Assimilate has released SCRATCH RED Pre-Post

[I Bloom]

Right now, I am avoiding that $50k+ price tag, not because I can't afford it or I won't make money in the here and now with such a system, but because I don't want to end up competing with a $10k package that runs on a $2k computer in 2 years.

At that point, I have to cut my prices so deeply, that it's not profitable enough on the second half of the lease. Just dollars and cents.

-Thor

Hard to argue with that. Though I have to say, I have my doubts about Color. Apples track record with Motion for example, has not been good.

IBloom

 

[I Bloom]

I was previously shut down for saying that a quicktime only pipeline is the solution to most of our problems. i.e. we don't need a conforming tool like Scratch if a FCP redcode timeline can be exported with a full debayer and 12bits of depth, after that there are programs that can make it happen right now.

Scratch is fast, I'm hoping to buy it here, for the here and now. But if I cut something native redcode, i.e. we are cutting actual material and it's all lined up in FCP. Then why would I have to line it up again in Scratch, we don't need to conform, WE ARE CUTTING THE ACTUAL MATERIAL. Cineform might factor into this equation as well.

Based on my calculations, and Moores law, Scratch will be economically viable at it's current price until April 2009. Set your watches team. If you have Euros then June 2009.

Anyways, seriously Rob. Enough with the sleeping. That's not helping the situation here. Really

IBloom

 

[Mark L. Pederson]

Right now, I am avoiding that $50k+ price tag, not because I can't afford it or I won't make money in the here and now with such a system, but because I don't want to end up competing with a $10k package that runs on a $2k computer in 2 years.

At that point, I have to cut my prices so deeply, that it's not profitable enough on the second half of the lease. Just dollars and cents.

-Thor

I try to live by a 6 month ROI rule for hardware and software.

That doesn't mean I pay it all off in 6 months, but I run the numbers and need to know that by making that purchase I create enough revenue to cover the investment in 6 months or less.

As far as competing again a $10K package in two years ... well ... as far as I am concerned ... two years is a LIFETIME away.

Scratch ROI's itself with just a few small feature films.

Just dollars and cents.

 

[Mark L. Pederson]

But if I cut something native redcode, i.e. we are cutting actual material and it's all lined up in FCP. Then why would I have to line it up again in Scratch, we don't need to conform, WE ARE CUTTING THE ACTUAL MATERIAL.

So you can color grade it (directly from the .r3d files) with all the powerful tools Scratch has to offer - and then output in any format(s) with a full quality decode.

And "line it up in Scratch" is two clicks of the mouse and DONE.

And for the record, you are CUTTING actual material - but you looking at a much lower quality decode.

 

[I Bloom]

As far as competing again a $10K package in two years ... well ... as far as I am concerned ... two years is a LIFETIME away.

I think that is the core of your argument for Scratch, Mark. I think it has to be correct. If I was in your position I would make the exact same choice. If I could be a Scratch customer in the next two years, I would or will.

So you can color grade it (directly from the .r3d files) with all the powerful tools Scratch has to offer - and then output in any format(s) with a full quality decode.

And "line it up in Scratch" is two clicks of the mouse and DONE.

And for the record, you are CUTTING actual material - but you looking at a much lower quality decode.

True all that. But it's still the same 27mb sec data that is passing through FCP via the proxie component. Unlike in the case of conforming uncompressed film DI, where the actual material is too heavy to pass through the NLE. In your case the conforming interface on Scratch is not as important, it's more like a way of importing an FCP timeline. If the computer was doing its job properly the clips wouldn't fall out of sync (not including stuff coming in from compositing and visual effects, but FCP might handle those directly as well with the right codec.)

Do you find huge discrepancies when you bring in an EDL that is straight Redcode?

In any case, I'm not suggesting that the quicktime proxie decode is what I'm going to use, we need another path that gets us to full quality. If my last step in FCP is to export it with a full debayer and downconvert to 4K Cineform. Drop it on a RAID and then move to color correct, I'm not in as good of shape as you are, but I'm still pretty great. (Some issues with this today, I realize. Transitions especially... I'm hard at work with testing and some scripting. It might also be the case that RedCine must have a big role here as well with pull lists etc. Unfortunately there is no way to control RedCine but through their interface. A utility that decodes an .R3D and is accessable from the command line, would be very helpful for guys like me. A non-realtime quicktime component that does a full debayer, would also solve a lot of problems. Either one of these will allow us to distribute transcodes over several servers.)

There are quite a few boot-strappers for whom an extra few steps is worth $100K for 2 years. They might be taking on 500K features instead of 2-3 mil or more. In their case doing a full decode first, and then going to Cineform for example is a path worth investigating.

Until I get full access to the material I'm shooting then REDcode feels a little bit like buying an Iphone and finding out you have to switch to AT&T. AT&T might be great, but screw that.

