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 :)

What's the future of Red post?

My understanding is that they do access the R3D's direct but as RGB, not debayered RAW - I don't think Red have opened that functionality to anyone yet - except maybe Assimilate.

You are right, I forgot that "SDK gives access to RGB" part and not to RAW debayer directly.:nerd:

Patrick
 
Come on Stuart, don't get all Rumsfeld on us. You know what we've been asking for. Deanan's response was what I wanted to hear. I haven't heard anything about red post for months and that's not due to a lack of asking. Hardware accellerated debayer sounds great, but I didn't expect that. Cross program/camera LUT's, and audio in redcine are at the top of the list. I am able to do the transcode, thanks Mike Prevette, but I don't do post work. I'm a DP and I do just fine at that. It's clients that have problems even after I've shown them what to do. They find weird bugs and get black video files. Why is it so hard for people to understand that there is a learning curve and not everybody has the time or desire to figure out all the programs and the proper workflow. If you know how to work with Red and have posted 2 million projects with it, great for you tiger, but don't let that cloud your judgement and don't start talking shit.
 
Hi Folks,

Funny to see how a lot of RED-Users posting judgments and solutions here can't even see how different problems are in different sections of production (as in short projects, industrial films, TV drama, Feature Film). Most just shout out what they see from their own little perspective...

We find the reality of RED-One camera rental is that you have to compete with HDCam rentals and workflows for tighter and tighter budgets in the TV Drama section as your daily bread - Those who could shoot 35mm with their budget will more likely find a "workaround" for post issues - where as for TV we find ourselves forced to throw in "Renderpower" for free to at least stand a chance for someone to change over from HD. Its not so much the machine power but the time it takes to get stuff through into edit that makes many producers turn their back after lengthy quoting and supervision sessions.
I agree that post is where it fails most...
What might be easy for me to decide for my own project (how i do it and with what) is not easy for me to decide for my customers.
just my 2cent
best of luck
 
linux support for redline is nice and kind of huge. As far as crashing in red rushes most of the problems I've had have been with the interface, its nothing that can't be scripted out and be put into a traditional render manager, into an optomized non fragmented DPX workflow.

What's exciting is that with a linux version , you could quickly scale up to several time faster than realtime, and in a proper data environment realize all the benefits there of, and it could be done quite cheaply with iscsi. a small blade farm feeding into an rdma drive array.

Also I hate to put it this way because I rather like mac's but on the file organization front they're running far far behind NTFS, and some of the other linux file systems. I mean long clips having problems with system cache ? Thats an OSX issue. There's a reason most major renderfarms use linux, why the red's internal system ware is linux and why most on set devices are linux based.

Hmm stuart if you do get the linux version tuned anychance of you putting out a distro that you like ?
 
I mean long clips having problems with system cache ? Thats an OSX issue. There's a reason most major renderfarms use linux, why the red's internal system ware is linux and why most on set devices are linux based.

Hmm stuart if you do get the linux version tuned anychance of you putting out a distro that you like ?

The system cache issue turned out to have a similar problem on windows too but still retarded nonetheless. It's always surprising to find problems that are deep in the OS. The advantage to OSX over linux is you get real access to the full QT stuff (as painful as doing QT is on mac even) but at least doing stuff like writing timecode tracks exists.

red distro would be cool but no time. We would like to know what distro's people are using. So far most feedback is either RHEL 4x or SLES10.
 
in the view that we shoot digital 35mm format in 4k and i have my laboratory in my 8core mac, that's not bad at all. and when i think, that i have my 1080p!!!!! rushes of one hour footage lately 3 hours later, also very nice at all. to give out now selectable the avid codecs was also very helpfull. to speed up the import of them into avid would be a big wish. to put more functions into redrushes will be great, but i heard deanan is on it. and by the way thanks..... sometimes i think your keyboard must burn :)
and then show me your laboratory which process your 6 rolls and scan them in 1080p in 3 hours at the price of running my 8core for that time......
when edit is finished i take the edl, connect the drives and render with monkey the dpxs, as my lab it wants for their baselight 8.
if you need a faster workflow just buy a second or a third 8core...... :)

must go to sleep......

-> EDIT: A GPU ACCELERATION WILL BE GREAT !!!!!!!!!!
 
We would like to know what distro's people are using. So far most feedback is either RHEL 4x or SLES10.

No RED ONE here, but I'm looking at Scarlet. I use RHEL5 x64 and Fedora Core 9 x64, but I also have a Windows box and a Mac.
 
in the view that we shoot digital 35mm format in 4k and i have my laboratory in my 8core mac, that's not bad at all. and when i think, that i have my 1080p!!!!! rushes of one hour footage lately 3 hours later, also very nice at all. to give out now selectable the avid codecs was also very helpfull. to speed up the import of them into avid would be a big wish.

