EPIC to output PRO RES

[Cüneyt Kaya]

And I can do all that without touching ProRes. I can do CC'd dailies while delivering R3D's or real-time generated H.264 or AVC.

I'm not saying anything against an offline/online workflow.

Maybe I'm just saying that people think they know what they want and they're asking for the wrong thing.

if the debayering is done differently, so you have the same fluid editing session like using dnxhd or prores and it works with all the other departments, this new thing can come...(talking about longform here)

 

[Peter Matthes]

Jeff,

It is a safe assumption that you know more about the Red post workflow in your pinky that I know in my whole body. I say this with 100% sincerity.

With that said, it seems to me that there is a value of a decent quality ProRes copy that can be offline edited and then EDL re-linked to the R3D files for conform.

I work with several clients that shoot endless amounts of footage. Believe it or not, this can sometimes mean a shooting ratio of 100+ to 1. (Painfully long interviews, endless B-roll, etc.) If they can weed out all the throw-away footage without having to transcode ... in a separate step ... all of the footage to ProRes or DNXHD, that would allow them the freedom to instantly offline edit and remove the wheat from the chaff before utilizing the data heavy Raw R3D files.

Even though Raw footage is without a doubt the move, and Avid MC5 AMA can now natively handle Red footage, it can still slow down or disable many older machines and/or laptops running Avid.

 

[Hans von Sonntag]

Most current NLE systems support R3D files natively now. FCP is flat-lining hard and the only reason it's on life-support is due to massive throngs of fanboy users and vague rumors of promises of something better coming. FWIW, I rebuilt this Mac I'm working on right now about 3 weeks ago. I did not install Final Cut Studio on it when I rebuilt it. Haven't missed it.

I agree. The last project we used FCP for is pretty much one year ago. Since then we use PremierPro and Smoke. I only need FCP as a translator to make Smoke to swallow the PremierPro XML.

The problem with native R3D editing and finishing is that every seat can read R3Ds differently. While this can be managed within a facility it becomes problematic when other facilities are involved. Also, you need special knowledge to work properly with R3Ds. There are many pitfalls regarding the "right" development of R3D files. We had that in many shots with PremierPro which lead to extra work and re-edits. We now use 1/4 R3Ds for editorial work only and create for finishing either graded DPX or ProRes files depending on the project.

ProRes is tied to QuickTime. It can't exist without QuickTime. As nice a codec as it might be, QuickTime is a scourge upon our industry and should be dispatched with extreme prejudice. Nothing good comes from QuickTime. Other than a facility for which to use ProRes, can someone please tell me one good thing about QuickTime as it applies to our industry and workflows today? I don't care what it's done for us in the past... I'm talking about now.

Yes, QT is useless for finishing. We avoid QT as much as we can. But for editorial rushes it works quite well. We don't use it as a timeline codec as one would normally do within the FCP pipeline. There the results are pretty unpredictable.

ProRes is not cross-platform. At least not in a complete sense. It is read-only on Windows and essentially non-existent on Linux. So if you hand your footage off in ProRes format to half the industry, they still need to transcode or utilize some intermediate codec.

True. But again I use it only for editorial rushes and never export to it. We use ProRes often in conjunction with the Adobe CS5 suite on Windows on a CineForm AVI timeline which does not show all those troublesome gamma shifts.

I see that several people are making the ASSumption that I'm speaking from the point of view of cine-style workflows or RED-specific post houses. On the contrary... My work is mostly non-scripted, non-narrative, non-hardlined EFP, documentary type work. I choose to shoot RED for quality. I don't have a problem with speed or making deadlines and quick turn-arounds these days. I work with R3D native files in most of my post pipelines right up until final render for delivery. I deliver many of my projects with fast turn-arounds. Often 48 to 96 hours for the completed project, delivered on Blu-Ray and DVD. I'm a one man show with 2 part-time helpers. I don't survive by doing what others do. I have to innovate and when it comes to my competition, I either make friends and keep my distance or I go for the throat.

We are on the same boat here. But every now and then I shoot a commercial that we don't do the post production for and I need a compatible workflow. Most post houses are not on the edge of development, on the contrary...

