Delivery dates

[Jason Rodriguez]

Hi Joe,

I'm sure Steve will get back with you in a timely manner . . . right now we're really busy getting ready for the immient release of the camera (can't tell you when, but soon), and as a result, there may be a little bit of delay in response-but we are not intentionally ignoring anyone here.

[Joe Carney]

He contacted me. Thanks.

[Obin Olson]

How do the camera images hold up aginst the 1080p 3ccd cams on the market?

and the price for the entire camera? not the head unit, camera with body and viewfinder?

[Steve Nordhauser]

Obin,
You can see the images on our website under "Gallery" and decide for yourself. We have been met with enthusiasm but we encourage you to make qualitative evaluations.

For now, we are keeping pricing off the boards but it is similar to what was announced 6 months ago. There will be a new round of press releases within a month as the product is finalized. Lots of new features have been added.
-Steve

[Obin Olson]

So am I right in thinking that it will be more money for the camera body then the RED?

[Steve Nordhauser]

Originally Posted by Obin Olson

So am I right in thinking that it will be more money for the camera body then the RED?

It hasn't been clear to me what you get for how much when RED is ready. Our basic camera includes the viewfinder, 4 hours of media, extra drive carriers, shipping case, the Cineform RAW codec and we have a bundle with full editing tools for not much more. This is a complete and ready to use camera (add batteries and optics).

People who understand the Cineform workflow recognize the value that this brings to the SI-DVR camera.
-Steve

[David Newman]

Thanks Steve.

[Jason Rodriguez]

Originally Posted by Steve Nordhauser

People who understand the Cineform workflow recognize the value that this brings to the SI-DVR camera.
-Steve

I think there's one pricinpal that everyone is missing right now with any other camera's proposed RAW workflow compared to CineForm RAW. You Obin, with your involvement in the Andromeda project, should be very aware of what I'm about to talk about.

The situation is this . . . I've shot my footage, it's in a proprietary RAW format and I need to edit/composite/etc. with it . . . "How do I get access to my RAW data?"

Do you sit in a conversion application which is one large step removed from your editing/compositing application and render the footage to another codec that will then serve as your "working" format?

Let's say you take this option.

If 1 frame of RAW data takes 1 second to render over to your chosen actual "working codec", it will take 24 hours for 1 shot hour to be ready for editing, etc. That's at 24fps. It will take 30 hours at 30fps. Now I know you will tell me that Moore's law will remove that need, but honestly, even if you were to halve that amount and it took 0.5 seconds per frame, you're still looking at 12 hours for conversion of 1 hour at 24fps. Just as an indicator, Dalsa at Cinegear expo was showing off their RAW conversion tool of their 4K Origin frames on some fast systems . . . depending on the system, it takes them 5-7 seconds per frame!. That would mean it will take 5 days to convert only 1 shot hour of footage on a single machine! Now of course you can reduce those times by throwing a render farm into the mix, but realize this . . . decent render farms are not cheap. You cannot just set one up in your basement without proper cooling, power, etc. For instance, your standard dual Woodcrest Xeon server (let's say an Xserver) will consume around 350W total at 100% CPU. An Xserve RAID will consume 480W. A Sanbox 5200 fibre switch will consume around 120W of power. You can only add 4 machines and a RAID with associated fiber and gigabit metadata switch to a standard 20amp house circuit here in the U.S. Plus the cost of such a system will run you around $50K when you're all said and done. In addition, you've gone from being the artistic film-maker to being a technical data-wrangler.

Say you don't shoot RAW, and shoot directly out HD-SDI or to a 4:4:4 RGB codec, etc.? Well, in that case, you've now lost all the benefits of RAW, including the non-destructive nature of RAW conversion and codecs.

If you go to a "working codec" because you can't edit the native RAW codec, you're also stuck in the same situation . . . if you use something like DVCProHD because of it's real-time streaming capabilities, when you want to make adjustments to your footage with the maximum amount of dynamic range at your disposal, it's back to the RAW conversion program. Furthermore, you need to go through an offline-online workflow scenario, which will cost more time in rendering, and there will be the need for expensive fast disk-space as the online format will most likely be some form of uncompressed 4:4:4 RGB which takes up 180MB/s+. If you want to edit that in multi-stream, it's not going to be cheap . . .

