Entertainment Post in Burbank

[Matt Irwin]

I have a low budget music video coming up where we're mixing 16mm and DVCProHD for two distinct looks. (Finishing to SD and DVD)

 

We're looking to transfer 800' of 16mm to DVCProHD and Pro50 or digibeta. A friend of mine recommended Entertainment Post in Burbank as a good affordable telecine house, but I've never heard of them.

 

Anyone used them? We're trying to find a low rate, but not at the expense of a quality transfer...

 

 

Along the same lines, are there any places in LA that do telecine direct to a hard drive? I know about Bono and SpyPost, but I'm looking for something local.

 

Thanks,

[Kevin Zanit]

Don't know anything about Entertainment Post, but as for the direct to drive option, I was talking to the people at Magic Film and Video (another low-budget SD telecine solution), and they were mentioning being in the testing phases of a direct to drive option.

 

I may be testing this option for a project soon; I will post the results if I do in fact go with that workflow.

 

 

Kevin Zanit

[Matt Irwin]

Sounds great, I'd love to hear the results.

[Phil Rhodes]

Hi,

 

I tried desperately to use Entertainment Post when I was in LA doing a 35mm short back in April. We were all booked up, walked in, and found them all in a concerned cluster around a cloud of smoke in the bottom of the Shadow telecine's power supply rack. They booked us into Level 3 at the same rate with the same colourist on two hours' notice - try that in London!

 

I do hope I have an opportunity to go back and use them again - well, for the first time! Glenn's easy to get on with, and it's how they handle problems like this that is most revealing about how a company operates. Despite it all, I was impressed.

 

One of the things I did while I was there was to try and persuade them to do a hard disk service.

 

Phil

[Michael Most]

IAlong the same lines, are there any places in LA that do telecine direct to a hard drive? I know about Bono and SpyPost, but I'm looking for something local.

 

Just about anyone will deliver your material to you on a hard drive. Regardless of what you might read, nearly everyone goes from telecine to tape and then digitizes to a hard drive, followed by copying the files to your disk. This is necessary because there is nothing available on a Macintosh (which is what you usually want to use, because almost everybody requesting "hard drive transfers" is cutting on Final Cut, and many of them in DVCProHD) that acts as a "VTR emulator," so there is no way to control the computer like you would a tape machine - which is necessary for accurate editing from telecine. The way around this is to record entire lab rolls without an edit, which is what we do. But in order to assign time code, and in order to have a backup element (nobody wants to go back into telecine because a hard disk crashed), you almost always run videotape anyway. Not to mention the fact that with anything other than seriously compressed formats, like DV and DVCProHD, you can't reliably record in real time to a single disk drive anyway, which means that the files will need to be put on a drive array. Since most clients supply single drive Firewire units, the files then must get copied to that drive for delivery. This is not a real-time process, which clients don't seem to understand.

 

I would say that we currently deliver about 50% of the material we transfer on hard drives. In the case of DVCProHD, we go directly to the client's drive and run an HDCam backup tape. In the case of "Uncompressed" files, we go to a very fast disk array, then copy the files. In both cases, if there is no editing involved - i.e., it's MOS and a "print all" situation - we avoid the videotape digitizing step this way. If you want to talk further, contact me privately.

[Max Jacoby]

I tried desperately to use Entertainment Post when I was in LA doing a 35mm short back in April. .

And how was your 35mm experience Phil?

[Phil Rhodes]

Hi,

 

> there is nothing available on a Macintosh ... that acts as a "VTR emulator,"

 

Yes, there is.

 

> And how was your 35mm experience Phil?

 

Expensive.

 

Phil

[Michael Most]

> there is nothing available on a Macintosh ... that acts as a "VTR emulator,"

 

Yes, there is.

 

No, there isn't.

 

We use Virtual VTR, but it is not a VTR emulator for recording, only for playback. For recording, it only allows a hard record command, and must take time code from a separate feed. It responds only to Midi commands, so I set it up with a MIDI footswitch controller to trigger "record on" and "record off" directly from the telecine room, but in order for this to be useful, the reels must be recorded in their entirety (no editing) and a VTR must be pulled to generate the time code. It does allow us to send the video directly to the computer without playing back from tape, though, and because we're recording directly from telecine, it saves us time because we can avoid a separate digitizing step.

 

But don't let Gallery's website information fool you. Virtual VTR was designed to play back sync picture on a separate computer for sound mixing stages running systems like Pro Tools. It does not offer any kind of VTR emulation for recording beyond "start" and "stop." It would be nice if it did, but it doesn't.

