How to QUICKLY sell a production on the RED workflow?

[Skip Hobbie]

I'm an EPIC owner, and I've been hired to shoot a documentary TV series. Mostly sort of standard travel type stuff, some episodes a bit more heavy on human / run-n-gun stuff, but other eps more heavy on wildlife (my specialty).

The production is excited about the possibilities and capabilities that RED brings to the equation, but this is completely new territory for them. They are hesitant about the workflow. I'm mostly a stranger to the post world, and looking for suggestions of a good successful workflow that will make the most of my camera, and cause the least additional headache to their post department.

They are in FCP7, and unlikely to budge far from what they do already. So what is the best way to take advantage of RAW, knowing that they'll want to edit everything in ProRes. Is the best option to just do a first light color correct in REDCine-X during the transcode, and then abandon the RED unless they decide they need to come back to the original material during the online / grading?

Thanks,
Skip

 

[Jack Shanahan]

Thats a good option -- Or perhaps give them a redlog film via external recorder so they can dial in a look and will have prores files to work with every day without transcoding. Just keep in mind you wont be able to shoot slow mo unless you playback to the recorder.

 

[Bob Gundu]

Well if it's a client you want to keep in the future, and to cause the least amount of disruption for them, I would deliver a first pass color corrected transcoded ProRes files to them. That way they get to know you and feel good about the experience. Then educate them on the merits of working in RAW as time goes.

 

[Jean-Claude Benet]

+1 to what bob said.

 

[Skip Hobbie]

Transcoding to ProRes w/ initial color correction is probably the route we'll go, but I'm thinking about getting a recorder if they want to go that route.

Is the Pix240 the best option?

 

[Chris McKechnie]

Transcoding to ProRes w/ initial color correction is probably the route we'll go, but I'm thinking about getting a recorder if they want to go that route.

Is the Pix240 the best option?

Consider Atomos as well. Either the Samurai or Ronin. http://www.atomos.com/field-recorders/

 

[Bob Gundu]

NAB will have more options this year, but the Samurai is serving me well.

 

[Blair S. Paulsen]

PIX ated

PIX ated

Agree that a one light transcoded to ProRes should best fit their pipeline. Suggest that after the editing is complete, they put a copy of the project file (and a CMX3600 EDL) on the same drive they archive the R3Ds onto in case they get an opportunity to re-master in 4K for another payday down the road.

The great thing about the PIX240 is that the audio section is pristine. In your use case I would make ProRes with hero audio in the PIX and feed that media to editorial. The downside is that they might just finish right from the ProRes files and leave a lot of image quality on the table. One thing you can do to partially protect yourself if that seems to be the inevitable track, is to shoot a few seconds onto the RedMag, open it in RCX-P, do a conservative primary grade, push that RMD into the camera and use it for the SDI out to the PIX - though that may be impractical if you are too run 'n' gun.

Cheers - #19

 

[Eric Haase]

To keep it simple for THEM:

1) Transcode RAW in best light dailies session with RedcineX to a low bitrate Prores (Proxy or Standard). Edit offline. Dailies look nice with your dailies grade.
2) Transcode RAW as second pass with same one light correction as #1 but change GAMMA setting to REDLOGFILM and use Prores4444 or 422HQ. Use these when you conform and grade.

It's a bit more storage space with 2 sets of transcodes but the workflow is dead simple for them. All file names and timecodes will be the same and you will pretty much preserve everything in the RAW on that Redlogfilm version of the transcodes. You also won't have to deal with another device on set that may or may not work properly. Just get a rocket, a sonnet enclosure and a laptop with thunderbolt and you're good to go. Let the dailies cook overnight even. Super simple.

 

[Patrick Tresch]

Eric,

If you have a conform and grade pass, I would go back to the RAW and not the heavy Prores4444.

R3D trim from the EDL is so simple in REDCINE-X. A 90 minutes feature is about 300GB.

Pat

 

[Eric Haase]

Eric,

If you have a conform and grade pass, I would go back to the RAW and not the heavy Prores4444.

R3D trim from the EDL is so simple in REDCINE-X. A 90 minutes feature is about 300GB.

Pat

Going back to the RAW would be better but the R3D trim is another step that could shy them away. Replacing media with a separate set of files is dead simple and any underpowered system can more easily grade Prores vs RAW. Just trying to make it as simple as possible while keeping quality very high and staying away from low quality options like recording onto a separate recorder like the PIX.

 

[Blair S. Paulsen]

Eric's suggestion is an excellent one that I have recommended to several shows and is particularly well suited to in house color grading pipelines. If final color is done at dedicated facility then I usually push for an EDL from editorial with conform from the R3Ds. ProRes4444 in RLF per Eric's note as the next best option.

If the editorial department is also the de-facto color & finish department, I inquire about how well their hardware can handle ProResHQ as the single media path. If creative editorial needs a low bit rate codec like LT due to hardware limitations then Eric's pipeline makes more sense. If ProRes HQ is fluid enough for them, then you can save drive space and avoid the conform step. If latitude for grading is an issue in this pipeline, I just do the one light to RedLogFilm and force the editors to look at low contrast images (or use an inverse LUT if practical). If the contrast range is well controlled and the color temperature well managed, a one light to Rec709 ProResHQ is the simplest option of all - but can be a serious handicap if more than minimal grading is needed.

