ALEXA vs RED thread in Creative Cow

[Jeff Kilgroe]

But Red not releasing an SDK for two years was certainly a stiff arm, IMHO. Perhaps this was necessary to get Apple on board.

Two years from what point? I think your timeline is screwed up. Public SDK was delivered 10 months after the first cameras. I turned in my SDK agreement to RED on August 1, 2008. Major industry players who wanted to be on board had earlier access than that and also had a hand in shaping the SDK.

But it's even more than that. It's the tenor of the early Red presentations. Avid did just not appear relevant to them.

RED showcased their two supporters -- Apple and Assimilate. I think the real question here is why RED did not appear relevant to Avid? Avid could have been part of RED's presentations too, if they had jumped on board.

I don't mean to keep dragging this on either, and we see this with companies other than RED, Avid, etc... However, when software X doesn't support Hardware Y, it's often looked at to be Y's fault, even though it's up to X to write their own software and support Y.

Apple signed on early, but overall support from Apple has been very lacking.

Of the major software players, Adobe is the one that's really stepped up. They waited for the public SDK -- being part of the frantic development effort before that point wasn't in their best interest I would assume. They released native R3D support for CS4 in December 2008, just 4 months after RED SDK release... Where was Avid? I'm not sure where they were, but in January '09 Meryem Ersoz and I did a 72-hour short film contest with two RED Ones shooting an overly-ambitious film for the project. Erik Rangel and I felt this was a good opportunity to test the CS4 workflow under a deadline and we cut it entirely on a 15" 2.33GHz C2D Macbook Pro with Adobe CS4 and delivered it before the deadline. No transcoding. No proxies. Native R3D on the timeline, shot 4K, delivered in 2K for screening, screened in LA. Sure, it wasn't without glitches, but this was in January 16-19 '09, almost 2 full years ago, built on the publicly available SDK.

Sorry to pick at you, but you're demonizing RED here and re-writing history on your own terms.

 

[M Most]

Sorry to pick at you, but you're demonizing RED here and re-writing history on your own terms.

So are you.

Comparing Adobe's stance to Avid's is a bit ludicrous. Premiere Pro had virtually no presence in the broadcast or professional editing market that has been Avid's bread and butter for years. Adobe has long been looking for a niche for its desktop editing product, and perhaps it saw an opportunity. Avid had a much more mature, highly developed system, and integrating a completely foreign SDK into that is not the easy task that both you and Red at the time seem to think it is. The bottom line is that Adobe was searching for a market (especially among the hobbyists that were the core audience for Premiere over the years), Avid was not. There is little to no commonality between the two, especially at the time you're referring to.

 

[Jeff Kilgroe]

The last time I checked, you don't work for either Red or Avid, and you likely have no more solid inside knowledge of any of this than the poster you were replying to.

You're correct, I don't work for either. But beyond that I won't discuss any relationships I may have. My whole point here is that some people have their timelines skewed and often point their finger one way for no reason other than industry FUD guiding them.

I've heard lots of FUD and rumors over the last couple years regarding RED workflow and industry support. But regardless of internal politics or technical issues, Avid found a way to incorporate H.264 support in Media Composer almost overnight, licensing and all, and yet from the point of release for a public R3D SDK, it took them nearly 2 years. I'm not going to speculate as to why... That's just the way it happened.

 

[Peter Moretti]

OMG, did you guys even READ my posts? My first post on this subject was as respectful as I think is possible. My second post states that this was concerning the early days. And my third post states that Red's lack of Avid support may indeed be Avid's fault and have been necessary on Red's part to bring Apple on board.

I'm not "demonizing" or trying to "villianize" Red. Really.

 

[Peter Moretti]

... But regardless of internal politics or technical issues, Avid found a way to incorporate H.264 support in Media Composer almost overnight, licensing and all, and yet from the point of release for a public R3D SDK, it took them nearly 2 years. I'm not going to speculate as to why... That's just the way it happened.

Native H.264 support has just been added to MC in version 5 released in June of this year (the same release that added native .R3D support). And those H.264 files have to be wrapped in QT inorder to work.

 

[Michael Cioni]

Post Problem Solving

Post Problem Solving

I was at Kappa Studios when Red was first being introduced. And it was clear as day (for about two years) that workflow was absolutely disregarded as far as Avid users were concerned.