I believe that import into avid is now pretty much at the speed of the copy.
At least that what users have said. Michael?
 
Lunacie...your post is insulting. Please list the solutions you have for the problems I already listed on earlier pages in this thread. I'll give you a dollar if you can come up with them.

There are bugs. Of course there are work-arounds. But in our case these were extremely lengthy.

So...just because you have some things working (short clips probably), does not mean we're inept at getting a workflow going.

Please be respectful. There are absolutely still bugs and issues with RED workflows. They are working on correcting them.


Oh, I won't give you a dollar if you say spend $25,000 more dollars for Scratch. Nope...that is not the solution I'm talking about.

Slow down there. Never once did I address you or any of your posts, or mention the word inept.

I was responding to SFGeek's post that started this thread.

There are solid, proven Red post solutions that work now. You, personally, don't have to like them, don't have to pay for them, can even deny that they exist, but it doesn't change the fact that I have one I use daily... yes, it's Scratch - but what, exactly, is wrong with choosing a tool that actually works today?

I know of several other facilities in LA that have their own workflows, some use Scratch, several don't.

The Red Reel is filled with projects that were successfully posted, and there is a long list of productions here in LA and around the world that are finding a way to post longform material shot on Red.

You don't have to spend $25,000 more for Scratch, but it is the best tool for the job right now. Other tools are becoming available, many of which cost significantly more than Scratch.

I have successfully conformed & graded three Red based feature films, several episodes of 1hr tv episodic, several commercials, music videos, and short features. I can assure you that my pipeline can do more than just short clips.

It's quite valid to let Red know that something isn't working, and to encourage them to develop stable software to be distributed to camera owners and others who can make use of it, but that does not change the simple fact that there are proven Red post workflows available today in use at post houses and in production pipelines right now.

JT
 
If there were no post solutions, no on would be using the camera. And I don't think Scratch has sold thousands of seats to redusers because that's the only way to do post.
 
Perhaps my words were harsh, but I am getting a little tired of people defending the post solutions RED has put out, just because they have something working in their setup. I am NOT talking about Scratch...that is NOT a RED solution...it is 3rd party. It is also not something everyone wants to get into or can afford to get into.

We are a non-L.A., all Mac based, small production company. After the huge investment in RED, lenses and equipment...I may eventually go with a Scratch-based solution. But you and I both know RED is absolutely trying to deliver solutions that are non-Scratch based. It is this reason why some of us continue to mention the problems and lean a bit on RED to put some efforts into correcting them. If you have Scratch, great. No need to post. We all know Scratch works brilliantly...

So, I stand by my push for solutions that are already in place...but buggy and not quite up to par. I do think RED is on them and hope posts like this will disappear soon.

But there is no need for "I still can't figure out why people think posting Red is so difficult." Unless you have Scratch right now, there are absolutely difficulties and challenges in posting RED using RED's tools...4k RAW is an entirely new game in the post world, RED has no manuals, RED has app's out that are still buggy...THAT is why people think posting RED is so difficult.

Okay, off my soapbox. Sorry if I overreacted...I just want to make sure these apps work better. I would gladly wait another year for 28K resolution if it meant getting 4K working awesome now. If they only work some of the time or for some types of projects...they are not ready for prime time or professionals.

Kevin
 
Deanan,
As far as Distros obviously a 64 bit variant would be prefered as regarding the recent announcement we're going to need it for some of the larger sensor sizes as well as the higher variants of redcode. In regards to what I've been seeing in renderfarms, generally CentOS (redhat) is liked, as well as Suse. I mean a 261 MP bayer image....will need a renderfarm.

Thank god USB 3 will be available when you're coming out with that.
 
My take on the Red software situation is that they will definitely put out fantastic, stable software - I know the coders are more than capable of it. But it is a small group, and they have a lot on their plate, so as we have seen, it takes time.

I really like RedCine, for instance. And Redline is the ticket for massive, greater-than-realtime transcoding until GPU based debayering is possible with Redcode.

I also understand Red's inclination to proceed carefully with the SDK and allowing vendors access to Redcode, but again, I think it'll open up in time, and solutions will be forthcoming.

My money, in the short term for ultra-low-budget is Adobe/Red. Adobe's openness to outside development is what puts it in a different category than FCP, where one is completely at the mercy of Apple's decisions and schedules, which don't always coincide with the best interests of the pro-app market.


cheers,

JT
 
Kevin-

Try ClipFinder, it is a 3rd party ap (I know you want better red aps) that is free. It uses redline to render, and redalert to color.