I think this is what guys like Laguun are reacting so strongly to. It feels a bit like DRM...at least tody... emphasis on the words feels like.

Ian

 

[Mark L. Pederson]

Do you find huge discrepancies when you bring in an EDL that is straight Redcode?

None.

In any case, I'm not suggesting that the quicktime proxie decode is what I'm going to use, we need another path that gets us to full quality. If my last step in FCP is to export it with a full debayer and downconvert to 4K Cineform. Drop it on a RAID and then move to color correct, I'm not in as good of shape as you are, but I'm still pretty great. (Some issues with this today, I realize. Transitions especially... I'm hard at work with testing and some scripting. It might also be the case that RedCine must have a big role here as well with pull lists etc. Unfortunately there is no way to control RedCine but through their interface. A utility that decodes an .R3D and is accessable from the command line, would be very helpful for guys like me. A non-realtime quicktime component that does a full debayer, would also solve a lot of problems. Either one of these will allow us to distribute transcodes over several servers.)

For the record, I have also requested a full quality 4K QT wrapper - just so I can automate the distribution of transcodes and automate multiple, individually watermarked DVD's encode, FTP uploads and web dailies.

I think I understand what you want to do - but please FORGET exporting out of the FCP timeline until they move the FCP engine to CoreVideo - don't even get me started on that. We have been flagging export bugs for years in FCP.

Let's see how RED does with their "pull list" - because if you can use REDCINE to process just what's in your cut and re-conform in FCP or Premiere - you would be happy, no?

I also think it is possible to hack a "conform in REDCINE" script.

There are quite a few boot-strappers for whom an extra few steps is worth $100K for 2 years. They might be taking on 500K features instead of 2-3 mil or more. In their case doing a full decode first, and then going to Cineform for example is a path worth investigating.

I think you can go to Cineform now out of REDCINE.

Until I get full access to the material I'm shooting then REDcode feels a little bit like buying an Iphone and finding out you have to switch to AT&T. AT&T might be great, but screw that.

I think this is what guys like Laguun are reacting so strongly to. It feels a bit like DRM...at least tody... emphasis on the words feels like.

Ian

Well ... I think that's going a bit far. IMHO RED is 90% there. And they could be 100% there in a matter of weeks.

There are TONS of apps that use DPX, and you can get DPX from your RED footage with meta-data right now.

We are cutting one feature now, shot 100% on RED on a PPC mac off of ONE hard drive using self-contained 1K proxie transcodes with sound.

For the record, there is too much obsession with cutting from .r3d extractions. Sure, it's COOL. And for short form content, or FAST turn around, it will be VERY handy - but I can give you MANY reasons you should edit long form in a TRANSCODED QT. The most obvious is media size. An entire feature on a single, internal 750G drive in 1K prores is very nice, clean and simple. It also looks great!

 

[M Most]

None.

I think you're making it sound far too easy, and it generally isn't. With Red productions, there is one potential source of error that can be fairly annoying, and that's the embedding of the reel number in the R3d file metadata. When it's set correctly (according to whatever numbering conventions you agree on in preproduction), it works very well and allows for flexibility in terms of how you set up your storage. But if it is not, you'll have a very hard time when conforming.

In terms of general conforming issues, these exist regardless of what's being shot. Editors will often use undocumented material (VFX shots, stock footage, you name it) that you then have to locate. They will do effects in the timeline that aren't advertised (color effects are actually the least of these). They will do speed changes that are not accounted for. They will make editorial tweaks after they've generated your EDL. They will generate 30 frame EDL's from 24 frame projects. They will hand you a Final Cut project rather than an EDL, and the final sequence will have 8 video tracks, for no apparent reason other than "that's the way they cut." And numerous other things I won't even go into. Now, all of these things happen a lot less with experienced, professional editors, but...... well, you get the picture. Conforms in general are far less straightforward than you're making them out to be, Red or no Red.

For the record, there is too much obsession with cutting from .r3d extractions. Sure, it's COOL. And for short form content, or FAST turn around, it will be VERY handy - but I can give you MANY reasons you should edit long form in a TRANSCODED QT. The most obvious is media size. An entire feature on a single, internal 750G drive in 1K prores is very nice, clean and simple. It also looks great!

Completely agree, especially for longform, where you could easily have 8TB or more of original source files. Which is why I also suggested the ability to generate truncated R3d files.

 

[Mark L. Pederson]

I think you're making it sound far too easy, and it generally isn't. With Red productions, there is one potential source of error that can be fairly annoying, and that's the embedding of the reel number in the R3d file metadata. When it's set correctly (according to whatever numbering conventions you agree on in preproduction), it works very well and allows for flexibility in terms of how you set up your storage. But if it is not, you'll have a very hard time when conforming.