While RED had to start somewhere, (and Apple being one of the earliest adopters of R3D support) I whole-heartedly believe and have kept this belief since September of 2007 that the main problem with regard to RED post production is the post production community itself.
While it seems as though RED didn't have a cut-and-dry solution for AVID users, I have observed and published numerous solutions for AVID that were as effective in 2007 and 2008 as they are in 2010. It has pained me to see this happen, as I carry the utmost respect for my colleagues in post, but on the whole, the post production community has done a lousy job of implementing progressive philosophies, techniques and tools to support R3D for all platforms since RED announced what they were planning in 2006. To be perfectly blunt, I don't believe this is a fault of RED. It's safe to assume RED assumed that the post production community and 3rd party manufacturers would change as fast as they did (ie: Apple, Assimilate, and Quantel (amongst others)). I have never blamed RED for post production support because I think it's been there since day 1 - it's just such a unique workflow that most post production companies chose not evolve enough to support R3Ds in an ideal way (this is not true with all companies, but it is true with the overwhelming majority).

In early 2008, when we did "The Informant" for Warner Bros, we created a way to marry the workflows of producing HD tapes that carried meta-data through an ALE so that even a tape capture from RED files could be re-linked perfectly back to R3Ds. I used this RED-to-tape workflow in 2007 and 2008 on numerous jobs since that is what the Media Composer was most comfortable with. And that, in essence, is not only the job of not the post lab, but the filmmaking community at large: to problem solve and produce, evolve, and create fast and easy workflows that satisfy any type of user. Since then we all agree the support for AVID and other platforms has been made easier - yet many post infrastructures still struggle due to the philosophical differences between RED and the competition. I discuss this with my team on a daily basis and the discussion usually resolves with the same conclusion:
Typical HD RGB capture of any form is completely inferior to RED RAW image capture, yet RGB capture is something the community does not have difficulty implementing, hence the reputation of RED in post. But again, does this concept mean that RED is problem? Or is the problem with the minds, companies and agendas behind infrastructures?

Mike Most is correct, nothing changes overnight and companies do need to act within their own best interests. But as Most has proven with his words and actions over the past 3 years that he and some other companies (Kappa included) have recognized what is in our best interest is to change and embrace emerging technologies that are superior to traditional ones.
85% of my clients are AVID-based. I have not had issue serving them since the beginning - and I have enjoyed the problem-solving and inventive opportunities to make the superior format of R3D meld with pre-existing or even inferior tools. To be fair, that is one of my full-time jobs and I realize that is not everyone's job on this forum

But one thing that I've been expressing publicly for years is that to some degree, problem solving and forward-thinking is all of our jobs...and that is easily revealed-often on this forum every day. If we can all accept that we each are responsible for some small portion of progressive leadership and in many cases re-education of existing talent, I think the future will show that investments made in supporting RED are, in fact, in every filmmaker's (and therefore every post company's) best interest.

Here is a very simple, watered-down algorithm (and I admit, this is very simply put) that I've experienced are the 3 main components to supporting RED efficiently:

1. 33% pre-existing knowledge and experience
2. 33% new hardware and software tools & techniques
3. 33% philosophical change

If a company or person lacks one of these key components or one is grossly out of balance with the others, the implementation of advanced tools such as RED becomes far more difficult. But when they are in balance with each other, the solutions and results flow effortlessly amongst a team of people-even when they don't know each other. One thing is for sure, some groups tattoo to their clients the notion that RED is easy to shoot and hard to post. Others swear it's the fastest, cheapest, and best workflow they've ever experienced.
Well, we can't both be right...

 

[Meryem Ersoz]

Red wasn't necessarily the open, loving, caring and sharing entity you're painting them as, and Avid has been a major force in the industry for a long time, with a product that works a certain way and relies on technology that has been battle tested and proven for years. Integrating new things is not something that happens overnight, particularly when you're not necessarily privy to everything you need to do it the way you want to.

Exactly the point. Avid has been conservative in their response to the shifting tides of change. They weren't super speedy on the uptake, exactly as one would expect.

For that matter, neither was Apple, and they were an early partner - in fact, they proved the least adaptable. Assimilate adapted quickly, and they made a lot of sales based on early adaptation. Now Avid is on board, and I'm sure they are making sales, based on this adaptation.

On a relative basis, Avid was fairly quick to integrate the RED workflow.

On a relative basis, that is. Which translates as "faster than Apple" -

Of course RED has interests to protect - every company does. I think you mis-read Jeff's characterization of RED - I don't think he ever said loving, caring, or sharing. In fact, we have all been witness to the blunders along the way - I don't know any other camera company or post-production company, for that matter, with this open of a process, do you?

Most of those blunders have been corrected and RED continues to refine their process and their relationships. As does Avid with RED. That's what is important.

 

[Peter Moretti]

Michael,

I of course appreciate your perspective as Light Iron is on the cutting edge of all this. I do feel my initial post is being turned by some people here into a fighting a battle that's been over. Red and Avid seem pretty tight now and that's what's important.

 

[Gunleik Groven]

Allmost forgot, as I am rarely in Avid land, but back in april - august 2008, I set up Avid workflows for a string of commercials. I have never been an Avid user, but even then I managed to set up workflows and matchback to the raw files for online based on an Avid edit. There were issues, but easily solveable. The biggest was that I was told symphony supported dpx, but I had to re-run the material as tiffs... No rockets back then...