Hans

 

[scott william]

I wouldn't want a ProRes module... If RED did make one, then suddenly most of my clients would start demanding it. And directly recording ProRes would play right into the "good enough" mindset that so many people have. And why should we even need this now when we can create and load looks right into the camera anyway and then those are automatically transferred onto the R3D RAW footage. Other than FCP, every other mainstream NLE (the ones that matter, anyway) support R3D natively. If the next release of FCP/FCS does not have native R3D support I will not buy it. Case closed. L&T and QT wrappers is so 2008 -- time to move on. And maybe they'll shoot QT in the head and fix their gamma issues while they're at it. Not likely, but we can always dream of world peace and other impossible things.

And whatever happened to the RED working / RGB codec. If that's still on the table, why should they promote ProRes when they could release something better? Something that could be completely free to license and incorporate and use on any platform...

Absolutly!

 

[Tim Whitcomb]

ProRes module.... Um... 2008 just called and wants its workflow back.

Exactly

ProRes is tied to QuickTime. It can't exist without QuickTime. As nice a codec as it might be, QuickTime is a scourge upon our industry and should be dispatched with extreme prejudice. Nothing good comes from QuickTime. Other than a facility for which to use ProRes, can someone please tell me one good thing about QuickTime as it applies to our industry and workflows today? .

I can assure you Deanan and RC-X Team are in 1000% agreement with this assessment.

 

[Adam J McKay]

I too have recently done some tests with the Alexa and even at 422 hq you have an extreme amount of play in post. So 444 is beyond "good enough" in many situations.

 

[Zakaree Sandberg]

I think you can achieve good looking images with prores..

If you know how to light.. know how to work well with the art dept.. then you should have your shit down tight and a baked in image will not hinder your desired look.

Its the Guys who cant light properly or expose their image that sometimes use "RAW" as a crutch, and they may have an issue later on if they dont shoot raw

Not saying RAW doesnt have its place.. but for a good chunk of jobs that arent going through a film out , pro res is a great option that can be color graded to look amazing as is.


sometimes I dont need to dig that deep into an image.. sometimes i know what i want.. and i shoot for that.

I think Epic should have a prores module, or else i can see alot of jobs going to alexa to save on transcode time..

 

[Gunleik Groven]

I sort of totally disagree on your points on exposure, "raw as a crutch", "shoot for that" and the flexibility of prores, but I do think you're right on your last point... -

More from perception, though - than practical implementation.

 

[Zakaree Sandberg]

sorry it came off weird... haha
what i meant was it can be used as a crutch.. its more forgiving to shoot raw..

I love raw..

But some jobs just need to have a quick turn around.. and a pro res module would take care of this..

you can still achieve amazing images with prores..

Ive seen alexa pro res stuff that looks insane!

 

[Gunleik Groven]

And THAT is fully in sync with the OP's intentions, and proof of record.

The rest of the argument is sort of academical (well not really) , when stuff "just works" -

 

[Peter Moretti]

... BUT, lets talk about that Module. Will it do full 5K Debayer on transcode like the Red Rocket? Will it take advantage of all that is glorious of EPIC? will it have to be small and modular to EPIC? If yes to any of the above there is NO way this will cost less than $10 to $15K module. When a Rocket alone costs $5750. ($3K refurbed)...

Tim, software solutions using off the rack equipment are generally more expensive and require more power than hardware ones. Case in point, a Ki Pro ProRes 4:2:2 recorder can be bought today for $2K.

Second, a RedRocket card does both decompression and debayering. FWIU, the decompressing is more processor intensive than the debayering.

Third, related to the immediately above, the camera does a full debayer already, so the module would not have to do all that a RedRocket card does. It would "only" have to compress to ProRes.

Now I don't know how much more it would cost to make a 4:4:4 recorder, or if Apple would even license that codec to Red, but a viable means of recording ProRes exists here and now for $2K.

 

[Carsten Fenstermacher]

Or you want real fluidity while editing. Or you don't want to carry around a tower or external drives. Or you don't want to use 1/8 res images. Or you're a professional editor who doesn't want to deal with debayering or color settings. Or you want color corrected dailies. Or you're planning to go to preview on a movie and want to use the offline as a source for that. Or......

