[Erik Lindahl]

I very much think ProRes 4444 will find it’s way into different areas for sure, especially if it holds similar compression levels as HDCAM SR which is accepted as a high-end intermediate format. Given it holds similar quality as HDCAM SR, it will however still take some time until we really can utilize it as say a HDCAM SR replacement.

It’s defiantly not a MOOT point to discuss this however rarely do you capture, edit, master and deliver in the same format. Our general high-end workflow today for TV-adds are:

[Recording] 35 mm film > [Transport] HDCAM SR > [Edit] Uncompressed QuickTime > [Output] MPEG2

We then keep a Digibeta and AIT-4 tape master (AIT-tape holds the Uncompressed QT files).

The future could very well mean you erase HDCAM SR and Uncompressed QuickTime IF ProRes 444 holds good enough quality. With a tool like the KI Pro, interesting things CAN happen in the future.

Erik Lindahl
Freecloud Communication
————————

[Erik Lindahl]

Well one issue we have today is a lot of post-intense jobbs we do look worse in the end going down the Uncompressed 4:2:2 10-bit road than using Animation RGB (4:4:4 of course) even when our final output is Digibeta or 4:2:2 MPEG2 files (30 mbit i-frame). I think this is due to Apples 4:2:2 codec having no chroma filtering hence keeping 4:4:4 through our post-work-flow and letting either Episode or our Kona 3 card do the down-sampling to 4:2:2. The end-result is better.

Moving to ProRes the 444 codec would be the interesting thing there.

1. Color resolution
2. Choice of working in RGB

Moving to and from applications like After Effects a RGB codec is by far the preferred as rendering times and issues accrue with Uncompressed-non-RGB media. But perhaps the ProRes 4:2:2 HQ codec offers the same choices now?

Erik Lindahl
Freecloud Communication
————————

[Erik Lindahl]

Very solid replies Gary!

My first question is how well ProRes 4444 will perform in current established post-chains we deal with here when we scan film. We’ve used one of two workflows:

a) Film > Telecine > HDCAM SR > Uncompressed 10-bit 1080i50 in FCP
b) Film > Telecine > HDCAM SR > Digibeta > Uncompressed 10-bit PAL FCP

The Telecine we in general deal with state they scan ALL their work to HDCAM SR. I can’t testify if this is the general 440 Mbit 4:2:2 setting or something else however.

Now, if we where to ask for ProRes 4444 from the telecine I’m quite sure their workflow would be:

Film > Telecine > HDCAM SR > ProRes 4444

Wouldn’t this just introduce double compression transport stage? If a device linke the AJA KI Pro where to replace the HDCAM deck I think a small “revolution” might be in place IF ProRes 4444 holds it’s ground.

All the above is talking about ProRes 4444 a transport codec. This will also always be file-based with it’s pros and cons. My second and main question is, since it’s what I deal with a lot more on a day-to-day basis, it’s ability to work as a finishing format compared to Uncompressed or Animation across generations and in heavy post-works flows.

Erik Lindahl
Freecloud Communication
————————

[Erik Lindahl]

Just cause function A used multi-core function B does not. It’s been a source of performance issues for quite some time in a lot of applications. Say FCP might be able use say 8 cores when it plays back 20 streams of ProRes media. That doesn’t mean rendering a 3-way Color Correction on 1 clip will utilize the machine to the same extent.

After Effects has solved this problem with multiple instances of it’s self in the last two versions. This is similar to how Compressor solves it this way also. This is general gives a solid speed increase but not always.

Erik Lindahl
Freecloud Communication
————————

[Erik Lindahl]

The codec is multi-core “aware” and very scalable, that doesn’t mean the host-application is. I think about 5 people have quoted the ProRes spec in regards to FCP.

Erik Lindahl
Freecloud Communication
————————

[Erik Lindahl]

One wan always have hopes but that PDF is regarding a specific codec and not an application. But ProRes sounds like a very well thought through compressed codec. I also think it’s something Apple would really push if it was something that the Final Cut Studio applications took advantage of.

