Welcome to our community

Be a part of something great, join today!

  • 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 :)

The future of post....

Jay A. Kelley

member
Joined
Feb 2, 2007
Messages
3,803
Reaction score
0
Points
36
Location
St. Louis MO
Website
www.dreamwelder.com
Post work is getting complicated. There's no doubt about it. Tape was expensive to make, and a lot of work to sell. So manufactures were careful about how many formats were out there.

This is not the case in our new digital, tapless world. Everyone with a camera and/or capture card can design their own codec and call it God's gift to the editor. In the past codecs lived mainly with capture cards. Now with capture cards serving mainly as a viewing solution, the codecs the posses no longer have the same importance (Unless you are working with tape).

As consumers / editors, we're kind of screwed aren't we? The forest of post production was thick enough and now a bunch of companies have come along and planted a crapload more trees. In other words, we're getting a lot of formats.

This is the reason for Adobe's new program whereby they, and many others are searching for a "common standard" in terms of RAW workflow. The idea being that no matter what codec your camera uses, there will be one format you can transcode to, that will insure no matter what system you wish to work/edit on, it will support this codec. A lot of marketing is made of "WE OFFER DIRECT SUPPORT, NO TRANSCODING NESSESARY!!" And while this is a good thing for the small editor who may post AND deliver a project on one system using JUST that software family, it does very little for the editor who may need to run their project through a number of different programs.

Right now the number of choices for a "standard codec/worflow" are very slim. The one with the largest headstart would be Cineform. They offer solutions for both Final Cut Pro, and Adobe Premiere. I am sure this will be for both Windows and Mac then cs4 comes out. The only nut they have not been able to crack up to now is Avid. If they were able to get on board with Avid then Cineform would be able to make a good case for being a viable codec standard for post houses to use.

Another factor is that IF more people start using a standard codec, more pressure will fall on the editing software companies to support it well. What they don't want to hear is:

Nah, I want to use Final Cut, my codec is real time there and Vegas is slow as hell. If larger post houses start saying "we support this codec" that will also make a difference.

The problem here is that what's required for this type of idea to come off well is a strange little word called "cooperation". These codecs are not coming from little known guys in their garages. They are coming from mulimillion dollar companys that are in this game to MAKE MONEY. So it's hard for them not to see one another as competition.

The funny part is that as customers, we all know something the software companys are slower to learn: Easy post helps EVERYONE.

But let's look at that for a moment.. Does it really?

Right now a lot of RED people are moving to Adobe Premiere. The reason, as they have said time and again is the support Adobe gives RED. NOT the editing system itself. Don't get me wrong, Adobe has really improved over the years, but it's still far behind FCP and AVID in it's feature set. And is this making a difference? Not really, customers are more concerned with support for their camera codec than they are the features and benefits of the editing platform they are working on.. This strikes me as a little bass ackwards. A standard codec would put the editor's attention back where it's suppose to be, on the best tool for job, rather than the only tool that works.

Now this may be blasphemy to say on a RED forum, but in my humble opinion SI-2K did a pretty amazing thing by supporting Cineform as their Capture codec rather than make their own. Not only did they get to dump what is no doubt a buttload of R&D but they had a system that had very impressive post support before the camera was even completed! Now this works for SI, but may not work for Dalsa, RED, and the other high end companies. Nevertheless, I think it was a clever move, and I would not be surprised if you saw more of this down the line. As obvious as it seems, I find it interesting to look at why more companies do not do it. One reason could be cost, Cineform is not free, and every time a camera is sold, SI-2k may have to pay something. Perhaps this is not something other camera & capture card companies want to deal with, also they give up a certain amount of control if they rely on a 3rd party to handle the codec needs for their given hardware. The solution here is "size". The company that handles the "Standard Codec" would need to be large enough to insure two things:
1: That they have enough history, and capital to make any company that uses their workflow/codec feel safe.
2: That they do high enough numbers to offer this solution at a relatively low cost to the vendors and/or consumers that use it.

The cost factor is a huge deal right now, and it's the one thing that could keep a standard codec/workflow out of the hands of the masses. That said, if the "standard codec/workflow" began pulling in enough sales, they would be able to get a handle on the cost. My own opinion is that it would need to be around $500 -$800 for the general consumer. To get there would require some serious sales.

This brings up an interesting question: If someone buys an SI-2K do they have to purchase Cineform, or does the codec and all it's bells and whistles come with it. If the camera company, or the software package INCLUDED Cineform, or a Codec like it, then that would get around this problem completely. And get us one step closer to a "codec/workflow" for all.

Think about this for a second: If there were a standard codec supported WELL be all three of the big editors, then the only real choice you would have to make is which PROGRAM you like more. While we've been busy working, AVID lowered it's pricing so that it could compete with the other two. They are all closer than they have ever been in terms of being reachable by the average editor.

Up to now, only Cineform has positioned itself to be the codec for all things editing. Everyone else (For the most part) has attached themselves to a piece of hardware, be it a camera or capture card. Of course there's always Jpeg2000, but that a dinosaur of a codec who's time has long since pasted in terms of speed and efficiency. That said some companies are doing pretty creative things with it. For cameras and the like, Jpeg2000 can be a good thing, but for the speed and multiple generations that editors and various post people need, it's not the best choice out there.

