Working in TV - loosing Epic work to the other brand due to transcoding costs/ time

[Blair S. Paulsen]

I was an owner-op for more than a decade before I got RedOne #19 and the risk/reward equation is anything but simple.

One of the reasons I put a Scratch system in a van in 2008 was to resolve any concerns about the nascent RED technology by offering one-light transcodes to format of choice shortly after wrap. In the short term, in a lot of markets and genres, the owner-op ignores this element at their peril. That said, as I noted earlier, its 2012 and R3Ds are a common enough format that this should all be history by now.

For $400million USD RED could just buy AVID...

Cheers - #19

 

[shashbugu]

It is what it is. There are far more shoots using the Red workflow on a daily basis than the are Arri Alexa shoots. There surely may be more bigger budget shoots using the Alexa (nobody really knows). We have been assured that a prores module and dragon sensor are coming next year. What if Arri makes its own 15 stop 8k or 6k sensor....then what? Red has the winning card in hand. I don't see folks lined up at Sony's doors for their 8k sensor.

 

[Max M.]

Thank you to everyone for such an impassioned and polarizing discussion. I truly appreciate all the feedback and intelligent responses. I think its an extremely important discussion. Once a shoot is wrapped, i want to move forward to the next job...not be on the phone with an editor trying to figure out which version of RED Cine X was used to transcode, what version of Quicktime is on that computer and why the black level has jumped up 7.5% from where it looked in camera and why the audio didn't render in the transcode. (spent about 2 hours on phone and e-mail with that today from a shoot last Thursday for Spike TV)

Unfortunately the suggestion from another REDUser of just buying a new tower, a Rocket, a dedicated cart and bringing a transcoder to set and not billing for it just to convince a production to use the Epic is not an option. Also...dictating to my clients what editorial platform they must use because i want them to edit R3D's natively is also not an option.

But honestly...i really appreciate all the response and intelligent discussion. (not just ravings from fan boys - but real, practicing professionals)

Thank you!

Yep totally agree, you nailed it. Red can really make a comeback with a prores dnxhd module which shoots nice 1920x1080 444 and 422 formats (also in high speed!) My biggest concern is most clients hate 5k and the data is creates, thats a big dealbreaker. So with the module able to deliver alexa like data people will be really happy. And although redcine is great, the momet i mention it i see a lot of producers look to me like what are you talking about.

The funny thing the size of the alexa is for some dops also a pro istead of a con, it lays balanced on their shoulder without all kind of different shoulderrig constructions... But i love the small size of the red. I feel 2013 is going to win back alot of people hopefully

 

[Jeremy Wiles]

Yup I do it regularly. 4k r3d. No rocket. No transcoding. And sometimes same day delivery. And my edit computer is almost 5 years old. Thanks largely to adobe premiere.

I don't understand why everyone isn't switching to Premiere. No proxies. No transcoding. No rocket. Easy... easy.... easy. Edit right away and export same day.

 

[Rob Ruffo]

Sometimes Red isn't the best tool for the job. Like when you have quick turnarounds.

But it's SO easy to sue any fo the several NLEs that directly support Red in raw - you really only have to transcode the final edit - and how long could that take? Even without a rocket?

Besides, Epic + Rocket is still WAY cheaper than "Competing Cam" and can do so much more.

 

[Rob Ruffo]

Yep totally agree, you nailed it. Red can really make a comeback with a prores dnxhd module which shoots nice 1920x1080 444 and 422 formats (also in high speed!) My biggest concern is most clients hate 5k and the data is creates, thats a big dealbreaker. So with the module able to deliver alexa like data people will be really happy. And although redcine is great, the momet i mention it i see a lot of producers look to me like what are you talking about.

The funny thing the size of the alexa is for some dops also a pro istead of a con, it lays balanced on their shoulder without all kind of different shoulderrig constructions... But i love the small size of the red. I feel 2013 is going to win back alot of people hopefully

I agree with data amounts being a real problem - which is why the new 6K Dragon will be much more of a problem than a solution, and I wish they would stick to 5K or even drop to 4K but with an as-wide field of view.