There are still a number of good reasons for an offline/online workflow, even if they're not necessary for you personally.

Does it need to be on camera?
If I had a module on camera I'd still need to copy it so why not a realtime box that made copies for me but into prores? Then I don't need to add more weight and power at the camera end and add it where it's needed on the dit cart, post house, or editorial?

 

[Rainer Fritz]

I go only out to a job with my macpro and rocket as calibrated dual link HD-SDI monitoring. so the dit base is at this moment in the camera truck... I get a CF card, copy, onelight, PRORES 4444 file in realtime exported... everybody can have the fileformat they want....

 

[Gunleik Groven]

I like the transcoding for added security... -
Picks up more copy problems than anything I know.

I really cannot see how going online from prores is a good option when the images are challenging exposed, and most filmphotographers I have worked with uses every available little dynamic corner of the RED. They don't expose "for the RED" but "for the image they want", and there's no way I would NOT want to go RAW online (even when transcoding to dpx) instead of using the redgamma prores as a basis.

But for offline, I see it can be a huge timesaver.

And I agree with Mark, that for many projects it may well work well for online, too...

 

[Harry Clark]

I think flexibility is the key here.
I have 2 Epics coming my way (when???) to replace my 2 MX cameras, and it would be great to have ProRes recording in a module (or better yet, a module that lets you choose ProRes, DnXHD, or H.264) But I already own an Alexa, because that's just what some clients will want.
It may in fact be impossible for Red to try to be all things to all people. And based on the way Jim Jannard speaks with passion about their design philosophy I don't think they really want to be all things to all people.
I took delivery of my Alexa just over a month ago and I have to say it's a really really nice camera. Not without its bugs and issues (think Red 2008 but more finished) but really nice. When Arriraw comes around, it will be possible to have a camera that does both, but Arriraw to an OB-1 system DOUBLES the price of the camera, so I won't be going there. If clients want RAW, and they don't want Red, I'll subrent an OB-1 or whatever other device. Maybe I'd buy a Cinedeck eventually. But either way, it's more expensive to do RAW than Epic or MX, and we all know Red is a terrific bargain in that regard. But Alexa is pretty darn flexible.
Now, an Alexa is at a fairly high price point as a per-day rental, and Epic will be close to it, whereas Red was a different price point, and MX slightly more. Different clients have different post requirements, budgets, and schedules, and it just may be the reality of the situation that in 2010 (and 2011 and 2012) you either need different cameras for different clients, or from time to time you (GASP!) might need to rent a camera.
BTW, I'm also interested in the AF-100 and have put myself on the list at Abel, since that might be a good camera for applications that are now DSLR. I HATE H.264, but if someone wants that and wants me to shoot a job for them and does not have the $$$ for an Epic, then that camera might be great. Yes, the S35 Scarlet will eat its lunch, but the Panasonic will fill the gap in the interim.
Don't know where I'm rambling to, other than to make the point that I don't think we will see one camera that fits all. I think that we will see cameras from Arri, Red, Panavision, and the usual video vendors that all have similar, but slightly different, design philosophies and price points. Each will have strengths and weaknesses. Do I wish there was a ProRes module for Epic? Yes. And maybe there will be. But there will be other way to make ProRes movies on a short turnaround. So I'm not sweating it.
Cheers,
Harry

 

[Ben Holmes]

This discussion always seems to dissolve into a mostly philosophical discussion on 'good enough'. You'll have to forgive me - but what a load of nonsense. Everything in film and TV production is about 'good enough'. Especially in TV.

'Good enough' usually boils down to cost, and is entirely subjective. In TV land, we've made do with far, far worse than "Epic + ProRes 4444" up 'til now. The vast majority of this market has simple post pipelines utilising FCP, Avid, AE, maybe Smoke etc. It's quick turnaround, it's tied into equipment on 3-5 year cycles. It does not want to invest ANYTHING in supporting a codec used by one camera if it can help it (Red Rocket). It does not want to spend MORE time doing anything. It wants quicker, faster, cheaper. It wants standards.