Due to Apple getting lazy, and Adobe and AVID getting busy, the playing field is as level as it's every been between the big three. But the only real solution is going to come from the outside. And it's going to have to work on ALL THREE platforms.

If it does, then it will be a game changer for us, that's for sure.

Jay
 
Great post, Jay. I'm going to digest it a bit before digging in with meaningful response, but great post!

-Thor
 
How many people have really switched to Adobe from FCP for this reason? 20? 50? 100? Even if it's 200 people (way more than the number of regular posters to this forum,) that leaves over 900,800 FCP users, maybe 5000 of whom care, who will most likely wait it out for Apple to increase their R3D support.
 
Drop the codec... use 16 bit TIF.

Drop the codec... use 16 bit TIF.

In other words, we're getting a lot of formats.

No codec will give usefully better images than uncompressed 16bit RGB TIF files, and no filetype with more bits will give noticeably better images in normal use. DPX does not have enough bits to allow for extream image manipulation without getting histogram gaps on 10bit output, also if you only have 10 bits you cannot make transfer curve adjustments in the film recorder, digital projector, or display without putting histogram gaps into the images at the 10 bit display level, or even the 8 bit display level sometimes.

With uncompressed 16bit RGB TIF Log or Video Gamma data you have enough bits to allow for additional adjustments of the color correction and transfer curves along the path the images move through, and readjustment in the future.

I picked 16bit RGB TIF for my DI system's I/O since it is good enough for high quality results and is widely supported by industry infrastructure. The right balance between quality and filesize.

Now that I have my "freeish" system working at a reasonable speed, almost "anyone" can do an uncompressed 0K to 6K+ DI at home on PCs using an "open" file 16bit TIF frame file type for image I/O, and "open" uncompressed WAV for audio.

I have my internal formats, like 11 bits R, 11 bits G, 10 bits B, for archive of CCed files to save space, but those can convert back to TIF later if needed. Log or Linear makes no difference with my system, the images get S curves for film like grading and the output frames can have any curve you like if you reprocess them on a second pass for non video gamma results.

If you ZIP, RAR, LZMA folders of frame files you can also reduce the archive space required by uncompressed frames.

http://en.wikipedia.org/wiki/LZMA

http://en.wikipedia.org/wiki/RAR_(file_format)

http://en.wikipedia.org/wiki/Comparison_of_file_archivers

Double encoding losses from working with compressed data are a real problem for the images, I see yuck everyday now on compressed broadcast.
 
How many people have really switched to Adobe from FCP for this reason? 20? 50? 100? Even if it's 200 people (way more than the number of regular posters to this forum,) that leaves over 900,800 FCP users, maybe 5000 of whom care, who will most likely wait it out for Apple to increase their R3D support.

These things have to be weighed out over time. Adobe's support of Red is a small victory that could have far reaching effects. These battles are won over years, not weeks.

-Thor
 
How many people have really switched to Adobe from FCP for this reason? 20? 50? 100? Even if it's 200 people (way more than the number of regular posters to this forum,) that leaves over 900,800 FCP users, maybe 5000 of whom care, who will most likely wait it out for Apple to increase their R3D support.

I think you missed my point, which means I did not explain it well, sorry. The R3d situation is an EXAMPLE of a larger problem. I do believe large numbers of users are selecting editing programs not based on their features and toolsets, but on what they support.

How many, I cannot say, but even 1 is more than there should be. A codec should not be a deciding factor in selecting your editing program.

What if Avid only worked with DVCPROHD and FCP only worked with BetaSP?

Jay
 
That makes sense. I do think that precedent has shown Apple to be quicker to adopt new technologies in its software and for whatever reason they've been slower with some of this R3D stuff... but knowing Apple's had operatives working at a couple post houses specifically looking into 4K infrastructure for the past year means some of these complaints will soon be just memories...
 
RED has said repeatedly that one of their biggest advantages is REDCODE RAW. This format provides them with not only a competitive advantage, but a technological one as well.

I don't see them abandoning it anytime soon, if ever.


As for a standard format that we can transcode to, there is always a sequence of individual TIFF images if nothing else. When you consider the storage it consumes and the time it takes to write it out to the media, however, you lose the advantage of being able to use a "standard" media such as CF with the camera.

Everything in tech involves some sort of a trade-off.
 
RED has said repeatedly that one of their biggest advantages is REDCODE RAW. This format provides them with not only a competitive advantage, but a technological one as well.

I don't see them abandoning it anytime soon, if ever.
.

Nor do I. Every camera must have some sort of capture codec and REDCODE RAW belongs to RED, and I am guessing Scarlet, and Epic.

The trick here is that this is hardwired to RED. And optimized for shooting (Although that may be changing). A "Standard" would still be needed to line up different sources on to different platforms.

Unless of course it's RED's intention to announce that REDCode RAW will now be for sale and begin to bill itself as a universal Codec.

