Thoughts on RED and Color

[Ben Brainerd]

So I know there's the big thread about trying to get the Color plugin from Apple, and a lot of waiting for RED to produce something. Note that I'm not complaining here by any stretch of the imagination. I'm... pontificating, I guess.

I've been thinking a lot about the Color plugin, the various workflows that we've been using to get around the new color science, and the fact that other tools (BaseLight, Scratch, etc) have the new toys and we don't. One of the things that struck me was the fact that these are *other companies*. Places with internal development teams that do nothing but add hot new features to their product, a narrow product line, and management that supports said development.

Let's be honest. Apple's primary market isn't professional video. It's more or less at the bottom of their product tree. And there have been some disturbing rumors about the future of that product line. All that said, is it really all that surprising that support for new video tech isn't really all that forthcoming from Apple?

RED, on the other hand, lives and dies by their product. And I love pretty much everything they do. But they're not as big as Apple. And I imagine their software development team is significantly smaller. So they're busy. Something like RedCine-X pops up, and that's the development focus. I'm ok with that, since the result is pretty amazing.

So I had an idea. I know it's not really feasible, for a number of reasons, but I'll feel better if I get it off my chest and out into the wild.

Why not outsource, or, (god forbid) open-source the Color Plugin? Take the development burden off the team at RED and let them keep working on the hot new toys. Tie the plugin firmly to the SDK, like all the other post solutions. Color science gets updated in the SDK? We don't have to clamor at Apple and RED for a new plugin. The outside development team will take care of it. (We'd have to clamor at them, but it'd be better than trying to get a response from Apple. :P)

I see two major problems with this idea of mine.
1. Apple's response. Whatever architecture/hooks were used to build the Color plugin aren't exactly documented and open. So I'm guessing open-source would be out of the question.

2. RED's response. Again with the undocumented Color hooks. Also, I don't know how much of the Color framework is SDK based, and how much is proprietary.

There's also the issue of finding coders. It's a project I'd be willing to work on, but I doubt it'd be a one-man show, unless both the hooks and the SDK are a lot simpler than I imagine. And finding people to work for free on it might be tough. Then again, I would willingly pay a reasonable amount of money for a reliably updated plugin rather than having to wait on RED.

So that's my thought. Discuss, if you'd like.

[Neil W. Smith]

Ben,

Your raise some valid points and in an ideal world we would all implement your suggestions .... however, in the ISV software world (ISV - Independent Software Vendor) its a different proposition .... everyone is looking for competitive advantage and a way to protect their wetware (IPR) ... as soon as you publish the actual code and not the APIs you lose control of your IPR and platform .... maintaining published code is much harder than writing code that works for one iteration of the firmware ..... this is the dilemma that companies like RED face .... if you have an Open Source approach and you then change the underlying code structure, how do you make sure all the ISV devs out there implement the changes quickly ... SDKs are a way to to implement some kind of conformity but at the expense of innovation and swift change of direction.

When we were preparing for the Jan 16th RED Day and started to deal with wider DR options of the new M-X RAW file format it became obvious to us that we had to go directly to the .r3d files to extract maximum tonal values and lower noise floor .... unlike the HeavyIron guys who have spent millions of dollars on expensive hardware we were able to get a plug-in from Deanon and drop it into a beta version of some software from a certain Cupertino company and derive maximum color fidelity from our M-X footage using desktop applications.

We had fifty members of the ASC evaluate our M-X test footage last Saturday (23rd) in a tightly calibrated DI environment .... everyone of the members was impressed with the increased DR and reduced noise-level of the M-X sensor - no one expressed any doubts on how much better the new M-X sensor is ... however, when the discussion turned to workflow and color correction everyone assumed we were using an expensive HeavyIron DPX dependent workflow .... when we explained we were grading everything from the RAW files in Apple Color they were somewhat surprised (pleasantly) ... when we showed the members the film-out we did from Apple Color to HDCAM-SR (no DPXs involved) at Efilm they were amazed.

In this Moore's Law driven industry of Digital Filmmaking, the desktop applications providers will win the war .... volume matters (as RED have proved) ... let RED and Apple continue their plug-in development and we will all benefit .... let the HeavyIron solution providers continue to charge millions for their platforms and lets see how this all ends.

Neil

[Ben Holmes]

Whilst I don't want to appear to be an Apple apologist, the "Apple is just an iPod maker now, that doesn't care anymore about pro users" meme is so commonly expressed - and lies at the heart of your basis for a new approach. As such, I think it's worth attacking this premise.