[Phil Rhodes]

Hi,

 

Then I've been lied to, because I was told it did exactly that.

 

You know it would be rather easy to write a script control to tie together an activeX RS232 interface (connected to a 442 convertor) and the Blackmagic SDK.

 

If anyone thinks there'd be a market for it...

 

Phil

[Hal Smith]

It sounds like whoever designed the Orban Audicy DAW needs to get to work on a video version. The Orban is notable for having an interface that exactly emulates a professional reel-to-reel tape recorder - including scrub.

[Michael Most]

Then I've been lied to, because I was told it did exactly that.

 

Yes, you have.

 

You know it would be rather easy to write a script control to tie together an activeX RS232 interface (connected to a 442 convertor) and the Blackmagic SDK.

 

There's a lot more to doing a full VTR emulator than just machine control. First, you have to keep in mind that there's no "tape" loaded, so there's no source for time coded control. So you have to create a "virtual" tape of either a specified or open ended length - i.e., an endless 24 hour long duration - and "edit" into that. There are various devices around that do this, among them Drastic Technologies software (PC only) and various products by DVS. The Fostex DV40 "sort of" does this using DVD-RAM for recording. Avids also have some limited functionality for this purpose. Virtual VTR was not designed for this purpose and does not have functionality to support it. It's not difficult to use the computer to control an external device, like a VTR, and ingest what's on it. It's a bit more involved to go the other way, and at least to this point, on the Mac, nobody's done it.

[Phil Rhodes]

Hi,

 

Emulating almost any kind of edit operation isn't a problem once you have reliable device control emulation to begin with. I was pondering the possibilities of having it include some databasing and the ability to sit and do the nonrealtime layoff to client disks overnight, or whatever. It could generate proxies, EDLs, what-have-you. I don't have sufficient experience to predict how a particular facility might want it to react to the device control other than pretend to be a tape deck and "assemble edit" clips together onto an apparently timecoded "tape". I'll take any advice you can give!

 

Phil

[Michael Most]

Hi,

 

Emulating almost any kind of edit operation isn't a problem once you have reliable device control emulation to begin with. I was pondering the possibilities of having it include some databasing and the ability to sit and do the nonrealtime layoff to client disks overnight, or whatever. It could generate proxies, EDLs, what-have-you. I don't have sufficient experience to predict how a particular facility might want it to react to the device control other than pretend to be a tape deck and "assemble edit" clips together onto an apparently timecoded "tape". I'll take any advice you can give!

 

You might want to talk to Mark Gilbert at Gallery about 9 pin control and the myriad of issues associated with what should be a "simple" protocol. As for what you're suggesting, quite frankly, a simple "software tape deck" would be very useful, but you have to keep in mind a few things. First, when you enter the dangerous world of "data capture," you're entering a world of no standards, endless format requests, "wrapped" files (like Quicktime) and file based transfers (like DPX), obscure client requests, a lack of understanding about technical capabilities of things like Firewire drives, and expectations that are often counter to reality. I can't tell you how many times I've had to deal with clients who claim to need "uncompressed HD," and expect to be able to go directly to their 200GB single drive Firewire disk in real time with 2 hours of material. When you inform them that it will take more than 4 times that amount of storage and at least 4 hours of file copying, they get very indignant and claim "they read" that it was entirely possible and in fact is done at "all the other facilities." You then dig deeper and discover that what they REALLY want is DVCProHD, but they call that "uncompressed" and argue with you when you point out that it's not. Then they want to go directly to DVCProHD tape in 720p/24 format, and when you point out that 720p/24 doesn't exist as a video format, they argue again because "they read" that it does. That's the unfortunate reality of today's free form post world. Any more complication is not really needed, but what is needed is simple devices that are versatile by nature of their simplicity. I don't need to be able to automate disk layoffs (in part because clients generally want to walk with their material), but I could certainly use something that would allow us to do edited transfers to time coded file formats with frame accuracy, using current edit control protocols.

[Phil Rhodes]

Hi,

 

I know about the whole client-expectations thing to some extent; to be honest as the designer of a system there wouldn't be much I could do about that other than make the thing competent and flexible. Anything Blackmagic based would inevitably record AVI in their own codec, but you could then serve it off the machine as anything you liked (it's readily convertible to DPX). The only one that's tricky under Windows at least is DVCPRO-HD. If that's what you want, that's what you should capture, using a hardware compressor, because converting it afterward is going to be unacceptably time-consuming; that's then a whole other ball game from doing uncompressed work as the input hardware is different. You don't want your DVR to spend its time doing software conversions.