None fo our clients have EVER asked for a 4K deliverable - and be honest - neither have yours - and when will they? Ever? Even for photography 5K on MX already out-resolves a 5D (in stills mode) - and more importantly out-resolves most lenses . Even if at some time in the future a 4K deliverable becomes a real possibility outside of a handful of tentpoel films, - will there really be any difference that a human eye can perceive between 5K and 6?

In the real world as it stands today 5K is already too much resolution and more is just silly "numbers gaming" and not useful to making money on gigs with actual real world producers outside of maybe the blockbuster market - and Spiderman does not seem to have suffered any shot at 5K - did anyone on that set say "Gee - there are just too few pixels here?".

Our clients are not billionaires who can buy as many computers as they want. They have limited resources, and 6K is just too expensive.

I think i am not alone in saying we will likely not upgrade to Dragon. It will make our Epic unsellable to real world jobs (and yes I know I can select lower resolutions, but I don't want to have to crop in on my lenses more just to get the 5K data expense I'm getting now)

I would venture that most here (not all, just most) who say they want 6K are hobbiests who do not deal with real clients and tight budgets and deadlines. They just think 6K will be "better" without really thinking it through.

Please focus on other things, like dynamic range, noise, color rendition etc. Those are always useful. Resolutions that cost lots of money in post (faster computers, more data, need for new Rocket cards, room to store cases full of hard drives, LTO tape purchase and transfer time - all of these scale up with more res) and cannot even be seen by human vision let alone most lenses are not.

 

[Jeremy Wiles]

Please focus on other things, like dynamic range, noise, color rendition etc. Those are always useful. Resolutions that cost lots of money in post (faster computers, more data, need for new Rocket cards, room to store cases full of hard drives, LTO tape purchase and transfer time - all of these scale up with more res) and cannot even be seen by human vision let alone most lenses are not.

Agreed. But, you know what that's like... :beatdeadhorse5:

 

[Max M.]

Yep rob, i agree with you. Red has a name amongst producers and post houses here for producing alot of data because of resolution that isn't needed for most productions. Sometimes its great to have 5k and usefull. But for 80% of what i do it's just annoying to have so much data. I would LOVE to have a possibility to shoot 1920x1080 raw with the same field of view as in 5K HD, will this ever be possible? The epic sends a 1920x1080 signal out the sdi port, so if that signal could be recorded raw internally would be really nice and give red a big advantage to win from the alexa. r3d files in 1920x1080 resolution, that would be A LOT less data than 5k and much better to edit without transcoding (computers can playback easily without rr, even older machines. I really hope one day red finds a way to do this.

 

[Blair S. Paulsen]

Not to be pedantic but you can set the compression level as you choose. If data loads are the rub then shoot at 10:1 or 12:1 compression. For many deliverables the down-sample will kill any minor artifacts anyway so the quality of the master is still far better than final distribution. That said, IMO if you are shooting for theatrical release, pulling stills, want to push in more than 20% in post, creating compositing elements, etc then 6:1 (or even 4:1 in some cases) is worth the data load.

I also suggest you think about what you're actually shooting; if there's a lot of high frequency detail and movement then the codec is under more stress and needs a bigger bit budget to keep things clean, if its a talking head or other fairly static shot - particularly with little important image information in the shadows - then 12:1 should do just fine.

The RedTeam has created a very flexible and customizable camera system that can be rigged to handle a wide range of assignments. I also think its only fair to point out that the Alexa creates a pretty big data load as well. At some point perhaps the real discussion is about whether you think shooting in RAW is of value or not. IMO having the RAW data set available to me for processing in the "digital darkroom" is part of an authoring "process" I find very fulfilling creatively. Not everyone will look at it that way, and that's their prerogative.

Cheers - #19

 

[Rob Ruffo]

Agreed. But, you know what that's like... :beatdeadhorse5:

Well, no, not until I see "3 more stops of dynamic range for 6K" rather than "6K for 6K' on a Dragon promo.