Cheers - #19

 

[Eric Haase]

Eric's suggestion is an excellent one that I have recommended to several shows and is particularly well suited to in house color grading pipelines. If final color is done at dedicated facility then I usually push for an EDL from editorial with conform from the R3Ds. ProRes4444 in RLF per Eric's note as the next best option.

If the editorial department is also the de-facto color & finish department, I inquire about how well their hardware can handle ProResHQ as the single media path. If creative editorial needs a low bit rate codec like LT due to hardware limitations then Eric's pipeline makes more sense. If ProRes HQ is fluid enough for them, then you can save drive space and avoid the conform step. If latitude for grading is an issue in this pipeline, I just do the one light to RedLogFilm and force the editors to look at low contrast images (or use an inverse LUT if practical). If the contrast range is well controlled and the color temperature well managed, a one light to Rec709 ProResHQ is the simplest option of all - but can be a serious handicap if more than minimal grading is needed.

Cheers - #19

One transcode is the simplest for sure and if you just balance it well and don't go too heavy on a look probably the best option as you can use this set of files for grading.

I hate to deliver super flat redlogfilm for offline. As a DP it's unfair for everyone to be looking at something that doesn't resemble the intention of the look. Having dailies be close to final intention goes a long way in having the look "stick" when it comes to the final.

 

[Naim Sutherland]

Doesn't Apple Color support R3D? What do they usually colour correct in? It's possible that whatever they already use can handle r3ds and you just have to help them relink their picture lock to r3d for colouring. If they have a decent colouring setup their minds will be blown when they sit down to colour r3ds.

I think a hyperdeck or samurai is nice to skip transcoding on a computer if that's an issue for you... maybe you're in the field or something.

I saw a post around about relinking to the r3ds after recording Hyperdeck just using timecode, so that is possible.

 

[Rob Macey]

I use my Pix240 for exactly this. Burn ProRes or DnxHD. If they don't do any CC at all, just capture the look as you/they want it. Baby steps. Hold their hand. Then, show them what it can really do in post.

 

[Steve Sherrick]

All valid suggestions here, but I do have to address one thing and it's more of a business issue. I'm going to assume Skip is going to shoot a lot of footage as this is for a documentary. I'm assuming (big assumption) that the company that has hired him is not going to pay for transcode time. If that's the case, he is doing this on his own time. With a Rocket, this may be a reasonable tradeoff, depending on whether the dollar figures make sense. But if he doesn't already own a Rocket or other hardware needed, then not only would this be a time suck that doesn't get compensated for, but it's also potentially a few thousand dollars to get a capable system going. I wonder at this pointy if the dollars make sense. Again, all based on assumptions but I think this is a situation that happens often and is why the Rocket solution is up for discussion right now because it's an expensive option for many people, although it does simplify the workflow. And it's one of the reasons why RED workflow is perceived as hard. If you are set up properly for it, then it's not really that hard. If not, well it can be hard or at least time consuming depending on what you are trying to do.

In this case, I would recommend an external recorder if the client's expectations are to receive files as they are shot. You build an in-camera look that gives them nice looking editorial files. If they want to also grade off those files, then make sure there is some room to grade. You can encourage a conform to R3D but not everyone will be sold on this workflow. Also, make sure to do some test conforms with your files recorded to the external device to make sure you can get a perfect match with R3Ds during the conform.

I know in a world with native R3D editing in CS6 people wonder why someone would choose to cut in FCP 7 or Avid. I think we all know the answer to that. We are creatures of habit and to be honest those NLEs still to this day do some things better. CS6 also does some things better than both of them. But getting people to budge from something they know well is no easy task, and it's our jobs to deliver what clients want/need. We can't force square pegs into round holes. Sometimes compromise is the only way to go. The great thing about working with a client and delivering what they need is that as the relationship continues,a trust is built and that can lead to evolving the workflows. These things sometimes just need time.

Skip, if all my assumptions are wrong, then certainly doing a one light in RCX is a great way to go. Allows you to deliver whatever they need. As long as you have the horsepower to do it during a hectic production schedule.

 

[Patrick Tresch]

Well instead of transcoding all the rushes he just has to transcode the used shots. Trim the R3D and transcode them to prores4444. This is less than 90 minutes of rushes... Something like 20 hours of transcode in Full res?


Pat

 

[Blair S. Paulsen]

The offerings in the outboard recorder market are maturing nicely and price points are dropping quickly - look for some game changers at NAB. I always recommend the PIX when double system with a proper sound department is not in the budget, or is logistically impractical. The Sound Devices pre-amps in the PIX are state of the art. Epic on board audio on a single system show requires some TLC in my experience.

Cheers - #19

 

[Steve Sherrick]

Well instead of transcoding all the rushes he just has to transcode the used shots. Trim the R3D and transcode them to prores4444. This is less than 90 minutes of rushes... Something like 20 hours of transcode in Full res?


Pat

Not if someone else is editing. He has to provide client the rushes, in this case in ProRes. He has to have a way to transcode all of the footage to ProRes unless the client is going to do this on their own. That's why the external recorder may be the best option for him so he can just copy that to a drive along with the R3Ds and let the client handle it from there.

The conform of course will be less intense on the transcode side of things because it will be 90 minutes or so of footage depending on handles. But transcoding all of the rushes with a one light as he's going along will get unwieldy without proper system in place. And again, unless he's being compensated this could be a lot of extra time given away.