 

If people want to walk with the rushes then they're just going to have to wait; there is no solution other than solid state or having people carry RAIDs around. Whether you want to spec the system to a sufficient extent that you can then go on capturing the next job while offloading the last one is a choice worth thinking about.

 

The reason I bring that up is that there is a workflow issue regarding how this is databased. If you want to have more than one job on the system at one time, or if you want to use this device as a frameserver for other parts of the facility, you need some kind of job management. This is quite doable and I suspect that you would want the interactive component of this to be a front-panel for the tape room guys whose responsibility this sort of thing generally is.

 

9 pin might best be served with a "learn mode." Then, you end up building a library of what common controllers happen to think is the standard. This could be either an on-install setup option, or something workable by a reasonably savvy tape room guy. Either way you end up with a conversion table for the principal transport commands and the expected response.

 

The lack of standardisation in this sector is frustrating, yes. I think something computer-based could handle it, if set up with care.

 

Phil

[Michael Most]

The only one that's tricky under Windows at least is DVCPRO-HD. If that's what you want, that's what you should capture, using a hardware compressor, because converting it afterward is going to be unacceptably time-consuming; that's then a whole other ball game from doing uncompressed work as the input hardware is different. You don't want your DVR to spend its time doing software conversions.

 

Well, yes, but it's also the most common client requirement these days. For all of the talk about "uncompressed," in the real world people don't usually want or need it. What they want is to not have to buy a tape machine. In other words, this is ultimately all about money, not quality. As for Windows, there are already a number of these devices for both Windows and Linux (Rave HD is a very good Linux based unit). However, there is nothing on the Mac, and that is unfortunate because at least 80% of the clients we (and a number of other facilities) have requesting disk transfers are cutting on Final Cut systems, and usually want DVCPro HD or some other Mac specific codec. Which is a shame, because Apple systems are very open in terms of third party hardware support, making things like hardware support of DVCProHD a moot point (the Kona cards are particularly adept at this).

 

If people want to walk with the rushes then they're just going to have to wait; there is no solution other than solid state or having people carry RAIDs around.

 

Only if it actually has to be uncompressed, which it usually doesn't. Besides, telling clients "they're just going to have to wait" is only acceptable if there is no alternative. I hate to pull rank here, but Phil, I don't think you've worked much in the post production facility end of the industry. The name of the game is finding solutions, not emphasizing limitations, even when they really do exist. Besides, as I said, actually having to supply uncompressed files is relatively rare, because in spite of the hype, most users do not have high speed disk arrays to play these files back and are not about to acquire them. They just don't always seem to know that.

 

9 pin might best be served with a "learn mode." Then, you end up building a library of what common controllers happen to think is the standard. This could be either an on-install setup option, or something workable by a reasonably savvy tape room guy. Either way you end up with a conversion table for the principal transport commands and the expected response.

 

The command set is not really the problem in 9 pin control. The protocols themselves are pretty standard, but latency and various delays are not. That's where the problems come in. Like I said, you might want to talk to Mark Gilbert about this, they've dealt with these issues quite a bit - it's one of the reasons they went with a Midi command set.

[Phil Rhodes]

Hi,

 

The thing is, it's almost infinitely easier to develop this sort of thing for Windows; I'll look into DVCPRO-HD solutions.

 

> What they want is to not have to buy a tape machine

 

This is actually a feeling with which I identify quite strongly; avoiding tape seems to cut out an expensive and quality-compromising middleman. I'd much rather move toward doing it all as data, hence my interest in this.

 

> The name of the game is finding solutions, not emphasizing limitations, even when they really do exist.

 

Well sure, and that's why I'm an engineer not a facilities manager - but at some point, y'know, it will do what it will do, and from an engineering standpoint that's what I deal with. Of course if they're desperate for uncompressed and want it fast, you can always give them 720p24 8-bit 4:2:2, and you wouldn't be lying to them!

 

Alarmingly, very few of the desktop tools will actually do a conform from data, either - I'm pretty sure FCP won't, I know the Adobe stuff won't. Usually they expect to be recapturing the online from tape and don't have any method for doing a non-realtime assembly of what could be uncompressed data. So, it's hard to make it happen, even if people wanted to.