Erik Lindahl
Freecloud Communication
————————

[Erik Lindahl]

Hey hey, don’t complain now, Apple has introduced the new iChat theater feature for OSX Leopard in Final Cut Pro 7. Oh, right, that’s a “killer feature” for yesterdays OS release…

I like you Michael feel pretty much the same. It’s sad but I think it’s the truth. I was hoping Final Cut Pro X (what ever they where to call it) would be a small, new, NLE revolution with the advent of MacOS Snow Leopard. They’d raise the bar of the studio-package, require Snow Leopard AND an Intel Machine but really give it a new face list in a wide area.

Then again, as you say, releasing the “next big thing” is a bit risky given Apple has done this in a past with the whole OSX and later Intel transitions. These went relatilvy to very painless for the end users even in critical environments. I would also think it would be easier for a developer to focus on a new minimum required system (i.e. require Snow Leopard and Intel but also give features accordingly).

If they do anything for FCP 7.5 they should do what they are doing for Snow Leopard: faster, fine tunned, stable and bug free. Make FCP feel like a modern, fresh application.

Erik Lindahl
Freecloud Communication
————————

[Erik Lindahl]

An important thing to differ is however comparing equal codecs or formats vs on-another. DPX vs TIFF hold similar abilities. TIFF vs PSD hold similar abilities. 16-bit TIFF vs 12-bit ProRes, in pure quality, the former will most likely always win. But you end up comparing the container as well and a bunch of other things.

We’re an FCP house also and well… we’ll see what happens. There aren’t really any other options out there at the moment (Premier is as far as I’ve understood still playing quite a lot of catch up to FCP given it has developed extremely fast).

Erik Lindahl
Freecloud Communication
————————

[Erik Lindahl]

Delivery is of course a different story all together. I primarily question ProRes as a true high-end post format. Advertising-delivery in Sweden now fully file-based. MPEG2, i-frame only @ 30 Mbit and WAV audio. Our Digibeta use has gone down a ton I can tell you. Sad part here is Compressor can’t deliver legal files for this (Episode can however).

Erik Lindahl
Freecloud Communication
————————

[Erik Lindahl]

ProRes HQ and 4444 are considered to be a lossless codecs. If not, it’s as close as you’re going to get to it for now. If this doesn’t work for you, then you’re going to have to come to the realization that FCP’s uncompressed days are over. If you need uncompressed, you need to now look for something else to fill your need.
I beilive ProRes is “virtually visually lossless”, this has nothing to do with actually being lossless. And that FCP’s uncompressed where over, well, then they will lose a ton of serious post-people. No offence but that’s one of the most silly comments I’ve heard…

When dealing with DPX, RED or other 444 formats, the main problem with ProRes HQ wasn’t the compression.
I can’t speak for the nature behind DPX compression, but RED is a very heavily compressed codec. 12:1 or something like it (far more than DV). This is a HUGE issue behind using RED, not the 4:2:2 color format.

ProRes 4:4:4:4 is an odd-ball with it’s 12-bit but quite heavy compression-ratio. Also, comparing this to image-sequence formats is hard since these work very poor in Final Cut Pro (also something I was hoping Apple would look into either directly to FCP or just add solid support to QuickTime).

For FX work I’m very iffy about the RED-camera in the first place sadly. Greenscreen-work is pure hell with the camera. It, as ProRes, has it’s place in the industry – no question about it – but no man in their right mind will post in 4K on a low-budget level. 2K at post, but most likely to stick to 1080p25. I’d rather play with uncompressed 1080p than compressed 4K.

I will likely buy the upgrade and test it at home but since the new feature set is questionable and we lose PPC support I think the real “use” will have to wait.

And I’ll repeat that I think it somewhat a failure Apple would think the high end will settle for something second best. The phase “go look for something that suits your needs” shows up a lot in the discussions with people being somewhat let down by Apple’s latest “accomplishment”. It’s a bit I guess foolish thinking “well if you don’t like it go away!”. That’s what’s made key players fall hard before. The problem is I know Apple can do the tools if they set their minds to it. It just seems they’ve forgotten a bit about one of their “killer” apps once upp-on-a-time.

Erik Lindahl
Freecloud Communication
————————