In terms of general conforming issues, these exist regardless of what's being shot. Editors will often use undocumented material (VFX shots, stock footage, you name it) that you then have to locate. They will do effects in the timeline that aren't advertised (color effects are actually the least of these). They will do speed changes that are not accounted for. They will make editorial tweaks after they've generated your EDL. They will generate 30 frame EDL's from 24 frame projects. They will hand you a Final Cut project rather than an EDL, and the final sequence will have 8 video tracks, for no apparent reason other than "that's the way they cut." And numerous other things I won't even go into. Now, all of these things happen a lot less with experienced, professional editors, but...... well, you get the picture. Conforms in general are far less straightforward than you're making them out to be, Red or no Red.

I was just just commenting on Ibloom's specific question on conforming from .3rd files.

But yeah, MANY, MANY things can (and will) get screwed up on the source side - and everything else you mentioned - you and I fight the same fights - we need tool to edit .r3d files, etc.

So, I did not mean to say "conforms are easy" - just meant to say that a .r3d acquired correctly conforms for us with no issues.

 

[I Bloom]

For the record, I have also requested a full quality 4K QT wrapper - just so I can automate the distribution of transcodes and automate multiple, individually watermarked DVD's encode, FTP uploads and web dailies.

I agree with that. I want RedCine to be working primarily on metadata, and a quicktime wrapper so we can distribute transcodes with Compressor.

I think I understand what you want to do - but please FORGET exporting out of the FCP timeline until they move the FCP engine to CoreVideo - don't even get me started on that. We have been flagging export bugs for years in FCP.

I see what you mean, I have run into this as well. Very good point.

Let's see how RED does with their "pull list" - because if you can use REDCINE to process just what's in your cut and re-conform in FCP or Premiere - you would be happy, no?

That might work. In that case Premiere might have the upper hand.

I also think it is possible to hack a "conform in REDCINE" script.

In the works. But again, this is easier if the actual work is done through a full quality quicktime wrapper, with RedCine as just a place to define parameters. Would rather hack Apple automation then a standalone app like Redcine.

I think you can go to Cineform now out of REDCINE.

Small bugs apparently on the RedCine side, but yes true.

Well ... I think that's going a bit far. IMHO RED is 90% there. And they could be 100% there in a matter of weeks.

Point taken I think I'm speaking specifically about the exclusive agreements that are in place today. Cineform as a Redcode import layer for Premiere would be very useful and would likely be available very quickly.

There are TONS of apps that use DPX, and you can get DPX from your RED footage with meta-data right now.

Too heavy for indies though. Been messing with Cineform, seems like it opens a lot of doors here. Only problem is one generation loss. Not sure yet how problematic that is. I know the Red team doesn't welcome this, as Cineform RAW can be considered a competitor to REDCode. But the tool is here and now. I wonder if they will offer 4K REDcode RGB as an intermediate format for compositing at some point.

We are cutting one feature now, shot 100% on RED on a PPC mac off of ONE hard drive using self-contained 1K proxie transcodes with sound.

On a PPC, interesting. I'm putting my G5 up for sale today.

For the record, there is too much obsession with cutting from .r3d extractions. Sure, it's COOL. And for short form content, or FAST turn around, it will be VERY handy - but I can give you MANY reasons you should edit long form in a TRANSCODED QT. The most obvious is media size. An entire feature on a single, internal 750G drive in 1K prores is very nice, clean and simple. It also looks great!

Another advantage being Compressor can currently distribute the transcode around your facility, no? Your experiences with FCP are worth their weight in gold Mark. I respect you overall strategy, which seems to always be "right tool for the job right now, but no compromises on quality."

It's interesting because you've been so against RGB in camera. I have too. But now I'm seeing where this could drop right on a timeline with no need for debayer or rescale. For commercial work that's tempting.

Ugh. I can't believe I just made two cases for using Premiere. Apple can be glacial about solving some of their deeper problems I guess.

Ian

 

[Mark L. Pederson]

Ugh. I can't believe I just made two cases for using Premiere.

Ian -

Premiere is gonna surprise the #@%* otta lots of folks at NAB 2008, TRUST ME. Do NOT write that app off ....

 

[Joel Kaye]

Ugh. I can't believe I just made two cases for using Premiere. Apple can be glacial about solving some of their deeper problems I guess.

Yeah. Hate to say it, but Apple is an iTunes and iPhone company now. Adobe is a creative app company. I've been planning all along to get the latest Mac when the next Mac Pro comes out but now I'm second guessing myself and considering going with Premiere on a PC. In that case Cineform would be the codec though I doubt I'd work in 4K. I've also been planning to go the offline / online route. Maybe I'd just onlline 1080P.

What about Premiere on the Mac and then over to Color?

Mark - do you have some inside scoop on what Premiere is alpha/beta testing?

 

[Chris Kenny]

Yeah. Hate to say it, but Apple is an iTunes and iPhone company now.