 

I could talk to Gallery, but they'd have no reason to help me - I'm sure they'd rather develop it themselves, and frankly probably are already doing so. They're certainly far better placed to do it than I am.

 

Variable command and response timing is not a problem - I presume you don't want the controller to say "go here" and the device to immediately come flying back with "done!". This sort of issue causes problems when trying to use compactflash cards as ATAPI hard disks; the ATA controllers get very confused when seek commands happen practically instantaneously. It shouldn't be a problem to time things out right, although I suspect that it would take exactly as much fiddling about to set up as a conventional tape suite.

 

Phil

[Michael Most]

The thing is, it's almost infinitely easier to develop this sort of thing for Windows; I'll look into DVCPRO-HD solutions.

 

Good luck, but I don't know of anything on Windows other than Avid that supports DVCPro HD. That's why we don't use something like Rave HD or Drastic to do these things in the first place. Mac/Final Cut compatibility for the generated files is an absolute must in our part of the world these days, and most likely in other parts of the world as well.

 

Of course if they're desperate for uncompressed and want it fast, you can always give them 720p24 8-bit 4:2:2, and you wouldn't be lying to them!

 

Well, you can't do that if what you're starting out with is a video signal, because there's no such thing as 720p/24 video. There's also no such thing as 720p/25 video. These are data-only formats, usually generated from cameras that either shoot them directly (i.e., HVX200) or record them to tape as 720p/50 or 60 with frame ID flags (i.e., Varicam). The Kona 3 card can do a real time data downconversion from 1080p/24 (a valid video format) to 720p/24, but that exists only on the Mac. I'm not sure whether the Xena (the PC version of the card) can do that or not.

 

Alarmingly, very few of the desktop tools will actually do a conform from data, either - I'm pretty sure FCP won't, I know the Adobe stuff won't. Usually they expect to be recapturing the online from tape and don't have any method for doing a non-realtime assembly of what could be uncompressed data. So, it's hard to make it happen, even if people wanted to.

 

If you're talking about really cheap, consumer oriented tools like Final Cut and Premiere, you're correct. And I'm not implying here that something like Final Cut is a consumer only program, but in terms of price, it is not priced as a professional product and is not supported as one either. However, Assimilate Scratch, Nucoda Film Cutter, and even Baselight are all "desktop tools" in the sense that they run on commodity hardware and commodity operating systems, and they all do data conforming just fine. You still get what you pay for.

 

Variable command and response timing is not a problem - I presume you don't want the controller to say "go here" and the device to immediately come flying back with "done!".

 

Well, that's basically what the Fostex DV40 does, and it works quite well with just about any edit controller.

[Phil Rhodes]

Hi,

 

Matrox Axio speaks DVCPRO-HD. Rather overkill for this sort of job, though. Understand about the FCP obsession of the world.

 

I believe you can do 720p/24 blown down from 1080 on a Blackmagic Multibridge (which has a very good SDK); I'm just waiting for the drives to get here before I can try it.

 

But I'm not quite getting what you're saying. Are people using these DVCPRO-HD data files as proxies for a later uncompressed conform, or is this the actual production data we're talking about? I presume both happens. Also, do you see any activity with DNxHD or Cineform? Suspect the latter may be a bit premature, but both are really quite a bit better than DVCPRO.

 

Phil

[Michael Most]

I believe you can do 720p/24 blown down from 1080 on a Blackmagic Multibridge (which has a very good SDK); I'm just waiting for the drives to get here before I can try it.

 

I don't think that's true. The Multibridge is a video box, that converts video formats to other video formats. 720/24 is not a video format. The Multibridge does not talk data formats at all, as it's not designed for that. As I said, the only device I know of that currently does the conversion between 1080/24 and 720/24 without doing a frame rate conversion is the Kona 3, primarily because they designed it as a data format converter - in other words, it can take in 1080p/24 video and make it 720/24 data inside the computer, or it can take 720/24 data already in the computer and play it out as 1080p/24 video. I can't say it enough times because there are so few people who actually understand it (you're one of them, but there aren't very many), but repeat after me: 720/24 is a DATA ONLY format. 720/24 is a DATA ONLY format. Good. I feel better now.

 

But I'm not quite getting what you're saying. Are people using these DVCPRO-HD data files as proxies for a later uncompressed conform, or is this the actual production data we're talking about? I presume both happens. Also, do you see any activity with DNxHD or Cineform? Suspect the latter may be a bit premature, but both are really quite a bit better than DVCPRO.