I don't see that happening any time soon

Jay
 
There was talk in the past of a REDCODE RGB codec. It kinda went along with the RGB shooting modes in the camera that have been dropped. But I can see an RGB wavelet codec from RED as being a good solution for editing and finishing.

In this day and age, for a new codec to gain a foothold in the industry and actually establish itself, it can't be burdened by strict licensing terms or fees. It doesn't necessarily have to be an open format, but software developers need to be given adequate tools for implementing support for it and they need a good reason to do so. R3D REDCODE RAW now has many good reasons for developers to support it and many are doing just that. It's just not an overnight process. The SDK is a huge step in the right direction, but still falls short of what many will ultimately want or even require in terms of developing for REDCODE RAW. I know that companies like Iridas, Discreet, etc.. who already have good RAW tools and debayer algorithms would like to work their magic on the RAW data from R3D files. Hopefully the SDK will open up RAW access at some point soon.
 
Yes, that would solve a lot of problems.

Hope they're listening to your post!

Agree it'd help if Cineform lowered their prices a bit... of course they'd have to crack the Avid ice cube too for me to get totally on board ;)

Bruce Allen
www.boacinema.com
 
Co-dec-pendent

Co-dec-pendent

We have solid examples of modern codecs for clip based work in ProRes, DNxHD and Cineform. Apple and Avid have kept theirs behind the wall and Cineform has to charge something since its not bundled into a profitable product. Uncompressed HD (let alone 4K) and the popular frame based options, DPX and TIFF all carry significant storage expense and require expensive hardware to manage in real time. I feel Jay's frustration, it shouldn't be this hard, not in 2008.

In addition to the desire for a commonly available production industry codec I want to add my fervent request for a delivery codec that would be readily accepted by the major networks. Archiving could still be done to DLT/LTO/etc data tape for bonding/security.

I am not sure RedUser is really the forum for this rant but SMPTE, MPEG, JPEG, NAB or whomever could really make a difference by propagating a universal codec topology as Jay suggests.
 
Adobe has been awfully quiet about cinemaDNG since they announced it last April (link to release PDF) but it would seem that they want an open standard for cine camera workflow and have already shown one that works for still cameras with DNG. While it works well and is an open standard, DNG hasn't taken the still photography world by storm, probably because most of the post work is done by an individual and there isn't a great need for the image to "travel" as it is processed. With cinema, the post workflow usually has to go through several systems, operators and institutions so here a standard is far more important and I cinemaDNG is aimed at being that standard.

The problem I see is that while I believe most software vendors, including AVID, would have no problem building in support for it, Apple is a closed off company that does not play well with others and would probably pretend it doesn't exist or, if marketing demands it, they will say that they are working on it and will support it "real soon now". Apple knows what is best for their customers and that is Apple Quicktime.
 
I agree Clint, but with the choices out there, Apple could find itself writing FCP for iphone only someday.

Avid once thought it knew what was best for it's customers as well.

And look now..

Jay
 
One of the great advantages of Adobe DNG for still photography is that it is an open source format and Adobe has positioned it as a future-proof way to archive digital photos.

It's 20 years from now. Nikon has been assimilated by Google and they have gone through 20 different forms of Nikon Raw. You have upgraded your computer 10 times and your operating system the same number. Do you really think you will be able to open those proprietary, encrypted Nikon NEF files you made way back in 2008? Will Photoshop still have access to the magical software black-box that Nikon gave them in 2008 to turn Nikon Raw into images?

Transcoding proprietary closed formats to lossless, open ones makes sense from a future-access point of view.

So... What is the best future-proof format for moving images? Will cinemaDNG retain all of the information recorded by our 4K, 5K, xK cameras?
 
So... What is the best future-proof format for moving images? Will cinemaDNG retain all of the information recorded by our 4K, 5K, xK cameras?

The DNG format goes up to 32 bit and I would expect cinemaDNG to match that. They also allow metadata to be recorded in a few different standards as well as allowing for custom and proprietary (ick) metadata to be added using "private tags, private IFDs, and/or a private MakerNote".

Adobe looks to covering the bases. There's also word that Red was one of the companies asked in by Adobe so hopefully the guys will help ensure that is a forward looking standard that will play well with whatever is bubbling away in the Red Lab.
 
How many people have really switched to Adobe from FCP for this reason? 20? 50? 100? Even if it's 200 people (way more than the number of regular posters to this forum,) that leaves over 900,800 FCP users, maybe 5000 of whom care, who will most likely wait it out for Apple to increase their R3D support.

Cinewalt,

FCP has about a million users, but Premiere has about 2 million. My source is the copy of Post magazine that I pullled out of my mailbox today.

I don't think Adobe is working with Red because they need the numbers... they're doing it as part of a long-term investment in the future.

Back to your regularly scheduled thread...

-Thor
 
Cinewalt,

FCP has about a million users, but Premiere has about 2 million. My source is the copy of Post magazine that I pullled out of my mailbox today.

I don't think Adobe is working with Red because they need the numbers... they're doing it as part of a long-term investment in the future.

Back to your regularly scheduled thread...

-Thor

Hey Thor,

I think they are doing it because it's what they do. They are the "android" of editing platforms.

:)

Jay
 
Back
Top