RED, are you listening? The ROCKET is the bottleneck....

[William Robinette]

I'm interested in the current state of, and what that means for future of Rocket cards. RED has said T-bolt is a bottle-neck (and I take that to mean it isn't an option they are entertaining). Are they then saying that the only solution is Rocket cards in a breakout box for laptop workflows or is something being worked on in the meantime? I have no "entitlement" expectations for cheap solutions, and I went in knowing, but I will say spending almost 50% of the price of a Scarlet on a PCI card and box is hard to swallow for those of us who shoot mostly but have to post projects every now and again. I do budget the compromise time in, and work with PPro which has been great but I would certainly like to know where the tech is going.

As much fun as the Arri/RED debate is, can we take it to any one of another million threads about the subject and keep this one on track please?

[Fredrik Callinggard]

Interesting conversation to read. Some insults, misunderstandings but mostly interesting. I'd like to say my opinion in this. Main problem is that the red rocket is a little unreliable. I should know. We own 3. That makes it not that healthy for field work. It's possible and we have a workflow for it in the field, but it's risky business. We have one standing at our rental only to make sure that there's a "safe" transcoding at the base. Our rental has assigned DIT's that follow the camera as it goes out and they work as DIT's / camera assistants etc. Hard work for them but it also pays off in what kind of service we provide. For any shoot that uses our cameras we deliver knowledge and transcoding at the end of the day. Now with the economy and how things are most of the time production companies leave the DIT out of the equation and simply look for a data wrangler - or even worse - A LOT of them believes that a DIT IS a data wrangler only. This has affected our business, but we're slowly and steadily growing - as we are one of the only rentals that almost 100% delivers a RED shoot, without any encounters (you wouldn't believe the things we have to save for clients that went elsewhere. To use Epic's, without anyone, with real knowledge of the cameras - DP's and even Ac's are to lazy to learn it in full). Also, post production companies have invested in a lot of other workflows and are either pissed off/ reluctant or simply don't have the space for a RED Rocket for transcoding. That means that they let their machines chew the footage over night or longer and they charge quite a bundle for it.

The post houses I work with knows what they'll get from me from my Epic material on my shoots - they prefer the footage that we deliver to Alexa and therefor don't want me to shoot Alexa - but I have also fought with teeth and nails to get them to be used to it LOL.

The way I see it - is that for this workflow to become cheap/fast with the possibilities and price tag that needs to come down to "affordable". Only happens if it was catered for the data wrangler management. This is why Alexa has the upper hand right now. As the larger portion of "higher end" production in the world that doesn't look for the "highest" form of cinema release and they therefor use the prores codec instead of shooting RAW on the Alexa. This way they "only" (I don't agree myself, but that's how producers sees it) need a data wrangler that can copy and protect the files. That wrangler only needs a laptop, card reader and some storage. The Alexa is simple to use. So DP's and AC's don't have to be bothered to much about following reduser to catch up on all the kinks etc. Meaning very few "fuck ups" can be done.

BUT I do think it's the post house responsibility to provide with the service for whatever camera you shoot on to deliver the best possibilities in workflow when in on-line. That sadly seldom happen with RED footage (outside of my productions). I often see our rental transcode to prores for them to edit in off line, which they then use to grade for final grading as well. The horror... LOL

[jake blackstone]

Sadly, but I see this notion expressed with increasing frequency, that it is always post guys fault, that Red material isn't graded from RAW, they have no clue what they are doing, they are grading from transcoded Prores bla-bla-bla. Why does it always have to be the production agains post? Could it be, that Red's vision of post isn't helping? Unless it's Light Iron, Red wants nothing to do with post. Light iron can afford to buy plenty of RRs. How about Red starting to meet with some other post houses, that are not Red-centric and start asking them questions? What about manufacturers of post gear? I'm sure they would have some ideas.
I could go on...

[Hans von Sonntag]

I shoot Red, I edit Red, I sometimes edit Alexa ProRes files. Alexa make nice pictures, so does Epic. Both need transcoding when best quality is the the target.