You can render all of your footage to an offline format and cut away happily in FCP, then when you are done export the xml of the sequence and import it into clipfinder. You can match the clips back to the original R3Ds and do a decent color in redalert (natively in clipfinder soon, if not already there) Then render your final online, and correct in color or FCP depending on your time constraints. If you have a 1080 finish this is a very easy and flawless workflow, if you go to dpx there is a little more you need to do (not with clipfinder) but we have no real need for that unless we are going to a film out, then we render to dpxs, correct in color, and re-render to send to the film houses, you can even take those dpxs to you coloring house of choice.

P.S we render close to 90fps in offline on 3 octicores and 12-24fps for online depending on full debayer on 4k or not.....

sorry if this solution does not work for you
scott
 
Thanks, Scott. We use Clipfinder nearly every day. But I believe that had the same render issues on long clips. It seemed more reliable than RED Rushes, so we have rendered a lot out with it. I have never tried the XML back in, so maybe I'll experiment with that.

My hope is that they have some sort of native roundtripping with RED Cine and any XML based editor like Final Cut Pro. I'm not sure how hard this is programming-wise...it sure doesn't seem that hard.

I would happy to do 95% of my color work in RED Cine & Final Cut Pro. Only occasionally do I really need to go into something more powerful like Color.

Thanks for the thoughts!
 
Really?

Really?

Apple Color is more powerful for grading than RedCine??? AFAIK in Color you are working in 2K 10 bit YUV and in RedCine in 4K 12 bit RAW. While the toolset in Color may have more options I would prefer to tweak the substantially more robust data set RedCine has access to.

I wholeheartedly support the notion of tapping the potential of XML to create better integration between post tools like editing apps, compositing apps, color grading, finishing, audio layback, etc. Part of the promise of a data centric workflow with proper metadata support is to automate and integrate the entire post infrastructure. Yes, I know this is a future implementation that doesn't help you today.

Disclaimer - I use Scratch, its not cheap, but it works extremely well, yields better images and saves oodles of time. Just sayin'.

It should also be noted that desktop based DIY 4K, even with bugs, is kinda mind boggling. Its also of note that RED is actually a very tiny group of sharp cookies who can only do so much in a day - pleading with them is understandable but politeness is still a virtue.
 
Blair, as you know, RED Cine is more of a one-light tool than true color correction. Yes, Color is a MUCH more powerful coloring tool, though it obviously doesn't support RAW files yet. Keep in mind, they announced they would support it at NAB in April. I don't know what the delay is, but once they support that, it should absolutely blow away what you can do with RED Cine. (Secondary correction, tracking, keying, etc.)

I agree with you that what is happening is fairly mind boggling. I continue to give credit to RED for what they are doing and I will continue to POLITELY remind them to keep their eyes on "the other 50%" of this whole movement...workflow. :-)
 
Hey Kevin-

Yes Hans is working furiously on making all this happen, I imagine it will be soon. If you have some concerns try and send him a PM, he is a very kind man and very helpful. I think the released version has an import from fcp xml, then you chose replace mov with R3D, navigate to the folder and there it is. Very easy. With this process you re-render whole clips which bothers many (which is why he is making the conform) but I don't mind as I can edit to my hearts content with slo mo and other problematic conforms and it always re-links because you create identical files (well with better res and color)

I personally don't use redcine because it does not render as stable as redline, so with clipfinder we can do a pretty good onelight in redalert and hans will have a new version that allows you to lift and stamp the color onto multiple clips. I love this. Then you can do minor tweeks. With that said I think the usability of redalert is much better than redcine because your color is saved to the rsx and it travels with the clip. I like redcine, and it has more options, but I can get it close enough and do final tweeks in FCP, or color if I am feeling ambitions. Blaire is right (as you know) color is a much better grader, but the interface is not as "fun" and I only go there for things that need major color correction.

And of course scratch is the only "real" answer, but hey we spent our money on the camera ;) And IMHO if you light well minor color is all that is really needed to make the shot shine, and that is easy to achieve in redalert.

P.S. the long clip render errors I experience are from rendering across the network. This is not a clipfinder thing, but a redline (or red) thing. I believe deanan found the bug and fixed it, but if you render the file locally they should always work. (at least they seem to for us)

Just wait for the released version of clipfinder, it will kick ass. One great feature I love is it will know what you rendered and what failed and you can just re-render the error clips, very nice.......
 
"2K 10 bit YUV"

its RGB only, always has been. Now Final Cut on the other hand...

As far as final quality, debatable. Really it mostly depends on the downsampling algorithm you use to get it to DVD ;-)

Seriously though, you can get great results in either if you have a talented artist. It really comes down to how much time it will take. If you're making money on a project, well I've seen things in apple color take 30 times longer than on scratch. Then I've pulled my hair out when all I wanted in scratch was a simple timeline so I could recreate something some crazy editor made.

Best tool for the job, and make sure you have enough experience with your toolset before you go into the flames of production.
 
Back
Top