Apple is selling record numbers of Macs. Apple is a hardware company. Just with the iPod and iTunes, it does not make its money on software - it makes it on the iron. To this end, it uses software as a differentiator to sell premium priced hardware. Look at the recent release of Snow Leopard - given away practically by the terms of OS pricing.

Final Cut Studio is priced on the same basis. Color is a great example. Buy software for the suite that used to sell for thousands, and give it away 'free' with the suite. Price the whole package so that multiple licenses are a throw-away price for facilities. Then sell a dozen Mac Pros to that same facility. Pro Apps are a huge part of selling Mac hardware - be it Logic or FCS. They sell iMacs, Mac Pros and Macbook Pros in ever increasing quantities. FCP is now the most used editing software in the world. Do you really think Apple is not interested in keeping it ahead?

So why does this myth persist? In the main, I think it comes down to the slow pace of recent development. Even though there was a major release in 2009. What explains it? Firstly, Apple has had two major transitions since releasing FCP - OSX (I still remember booting OS9 to run FCP) and Intel hardware. Add another switch to 64-bit. This means that an overhaul of the suite means a complete code rewrite, which takes years. We all know this is coming eventually. Which brings me to the second reason: Apple is about stability and reliability. They do not do beta releases or beta programs - barring private ones in limited quantities that maintain Apple's (highly successful) competitive advantage. Whenever they have rushes an update to market they have been roundly condemned by pro users for making the software less stable.

When you think about it, it's pretty remarkable that every FCP update brings stability. FCP7 had only a couple of minor issues (like losing render files). Everyone who took the plunge had no real problems. And Apple manage this without a public beta program. I'd take that over innovation most days - FCP still basically does the same job (and makes me money) every day that it did in 2001. I want more stable, not more flashy. I made a business out of working alongside AVID systems

RED have the opposite ethos - get the software out there. Warn people it's a beta. Leverage the user base to get the kinks ironed out - this makes a lot of sense for a small company on the cutting edge. Who cares if Redcine spent 2 years crashing on larger exports, right?

This is what makes the RED/Apple partnership so interesting - Apple love to associate themselves with the cutting edge, but they have a much wider user base to consider. They don't publicly release beta updates. But when they DO integrate RED in Color - well, it pretty much works.

RED need to stand still for a few minutes so Apple can catch up. I have no doubt they will - just as I have no doubt that RED know a lot more about the Apple product pipeline than we do. Do I want the New Color Science in Color. Hell yeah! Do I want it stable and well integrated. Oh yes. I also want FCP to leverage all 16 virtual core in my Macs. And I want a big bag of cash and a Ferrari.

Apple employ hundreds of pro apps developers, keeping Apple hardware on the buying lists of professionals year after year. No one is losing focus on that aim - to sell more and more Macs. Sure, as a percentage of Apple's business, Mac sales have declined. But in real terms it's still growing. And finally - do you know how many of them can run FCP? All of them (barring screen resolution issues for Macbooks on Color - don't let that spoil a punchline). Try saying that back in the 'glory days' of 2001.

All the best.

Ben.

[Cüneyt Kaya]

i dont believe in the apple doesnt release beta stuff, actually without clipfinder, the apple workflow is a headache...and what about long timelines in Color, doesnt work.

i wait for the next update of smoke, then i switch.
i am getting tired of the you have to use r-x,fcp,clipfinder,color workflow for daily work...i believe there should be no offline/online workflow...one in all...and if you need special programs, then use them for the creative goal you are after, but not just for the sake of having a working pipeline.

all in one box for 90% of the work...this is what i want in 2010.

[Ben Brainerd]

Well reasoned discussion on RedUser? What's the world coming to? :)

Lots of good points to address here.
Neil:
I didn't intend to imply that RED should Open Source their entire codebase. That would be the wrong course of action, for the reasons you described. I guess I was suggesting that, if the plugin development were to be outsourced, the hooks into Color/the current plugin source might get released, to serve as a platform for tying the plugin into the SDK.

I agree wholeheartedly with your last point regarding desktop apps vs heavy iron. In fact, that's really where all this came from. As a starting post professional with a focus on color, I've invested what is to me a lot of money into building a fairly full-featured post workstation. And that would never have been possible were I forced to buy heavy iron.

