeful the RED team will respond.
David
I’m writing this post in a public forum because I know the RED team reads this stuff and because I firmly believe that the improvement of REDCODE will have more of an impact than the further improvement of the cameras themselves. I also want to flush out what I don’t know or understand.
With all of this talk about new camera’s and the potential of new cameras, I wanted to take things in a different direction and back toward post production. The most attractive thing to me about the entire RED system is not the Camera. It’s not the RED One, it’s not the EPIC or the Scarlet, in fact it is not any of the camera offerings of RED. What I love about the RED system id REDCODE RAW.
The very idea of RED is fantastic. The whole system, as it is laid out is really intelligent and that is what attracted me and many of the other early adopters to it in the first place. The camera records REDCODE, then through a de-bayering process gives you access to the files through some RED software or you can watch your footage quickly and easily with QuickTime and you can (not really) edit those proxy files in Final Cut Pro, etc.
Shooting RAW gives RED the advantage over many systems, but the disadvantage has been a lack of REALTIME performance of REDCODE. This lack of performance has been the bane as well. The potential is still amazing.
I think that the single thing that will make RED a stronger force in all it’s marketplaces is the improvement and optimization of REDCODE on different platforms.
But REDCODE, the cornerstone of the program and the one that people rely on for the post production half of RED, is not real time. I know that there are those who will say that a scratch system plays back redcode real time but please hear me and read on before you jump all over me on this.
Imagine if the R3D debayering was realtime all the time?
Here is what I mean. I know that REDROCKET helps accomplish this but read on to see how great it could be if it was truly a realtime system.
The Scarlett -
If the fixed lens Scarlett was released with a realtime REDCODE codec I think that would make a much more attractive platform for people who want to go the DSLR route. You’d get RAW motion image in REALTIME instead of a baked in H264 codec that the current DSLR’s offer. Attractive and powerful tool.
RED ONE -
Jim has stated that with the release of EPIC they want to avoid the “beta” problems they faced with the release of the RED One program. I would ask RED to seriously not underestimate how much of those growing pains can be attributed directly toward workflow, post production, the lack of realtime performance and the fears producers had about shooting RED because the post tools were lagging so far behind the camera department. This forced people into complex workflows involving third party software did not match the elegance of the rest of the system.
Having a post background and owning a RED is a major advantage, but we can all agree that the early days of having to transcode footage using Apple compressor, creating DPX files, rendering out of REDCINE or REDALERT, all of which were clunky and unsophisticated ways of extracting images from RED and, dare I say, completely defeats the purpose of shooting RAW.
EPIC -
The high end future is bright, but I wonder about what bringing 5K into post will do and how that will tax existing storage and backup workflows. We all want the best image quality, more latitude and better images. I think there is a great need to push the optimization of REDCODE even further. Why? We are pretty capped out right now in terms of computer speeds, RAM and processors and even a REDROCKET card (a very great move forward) still taxes systems and puts things at their limit. I actually believe that this is exactly what the EPIC program and future camera offering should do.
What I’d like to see:
Release SCARLET with and out-of-the-box optimized REDCODE that is real time. It would be nice if the CORE of the Scarlett program was to give users all of the functionality of P2 or SRS data based cameras but with an optimized REDCODE that works the same way as these systems.
So I would shoot R3D files copy them to my system and edit in realtime in FCS, Premiere. Think of it as the high end DV codec of the future, offering RAW in Realtime would widen the use of the RED system.
For RED ONE and EPIC owners release a version of REDROCKET that works with all of our existing tools. This means that if I shoot on my REDONE and have a REDROCKET, I can edit R3D natively in FCP or premiere, etc. FCP has over a million legal licenses sold. The ability to interact directly with these applications in REAL TIME would be a giant leap forward.
Release CAMERA builds and post tool, QT updates less frequently and simultaneously.
Make transcoding OBSOLETE. I want to shoot the footage, back it up, edit the footage in R3D in FCP and finish it right there or send and XML and the r3d files to any high end system from discreet systems, Flame, Smoke, etc. to baslight, Apple Color, or any number of finishing solutions.
Transcoding to DPX is a major drawback. Do everything to encourace the use of R3D from shoot to finish. Make it the number one priority of the SDK program to ensure R3D implementation is as many systems as possible. Be on top of this as much as you are at releasing camera firmware builds.
Along with MAX mode, introduce a MIN mode that would allow for more realtime performance if I want it.
I know that these are things that many people have asked for but I think it is important to open up the discussion on what kinds of things would repair some of the myths and bad karma put out there about workflow.
As always I am sure this might spark debate. If you agree or want these features please include that too. I am hopeful the RED team will respond.
David