That's a little harsh. FCS2 is a pretty nice upgrade, as is the new Logic Studio. These kinds of fundamental architectural issues are just hard to fix. Sufficiently hard that, in trying to do the same sort of thing with Shake, Apple decided the best approach was to end-of-life Shake and go write a new compositing app.

I suspect we'll eventually see FCP ported the same modernized architecture as that new compositing app. But that's not a trivial undertaking and it'll probably be awhile.

 

[Steve Sanacore]

Well Adobe certainly did a better job with Lightroom than Apple did with Aperture, and support for new camera formats from Apple is so lagging it is un-usable on the pro side. Now that Premiere is running on the Mac again - is that a viable alternative to FCP2?

 

[Thor Wixom]

Deanan (working on thanksgiving day in the car debugging software with Rob )

I love you guys! Seriously.

-Thor

 

[Joel Kaye]

I suspect we'll eventually see FCP ported the same modernized architecture as that new compositing app. But that's not a trivial undertaking and it'll probably be awhile.

I'm not arguing that what Apple has to do isn't trivial. I'm arguing that they don't live in a vacuum. You think Adobe isn't full steam ahead working on their creative apps? I think they smell blood in the water and are moving much faster than Apple right now.

And what resources is Apple spending on rewriting FCP as opposed to the next iPhone or ipod? Meanwhile Adobe is releasing Audition three, has After Effects using multicore Macs like a render farm and keep tightening up their studio integration.

And Premiere is open to things like Cineform replacing their rendering engine.

It's hard for me to get excited about FCP anymore and it's hard for me to bet against Adobe. RED's workflow may warrant using FCP but I'm going to test that to be sure buying the newest Mac truly makes more sense than tricking out a PC.

I still really like Color though. It's a cool low end CC solution even though it's been pretty buggy.

Adobe on the Mac may just be the real ticket.

 

[Steve Sanacore]

I
Adobe on the Mac may just be the real ticket.

Adobe on the Mac is certainly the winner in the world of still photography. It would be a great option if they achieved the same level of performance with Premiere on the Mac. But I think Apple has a lot of energy behind FCP.

 

[Chris Kenny]

And what resources is Apple spending on rewriting FCP as opposed to the next iPhone or ipod?

It's not a meaningful question. The people who'd work on one mostly aren't the people who'd work on the other. And throwing more people at a software project to try to get it done faster often doesn't actually work. Significantly or completely rewriting an app that a large number of professionals have built workflows around is also something that one wants to do very, very carefully.

Meanwhile Adobe is releasing Audition three, has After Effects using multicore Macs like a render farm and keep tightening up their studio integration.

Adobe isn't exactly executing perfectly these days. A lot of people don't see the CS3 version of Photoshop as a particularly impressive upgrade, for instance. (Photoshop is, IMO, in a situation similar to the one Final Cut is in. It has been around for a while and its basic architecture is starting to become a bit of a liability.) It's also worth noting that Apple made a lot of the inroads it did because Premiere was, at one time, stagnating badly. Adobe basically rewrote it, and as a result now has a somewhat more modern foundation to work with. Things could easily tip back the other way in a year or two.

The big reason I'd bet on Apple long-term, though, is simply that I think controlling the platform their software runs on (and having a willingness to take advantage of features of that platform that don't exist elsewhere) is a major advantage for them.

 

[Joel Kaye]

The big reason I'd bet on Apple long-term

Long term? I'm just trying to figure out the best course of action for the next 6 months to a year.

I'm not religious about this stuff. I just re-evaluate every year or so. I've had as many Macs as PC's over the years and I can't really say I like one over the other. I just get whatever does the job best. I'm not marrying a platform or an app.

There's a reason Scratch doesn't run on a Mac...

 

[Mark L. Pederson]

Long term? I'm just trying to figure out the best course of action for the next 6 months to a year.

I'm not religious about this stuff. I just re-evaluate every year or so. I've had as many Macs as PC's over the years and I can't really say I like one over the other. I just get whatever does the job best. I'm not marrying a platform or an app.

There's a reason Scratch doesn't run on a Mac...

Very, very true.

Bet on the "Triple A" -

Apple, Adobe and Assimilate.

(yeah I know, Avid starts with A too - but I wouldn't take that bet)

 

[Thomas Mathai]

There is definitely a leap frog game going on with Adobe and Apple. One gains the upper hand then the other. Premiere coming back to the Mac will make it an even more interesting race.

I see it as Adobe catching up more in some aspects than getting ahead. I see this as a result of being unchallenged for so long and then suddenly finding that they need to catch up.

This is all pluses for the end user. I think having both Premiere and FCP is a good thing.

The integration between Adobe apps is nothing short of amazing. The FCP Studio 2 has some great tools. So to have both suites makes sense to me, sort of like one uber suite.