 

[M Most]

But as Most has proven with his words and actions over the past 3 years that he and some other companies (Kappa included) have recognized what is in our best interest is to change and embrace emerging technologies that are superior to traditional ones.

There are some very good reasons why I work for the company that I do (Next
Element by Deluxe). That's one of the primary ones.

 

[David Battistella]

Guys. I am following this closely and what seems most important is the way things are right now. The AVID stuff is old, it doesn't matter much to how things are now.

RED commands attention.

next.

Don't pound on each other over what was, blah blah three years ago.


things change. change is good.
David

 

[Gavin Greenwalt]

The only thing RED has ever restricted is people using their own debayer algorithms and direct file-level access. It seemed odd at first, but after seeing a few of the first third-party attempts for these things, I totally understand why RED did this.

I disagree. I can think of no sensible reason to obfuscate the metadata for instance.

Also if the Bayer stream was open you could do things like generate rough thumbnails in apps without loading the SDK. Even if it was monochrome there are a number of applications to a more open file-level access.

And we only got trim access... 2 months ago? (FYI my trim tool's trimmed r3ds still work. Just saying.)

 

[Rainer Fritz]

"With both EPIC and Scarlet, you can record RAW to CF or SSD AND record 720P, 1080P or 2K RGB (your choice of flavors) out the HD-SDI. There are options to do that now from 3rd parties and we will have our own soon. That gives you the best of both worlds if you need it instantly. "

Means that we get a small, handy dual link hd-sdi recorder with a SSD slot??? which can record up to 2k RGB or RED compression...??? omg I see a lot of Alexas and Sony F-3 out there with a RED DVR on it..... that would be funny.... something like the cinedeck... only smaller....

btw. we shot with beginning of april 2008 a big feature film and edited on avid without problems... mediacomposer 3 on mac and imported the camera proxies with ALE... worked like a charme....

 

[Rob Lohman]

I disagree. I can think of no sensible reason to obfuscate the metadata for instance.

Obfuscate metadata?

Also if the Bayer stream was open you could do things like generate rough thumbnails in apps without loading the SDK.

And this would be faster than letting the SDK do it.... how? This is exactly what the SDK does when you ask for a lower resolution decode.

And we only got trim access... 2 months ago?

Unfortunately that one took a long time indeed. We're trying really really hard to get the tools out that everyone wants, but there are only 24 hours in a day.

Hacking 3rd party trim tools together creates some new problems....

(FYI my trim tool's trimmed r3ds still work. Just saying.)

Except that it'll break a new feature introduced in the just released SDK

 

[Matt Fleming]

Obfuscate metadata?

Hey Rob,

Obfuscate may or may not be right word but perhaps you can clarify things. Is the lens metadata available (as in, anyone with the SDK can freely access and use the data)? Note: I'm not asking if it is "present"...it's been confirmed by Deanan and Stuart that the lens metadata is in the R3D files...the issue is if we are given access to it. It's been suggested that this data has been hidden in some way. Perhaps that is totally wrong but would love for you or any one at RED to please clarify this. Thanks for ANY information regarding this!

Best,
-Matt

 

[Aaron Newsome]

Hey Rob,

Obfuscate may or may not be right word but perhaps you can clarify things. Is the lens metadata available (as in, anyone with the SDK can freely access and use the data)? Note: I'm not asking if it is "present"...it's been confirmed by Deanan and Stuart that the lens metadata is in the R3D files...the issue is if we are given access to it. It's been suggested that this data has been hidden in some way. Perhaps that is totally wrong but would love for you or any one at RED to please clarify this. Thanks for ANY information regarding this!

Best,
-Matt

Wouldn't the SDK allow you to access the metadata in a straightforward way?

 

[Matt Fleming]

Wouldn't the SDK allow you to access the metadata in a straightforward way?

No, not as I understand it. That has been the assumed ability and high hope for many of us, but that's apparently not the case. The lens metadata is "present" in the R3D files, but it is not accessible for some reason. There have been a lot of posts asking for clarity on this but there's not yet been any answer from RED.

On the bright side, there are /i pins on the Epic PL mount! :smiley: There's hope accessible lens metadata in R3D files may someday become a reality!

 

[Rob Lohman]

S4/i compatible lens data is indeed in the R3D files but not yet accessible through the SDK.

It will be in the next big non-beta SDK release.

 

[Matt Fleming]

Hey Rob,

It's great to have official, RED feedback about this topic. Many thanks for taking the time, it's very much appreciated as this is something very important to me.

Any rough ETA on when thenext big non-beta SDK release will be (when lens metadata will be available)?