Red's raw files have to be transcoded. Alexa's raw is trascode in the camera but mostly with a LogC curve that NO editors wants in hers/his system. The result: both camera files undergo a telecine process and then get transcoded to something meaningful, DNxHD, ProRes you name it.

Bottle-neck RedRocket: I never bought one but shoot red Red since early 2008. The reason: I mainly shoot 4kHD. Half debayered I'm realtime in PP and Smoke, transcoding 1/2 4kHD to 1080p on CPU is pretty quick. Nonetheless all system work better directly off RGB than RGB transcoded to RAW. The RR helps here but does not allow streams faster than 30fps or so or two layers at once. Hence I transcode. The quality loss to full debayer is hard to se for me, actually, IMHO, it's not visible. Personally, I find the software debayer more organic then the stuff from the RR. This and the rapid development in the IT world stopped me to invest in RR.

If I were to transcode hours of footage on a daily base I would think differently. But then the expense of a RR is the least investment.

Hans

[Hans von Sonntag]

Also, post production companies have invested in a lot of other workflows and are either pissed off/ reluctant or simply don't have the space for a RED Rocket for transcoding. That means that they let their machines chew the footage over night or longer and they charge quite a bundle for it.

Yes. This is common practice here in Germany. But, and this a very important point, only a very, very few know how to set up RedCine to squeeze the best possible out of the RAW. 98% use the funky metadata form the shoot. It's sad how much is already screwed up before it sees an colourist. BTW, most colorist I know don't care about grading off the raw. They prefer Arri LogC anyway.

I've been a longtime advocate of releasing a RedCine-Light. A batch-render tool that bypasses metadata, uses only default settings such as 320 ASA, 5000K and lets the user only choose between RedLogFilm or RedGamma3. The latter I would Rename in RedRec709.

That would push the over-all quality of Red based productions significantly. Keep things simple.

Hans

[Kwan Khan]

RED's r3d = two choices, decompress and debayer in CPU or both in the RED Rocket card. There are no opportunities for debayer of the r3d in GPU.

[sergio arguello]

what is confusing is that all this dialogue and nobody from RED is chiming in....zzzzzz I bought a red rocket years ago and in fact owned 5 from that one purchase. Something has to be done asap. Come on guys this is getting rather obvious. We (loyal customers)
deserve this to be addressed in a logical fashion. Does that make sense? too expensive, too unreliable and upgrades unpredictably brick them.

[Evin Grant]

Sadly, but I see this notion expressed with increasing frequency, that it is always post guys fault, that Red material isn't graded from RAW, they have no clue what they are doing, they are grading from transcoded Prores bla-bla-bla. Why does it always have to be the production agains post?

I've had the opposite experience, where producers I've worked with are disuaded from shooting Red because an editor tells them it's more complicated/expensive when they really just don't want to learn a fairly straight forward if not more direct workflow. This compromises the overall quality and future of the project and is the definition of FUD.

Ideally camera, post and the producers should be on the same page, and all parties sould be open to the options that are available. I have not had any of the issues with the Rocket that others have mentioned here. I can spit out pretty much any editorial codec a particular project desires. I can make it fit an Alexa style Prores 444 Log workflow or a DNXhd Avid workflow but really it's becoming clear that Premier native R3d editing will be the future and no transcoding will realy be necessary regardless of camera or codec.

[Michael Cioni]

Unless it's Light Iron, Red wants nothing to do with post. Light iron can afford to buy plenty of RRs. How about Red starting to meet with some other post houses, that are not Red-centric and start asking them questions? What about manufacturers of post gear? I'm sure they would have some ideas.
I could go on...

Hopefully it's clear that the the above statement represents Jake's opinion, even though it's stated as if it were a matter of fact, which it is not.
Jake, please be respectful...and careful.
m

[Johnny Friday]

You'd be totally wrong Johnny. Nice way to turn this into a personal attack, rather than trying to examine why some post people may have a problem in unnecessary spending $5k.
In case you care, my system does debayer in background and I always work with full resolution in real time without RR.
They are other grading systems, beside Resolve you know...
Unfortunately, for majority of users, Resolve is it and without RR it's 1/4 resolution...

Sorry Jake, it was meant to lighten up this thread.....a Joke....Personal attack? What?