 

Some use it as proxies (not many). Most use it as their offline, online, and final format. The fact that it's heavily compressed doesn't seem to bother them a bit, regardless of what you, I, or anyone else might think. Sometimes it's from the camera, if they shot on a video camera. Sometimes it's from a film transfer, if they shot film. As for DNxHD, that has never yet been requested. I suspect this is primarily because in spite of Avid's efforts to integrate it into the low cost software only products, most of Avid's users are, shall we say, a bit more professional in their approach, and thus use an offline/online paradigm - meaning that the projects are finished by professionals on higher end equipment. I have never yet had a request for Cineform, and I suspect this is primarily because users have to actually pay for it - which most people in the categories we're talking about don't want to do. Not to mention the Final Cut dominance in this area in the first place. The fact that they both might be a bit superior to DVCPro in a number of ways is irrelevant. It just doesn't matter. Final Cut has won the battle of the desktop, and the technologies Apple has embraced - SATA drives, Firewire, Kona, Blackmagic, DVCProHD and to a more limited extent HDV - have ridden the wave with it. Cineform is working on a Quicktime implementation, but my feeling is that it's just too late. Any success they find will likely be as a niche player. Hopefully that's enough for them to survive.

[Alessandro Machi]

I have a low budget music video coming up where we're mixing 16mm and DVCProHD for two distinct looks. (Finishing to SD and DVD)

 

We're looking to transfer 800' of 16mm to DVCProHD and Pro50 or digibeta. A friend of mine recommended Entertainment Post in Burbank as a good affordable telecine house, but I've never heard of them.

 

Anyone used them? We're trying to find a low rate, but not at the expense of a quality transfer...

Along the same lines, are there any places in LA that do telecine direct to a hard drive? I know about Bono and SpyPost, but I'm looking for something local.

 

Thanks,

 

 

I don't have any direct experience with the Kona Card but associates of mine are really happy with it. We shot a betacam sp project and then they transferred via a component video kona card to Final Cut Pro. You'd probably want digibeta and works just as well, although you would need a digibeta deck to do the transfer.

 

Interface compression options from the kona card included firewire, DVC-Pro 50, uncompressed 8 bit, uncompressed 10 bit. Because they had several hours of live program I suggest the DVC-Pro 50 option rather than a straight conversion to firewire. They were ecstatic with the result. When they made a DVD from the timeline and then looked at it on a 50 inch television they were still ecstatic.

 

If you have a way to investigate this method it might be worth considering.

 

My concern is it could become a shortcut for transfer facilities to offer "uncompressed" transfers as a gimmick while they let their actual transfer quality drop by letting their tube get old and certain other components that need maintenance go unmaintained.

 

My opinion is a video transfer from Spectra Film and Video to either betacam sp or digi beta, transcoded by the component signal into a kona card at DVC Pro-50 quality or higher (which includes 8 bit uncompressed or 10 bit uncompressed) may be the best current combination quality and cost, available in the LA area.

[Sam Wells]

Cineform is working on a Quicktime implementation, but my feeling is that it's just too late. Any success they find will likely be as a niche player. Hopefully that's enough for them to survive.

 

If Apple "blessed" it and the price dropped it would appeal to me (in this context) -- ok there's one vote...

 

-Sam

[Thomas Worth]

Well, I am jumping into this a little late, but here are my two cents:

 

First off, in response to the whole point of this thread, Glenn at Entertainment Post is a super nice guy and very knowledgeable. When Phil and I were in there the couple times we visited, he stopped what he was doing, showed us around, and then talked tech stuff with us for about an hour. He even offered to let me bring some of my 35mm negative in and throw it up on the Shadow just to check it out. I haven't actually taken him up on the offer, but simply throwing that out there was a very nice gesture.

 

They have Ranks, but as Glenn pointed out, they're really for dailies only. He mentioned something about underdriving the tubes so they last longer, but the quality isn't 100%. Instead, he offers quality transfers on the Thomson Shadow (almost identical to a Spirit, but a step below). He has the Shadow hooked up to a da Vinci and then an HDCAM SR deck. So, if you wanted to, you could telecine 35mm to HDCAM SR. The last time I was over there, he had an unheard of hourly rate for the Shadow. You should call him up and inquire.