 

[Rob Ruffo]

Not to be pedantic but you can set the compression level as you choose. If data loads are the rub then shoot at 10:1 or 12:1 compression. For many deliverables the down-sample will kill any minor artifacts anyway so the quality of the master is still far better than final distribution. That said, IMO if you are shooting for theatrical release, pulling stills, want to push in more than 20% in post, creating compositing elements, etc then 6:1 (or even 4:1 in some cases) is worth the data load.

I also suggest you think about what you're actually shooting; if there's a lot of high frequency detail and movement then the codec is under more stress and needs a bigger bit budget to keep things clean, if its a talking head or other fairly static shot - particularly with little important image information in the shadows - then 12:1 should do just fine.

The RedTeam has created a very flexible and customizable camera system that can be rigged to handle a wide range of assignments. I also think its only fair to point out that the Alexa creates a pretty big data load as well. At some point perhaps the real discussion is about whether you think shooting in RAW is of value or not. IMO having the RAW data set available to me for processing in the "digital darkroom" is part of an authoring "process" I find very fulfilling creatively. Not everyone will look at it that way, and that's their prerogative.

Cheers - #19

12:1 at 5K looks way worse than 3:1 at 1080p on a 1080p monitor. On many gigs I'd rather have the 1080p with no macroblocks that I can grade hard with dramatic secondaries than the resolution I don't need or want.

It doesn't matter hwo I look at things- it's my clients and the beancounters behind them who actually decide what happens much of the time. They often could care less about anyone's "authoring process" or "creative fulfillment" and just rent another system that they think will be faster and cheaper.

 

[Blair S. Paulsen]

12:1 at 5K looks way worse than 3:1 at 1080p on a 1080p monitor. On many gigs I'd rather have the 1080p with no macroblocks that I can grade hard with dramatic secondaries than the resolution I don't need or want.

It doesn't matter hwo I look at things- it's my clients and the beancounters behind them who actually decide what happens much of the time. They often could care less about anyone's "authoring process" or "creative fulfillment" and just rent another system that they think will be faster and cheaper.

I'd love to see the side by side comparisons of the 5K at 12:1 vs 5K at 3:1 source material after its been mastered into the deliverable format. I have no doubt that some material would give it away but for images that compress fairly easily I'd be surprised. In any case, 10:1 vs 6:1 is a more "real world" comparison and I don't have the chart in front of me but I think some of the ProRes variants create as much or more data at 1080P than 10:1 4KHD RedCode files.

If you're trying to convince me that many clients don't care about the same things that I do and don't want to spend one nickel more than they have to as long as the piece gets signed off and through QC then I can assure you I am already painfully aware of that. As others have noted, units like the PIX240, provide a shortcut for 1080P projects and while it may require extra time to playback overcranked footage into an attached recorder at least the camera can do higher frame rates than 60/sec.

If your goal is to keep pushing RED to put out a "proxy" recorder as a module then that's understandable, but FWIW I think they already got that memo ;-)

Cheers - #19

 

[Mark Pugh]

I'll happily pay $6k for the 3 extra stops, plus sensitivity.
I don't think I'd pay much at all just for the extra resolution.
That second possibility is just hypothetical, though.

 

[Marc Wielage]

I'd love to see the side by side comparisons of the 5K at 12:1 vs 5K at 3:1 source material after its been mastered into the deliverable format. I have no doubt that some material would give it away but for images that compress fairly easily I'd be surprised.

I had an occasion awhile back where I was involved in some post tests with material shot at various compression rates. A night shot with a guy carrying a flaming torch (unclipped) showed all kinds of weird macroblocking at high compression rates. I seem to recall Jim Jannard saying the following:

The truth is that there is a steep curve of improvement with lower compression starting at 20:1 (acceptable) to about 7:1. After that the curve noses over. While there continues to be improvement to 5:1, it becomes almost impossible to see much of a difference from there to 3:1… just more data to deal with.