More and more of TV's post infrastructure is owned and run by production companies and broadcasters - the camera systems on the other hand are not, especially in an age of competing tapeless formats, barring the news organisations.

In the end - what does it matter? RED wants to sell cameras. Some people will want RAW, some will want ProRes. Let them have either, or both. Sell more cameras.

Will they sell more? Oh yes. Alexa is sold as 'a RED that shoots ProRes' to producers. And by any stretch it looks 'good enough'. They can't get enough of it.

Trust me - everytime a RED expert on here points out how easily RED native post can be achieved, something I totally accept by the way, the majority of the TV industry rolls its eyes and goes back to their terabytes of ProRes or QT archive footage on FCP and (now) AVID systems and already paid-for workflow muttering 'debayer?'

 

[Thomas Wright]

Well said Ben

 

[Gunleik Groven]

Ben. You're probably right on all this.

Even Rec709 on Alexa is pitched quite aggressively, evn though I cannot see how anyone really would like to use it that way for anything but high-end doco's and possibly musicvideos. But the fact of the mater is that Arri haven't really sped up the Alexa-RAW option, and I think that is because they WANT Alexa to be used this way.

I also happen to think that all those claiming to see the fast demise of FCP, probably are wrong, and that ties very well inn with Alexa, KiPro and the like. I thnk Appløe may have changed their strategy a bit with the release of prores, and tries to propagate that format into more and more recording devices, thus making FCP the simple and cheap solution.

That said.
It is really simple and cheap and actually good enough for a lot of uses and ties very well into large-organisation workflows.

AND:
I really cannot remember last time FCP crashed on me.
It's rock solid.

The alternative codecs are very often NOT RAW but XDCAM-HD or AVC-Intra (which I presumed was quite cool without ever really using it, untill I had a blue-channel peak in a Nuke suite. That really makes the R1 M blue-channel shine!)

Prores is both workflow-wise and quality a big step up - for the most occations.

I just don't want to see it in a feature/TV-drama setting as the main codec.
Post with RAW in those settings is so easy, good and flexible that I have to really bite my thumbs sometimes to believe it.

The alternative here is HDCAM SR'ish. AND I would never swap RED RAW for that...

Edit:
I guess that we'll see tons of (even feature) productions on the F3/AF-100 over the next years with a Ki-pro slapped on the side. I have even seen productions doing that with the D-21... -

The thing I am uncertain about is whether this is a good or bad argument for RED to provide the option.

For the Original-Concept-Scarlet, I'd have said it is a no-brainer.
But RED has allways had a stubborn "educative" side to its philosophy, and the prores module doesn't go too well with that....
I have really enjoyed that part of RED, though.

 

[Cüneyt Kaya]

if this module helps to not lose jobs to an alexa, i am all for it.
if someone wants instantly prores files with rfl, he should get it, instead prefering to pay 800€ a day for an alexa

the more we earn with the cameras the more we are able to keep up with reds new toys, which will be awesome anyway

 

[thomas s stanton]

My last few jobs are a perfect example of Ben's post. I hope the Epic can be a one-stop shop for a few of these scenarios. Producers often do not want to be convinced that their editorial guidelines/deliveries are archaic.

Job 1 -
ARRI ALEXA Direct to SxS Card: Record Quicktimes at 1080psf23.98 Apple ProRes 4444 in LogC format to editorial and colorist
Job 2 -
RED MX (3k, 16:9, 50fps, straight to edit with proxies,back to R3Ds for graphics/effects and coloring by effects house in color
Job 3 -
Canon 5D with PL Mount 1080p 23.98/h 264/converted in Media Composer 5 to Avid DNx HD 115 MXF/audio synced prior to editorial/no cc
Job 4 -
ARRI ALEXA - SXS cards 1080p @ 23.98 4:4:4 Log C/ transferred at Lightning Media to HDcam SR Tape with Rec709 Lut applied/ down converted each HDCam SR tape to it's own Beta Sp via SDI or component connection/time codes must match on all media
Job 5 -
RED MX + (5) Canon 5D - 4k 2:1 23.98 converted to Apple Pro Res 2k (hq) and h264 converted to 422