It's just frustrating to have built this system, gone out and acquired a client-base built on the principle of doing RED post, and then have it not really work as advertised. I can't just L&t, edit and grade anymore. Not and have access to the best possible images. I've got to jump through a number of hoops to get there. And it's annoying to have to jump through said hoops when I *know* the software is out there, based on the fact that via the SDK it's already built into the heavy iron systems.

Hearing about your experiences with RED Day prep gives me some hope. It also makes me kinda wish that the beta plugin experience was a bit more spread out. Maybe not a full open beta, but maybe a sign-up beta? Dunno.

Ben:
I wasn't trying to base my argument on the "Apple makes iPods" thing, so much as some more specific items. Losing the Expresscard slot on the new MBP. The Shake development meltdown. Rumors (and yeah, we all know about Apple rumors) about the development direction of the pro video apps.

Don't get me wrong, I like FCP et al. For most of the reasons you list. (Stability, etc) But I also think Apple's culture of secrecy causes me some headaches. It's like someone within Apple is saying "I know what the pros want" when in reality, they don't. How long have we been clamoring for a better media manager? Other developers at least open up their roadmaps, so we know what they're working on, where they're going, what's coming next. Apple? They could be riding unicorns around Cupertino for all we know. :)

It seems that, to a limited amount, this culture has affected RED too. Sure, we hear a lot about the new sensors/cameras. But with certain notable exceptions (RedCine-X) we don't hear anything about software. For more than six months (I remember expecting the new color science in the FCS3 plugin) we've been waiting for an updated Color plugin. And we have heard *NOTHING* until like the past week. Now, I've been part of a couple product launches and some coding teams, so I know how dangerous too much information can be, especially if you miss your target date. I'm not necessarily asking for "This software will be released in a 3 weeks" so much as "Hey, we're working on it". And we didn't even get that.

I guess what concerns me is that they'll release the new plugin, which they've stated won't have FLUT and some of the other new stuff like RedGamma/RedColor, and then we'll have to wait *another six+ months* for a new one. There are a lot of us who can't afford to move up to the heavier iron that's supported by the SDK and external dev teams. And to be honest, even if I could, I wouldn't want to. I like FCP better than Avid or Premiere. And the pricepoint difference between Color and its nearest competitor (Scratch, I think) is pretty significant.

[Nick Shaw]

Neil, are you able to say whether the Beta plug-in you had from Deanan supported the new new colour science, FLUT and all? If so I guess there is hope it might be released to the rest of us some time soon.

[Ben Brainerd]

Rob Lohman is on record as saying the Color plugin that we'll be getting won't have FLUT™. http://reduser.net/forum/showpost.ph...1&postcount=32

EDIT - So RedUser automatically TMs FLUT. That's kinda cool.

[David Battistella]

We had fifty members of the ASC evaluate our M-X test footage last Saturday (23rd) in a tightly calibrated DI environment .... everyone of the members was impressed with the increased DR and reduced noise-level of the M-X sensor - no one expressed any doubts on how much better the new M-X sensor is ... however, when the discussion turned to workflow and color correction everyone assumed we were using an expensive HeavyIron DPX dependent workflow .... when we explained we were grading everything from the RAW files in Apple Color they were somewhat surprised (pleasantly) ... when we showed the members the film-out we did from Apple Color to HDCAM-SR (no DPXs involved) at Efilm they were amazed.

Neil

THANK YOU THANK YOU THANK YOU!

Thank you for this Neil. I have been using a very effective Apple Color workflow for MONTHS but people still seem to be non believers because Apple Color has been crapped on by many "pros".

A properly set up and calibrated Color suite is a very nice tool in the RED arsenal, that's why we want to ensure continued, excellent support of this product.

To me COLOR is a very powerful tool and could be to grading film what desktop publishing was to print many years ago.

David

[David Battistella]

i am getting tired of the you have to use r-x,fcp,clipfinder,color workflow for daily work...i believe there should be no offline/online workflow...one in all...

all in one box for 90% of the work...this is what i want in 2010.

I think that is becoming less likely with the release of REDCODE 42 and REDCODE 100. These codecs simply will be LESS real time than the existing ones. The RT debayer would be nice, but you really have to use rendered media over proxies and it will be that way for a while.

David

[Charles Angus]

i believe there should be no offline/online workflow...one in all...and if you need special programs, then use them for the creative goal you are after, but not just for the sake of having a working pipeline.

Until the day that we can edit native footage (whether that be 10-bit DPX, RedRAW, ArriRAW, etc.) on a high-end consumer PC, there will be a place for the offline-online workflow.

Really, there's nothing wrong with it.