I think 5:1 is probably the sweet spot. Anything less than that is probably overkill. I don't recall us being able to see a lot of difference beyond this. 10:1, 12:1, you can see it in some cases, especially with smoke, explosions, or (my personal torture test) sports bleachers filled with thousands of people.

As far as Red losing work to Alexa, TV is often more about hitting deadlines than anything else, and I think in the right hands, Alexa can look great. At the risk of sounding like a broken record (remember records?), the final image is more about the lighting than it is about the camera.

 

[Ned Wilson]

I don't understand why everyone isn't switching to Premiere. No proxies. No transcoding. No rocket. Easy... easy.... easy. Edit right away and export same day.

I recently got Premier. Time to give it a shot.

 

[Ned Wilson]

It is what it is. There are far more shoots using the Red workflow on a daily basis than the are Arri Alexa shoots. There surely may be more bigger budget shoots using the Alexa (nobody really knows). We have been assured that a prores module and dragon sensor are coming next year. What if Arri makes its own 15 stop 8k or 6k sensor....then what? Red has the winning card in hand. I don't see folks lined up at Sony's doors for their 8k sensor.

What data or evidence do you have that backs that up? I don't see it all in NYC.

 

[Curt Pair]

YES! There is an INCREDIBLE "RED Proxy" module in the works... it will record on both SD cards and "mini" Red Mags! It's killer! It does AVID DNxHD and Pro Res! Hold on! It's coming... when it gets here.... good bye Pix 240 and nanoFlash! There are pictures of it floating around... here's one:

I have an EXCELLENT contact at Red that says this is coming sooner rather than later... be patient!

Anyone else noticing the Epic work getting tougher to sell to producers in television/ broadcast, tv advertising because they won't spend the time or $$ to transcode?

Will there ever be a module for the Epic that records to a ready to edit format (high speed footage, too)?

Also have heard ad agencies in my market dictating which brand of camera their vendors can use because they don't want to pay for transcoding.

Anyone with similar obstacles?

 

[Curt Pair]

how do you get high speed footage onto the pix 240 or samurai without the task of playing back all the clips after the shoot? - if that's the case...you might as well transcode if you are going to sit through that (or be required to hire editorial personnel after the shoot day to play all the clips into the HD Digital recorder) Also ties up the camera body if playback is your method of getting high speed footage to the HD recorder.

not really a good option for me...but thanks for the input.

We do this on "golf shoots" all the time... we shoot the over-cranked footage.... then when we move from one hole to another... "playback" the footage out of the camera and manually capture on the Pix 240 or nanoFlash... It's the only option we have right now. It does work though!

 

[Curt Pair]

Quick Turnarounds with Red are possible, but require a lot of support and planning. It also takes away some of the advantages of working with Red, mainly the RAW files. How much Color correction can you do in such a short time? When you take away the main reason to shoot Red, which is RAW, then it becomes similar to many other cameras.

The Samurai and Pix are great solutions, especially if you can create your desired look in RCX before hand and apply the look in camera. You could also offer to do the transcode for free if the money is the only concern.

The final "look" is what many TV and broadcast clients "want." This is especially true these days for "shallow depth of field..." I have clients that WANT THAT LOOK, and the Epic is the best tool for the job. You're right... planning is in order, but it just ultimately makes the shoot much smoother in the process!

 

[Curt Pair]

The only way to combat the argument is stop them dead in their tracks.

Whenever I shoot I include transcoding as part of the price. Not an option, not an extra, just $X to shoot and your footage will be delivered same or next day in whatever format and resolution they like.

With the Rocket it's a breeze.

Here, here Jim! I do the same thing! I think this is the ONLY way to do it! In fact... I believe that truly producers and broadcasters are looking for "specific" codecs as deliverables to them! I have a client that uses XDCam HD... no problem... I shoot, grade and transcode with Adobe. No problem... I export as the "file of their choice." It's a win/win, for sure. I get to use the camera I WANT... they get great footage, and everyone is happy! I don't charge extra for it either... the Rocket is so fast, I don't have to!