 

The very first thing thing out of Glenn's mouth when the subject of direct to disk transfers came up was "RS-422 control." I agree with Michael -- for this to be a reality, you need a software-based "deck." Regarding the "virtual tape" arrangement, I think we have to graduate beyond what we're used to and look at the situation from a new perspective. Certainly the whole concept of "tapes" or "reels" would not apply in the way that it does now, so then the question is, "what is the most efficient way to manage the data?" The 24 hour "tape" is a fine idea, but now the problem is, "what file format?" And if transferring uncompressed 1080p/24 10 bit 4:2:2 to disk is your plan (it should be), you will most certainly want to generate an offline DV version of this to give to the editor. Now how do you do that, while preserving the timecode from the uncompressed file? Can the timecode be STORED IN THE FILE? And if "no," WHY NOT? Seems pretty basic.

 

Of course I don't claim to be a post production veteran of 15 years, but I don't see any easy data management solution for the little guys. The days of the "tape" workflow are numbered, and it doesn't make sense that there isn't a swift and easy solution to this using off-the-shelf hardware and software. A reasonably fast PC with an 8 drive RAID 5 array (probably around $1500 to assemble), is fast enough to online uncompressed 1080p/24 10 bit 4:2:2. Of course, it's pointless to edit at that resolution, but if I did "render" a DV version of my uncompressed footage for the purposes of an offline, I would need someway to insure that both files could reference each other. As we all know, traditionally this is done by re-dig'ing the TAPES back in on a more capable system. But we're not using tapes! So what, then?

 

I'm no engineer, so go easy on me. But I do recognize the technical hurdle and would like to do all I can to come up with a solution.

[Stephen Williams]

They have Ranks, but as Glenn pointed out, they're really for dailies only. He mentioned something about underdriving the tubes so they last longer, but the quality isn't 100%.

 

Hi,

 

Running the beam current low on a Rank makes rather noisy pictures. I once worked at a facility that did that, funny thing was the tube did not last any longer but they lost most of their telecine clients!

 

Stephen

[Phil Rhodes]

Hi,

 

This all depends what you mean by a video format. The Xena 2K has a 720p24 input option which will talk to JVC's HDV cameras. So, I guess the need for it has been recognised and it's beginning to be implemented. So it's now a video format, if not a tape format.

 

No reason it -couldn't- be done, after all.

 

Phil

 

Hi,

 

This all depends what you mean by a video format. The Xena 2K has a 720p24 input option which will talk to JVC's HDV cameras. So, I guess the need for it has been recognised and it's beginning to be implemented. So it's now a video format, if not a tape format.

 

No reason it -couldn't- be done, after all.

 

Phil

[Michael Most]

Certainly the whole concept of "tapes" or "reels" would not apply in the way that it does now, so then the question is, "what is the most efficient way to manage the data?" The 24 hour "tape" is a fine idea, but now the problem is, "what file format?" And if transferring uncompressed 1080p/24 10 bit 4:2:2 to disk is your plan (it should be), you will most certainly want to generate an offline DV version of this to give to the editor. Now how do you do that, while preserving the timecode from the uncompressed file? Can the timecode be STORED IN THE FILE? And if "no," WHY NOT? Seems pretty basic.

 

The only real way to move into an IT based environment is to go "all the way." That means that one must let go of the idea of "wrapped" file formats with specific resolutions and codecs for original material and go to a system where each frame is self identifying, both in terms of resolution and metadata, which would include time code. The common file format for this is DPX, which not only carries picture information, but a bunch of metadata, including a standard location in the file for SMPTE time code. This is the file format used for virtually all digital intermediate work, in part because it's closely related to Cineon, but more importantly because files in this format can be automatically conformed, provided that the directory structure makes sense. The beauty of DPX files is that issues such as "422 vs. 444" disappear, as do frame rates and codec choices. The bad news is that they take up a lot more space, but that becomes a storage issue, not a compatibility issue. For this reason, most DI facilities (including ours) use streaming data tape backup systems, like LTO3 for this purpose. These formats allow for selective file restoration when the material is finally conformed. For offline purposes, there are a number of software and software/hardware solutions available for converting file sequences to "wrapped" movie formats, with time code stamped on the movie file that matches the original DPX files. As we move towards more data-oriented workflows, this is likely to be one model for an offline/online approach.

 

In terms of a workflow for original film transfer, however, it is necessary to basically do what I originally said - create a "virtual deck" that records the material as a series of DPX files. The DVS Clipster and Pronto are two such devices. The Drastic Technologies QuickClip is another. Both of these devices create lower resolution proxy files in just about any format (although Quicktime is the most common) for offline purposes as either a background process or a post process.