With all the workarounds and headaches mentioned above, you start to wonder why [i]anybody[i/] in their right-mind would settle for a proprietary RAW conversion applicaton workflow and wouldn't want to have direct access to their RAW data in all their favorite applications, and not have some intermediary step in-between?

CineForm RAW gives you direct access to the RAW data in a universally accessable AVI or Quicktime wrapper, it gives you 10-bit precision with 4:4:4 RGB decodes and 32-bit floating point internal processing so no data is ever clipped, and furthermore it gives you real-time multistream editing without the exhorbant costs associated with uncompressed and the disk arrays required to move that stuff around. All you need is a modern-day desktop computer. $3K for a new MacPro is a lot easier to swallow than $50K for a render farm so you can see what you shot by the end of the day.

Frankly, RAW conversion applications are worse than the film-lab . . . I'll get my film dailes back faster and cheaper than spending money on render farms, etc. to-do the same thing.

The question shouldn't be do I save $3K or whatever on the price of a camera, but "What do I save in my very precious and expensive time", and "How much does the workflow and associated infrastructure cost me?"

If time isn't important to you, then I guess there is no argument. I think for many working these are very important questions though, and the fundamental flaws associated with RAW conversion applications will rear their ugly head in the near future. So in the end, you purchase a RAW camera, but you'll never be able to take advantage of all the benefits that RAW could deliver.

[Nate Weaver]

I thought of that too when I saw the RED flowchart released at IBC. You can't go converting everything BEFORE you edit. That's nuts, and very inefficient.

I'll give them the benefit of the doubt and assume for now they are not dumb, and know this. All that would be required to get around this would be to make a Quicktime codec that can do a quick display of the data stream, but not at full quality/bit depth, etc.

[Jason Rodriguez]

If they put it in a Quicktime wrapper that decodes to an "offline" proxy, you're only delaying the pain if you still have to go through a conversion application and not your editing/compositing application when it comes time to get access to the RAW data.

You're still back to an offline/online workflow . . . you'll still require super-fast disks for the online file formats, and if the RAW conversion application can't ingest your timeline like a Final Cut Pro XML or Adobe AAC file, you're stuck having to figure out timecodes and export clips one-by-one from your timeline for conforming (so you don't have to re-render all your footage for online). This still sounds very messy to me.

I can see that you might be able to edit the offline files in Final Cut, etc., but it's matching back to the online that's going to bite you. You will have to export one-by-one from the timeline a Quicktime Reference movie file (so you don't destroy the RAW data in the codec), then import that into the conversion application, batch convert it to an online format, and then re-conform.

The only advantage to this process is that you're trimming down the amount of footage that needs to be converted, and you can do an offline "creative" edit, or at the very least pick your shots. But in the end you're still back to offline/online, conforming, etc., all of which waste time and money, and require significant hardware investments in fast storage, especially for multi-stream uncompressed RGB 4:4:4 work.

To simply look at the price of the base camera and make your final decision misses the whole point of what it takes to actually work with that piece of equipment, and how well it integrates into common workflows.

[Nate Weaver]

No, I was talking about the concept of a codec that presents two different faces to different apps.

A quicktime codec that handles things in the fastest way possible, for quick decoding to screen or whatever. And then a more complete codec implementation for their conversion utility.

I should say, this is the way I've always understood Cineform raw to work. Cut in the codec in Premiere or whatever, then finesse the data in the codec after you are done cutting into an uncompressed format.

[Jason Rodriguez]

Oh No!

That is *definitely* not the way that Cineform RAW works. There is absolutely no need for offline/online or re-encoding to an uncompressed codec anywhere in the pipeline . . . the Cineform RAW codec is already a high-quality 10-bit online codec with 4:4:4 RGB decoding and a 32-bit floating point internal processing pipeline of the RAW data. Furthermore metadata control over your RAW data is available in your primary working applications. You shoot, edit/composite, and output to whatever or wherever, all with the same file . . . there is no need to convert to anything else.

Read the white-paper here for a better explanation:

http://siliconimaging.com/DigitalCin...Whitepaper.pdf

The main problem is when those "two different apps" you mention are not your editing/compositing programs, but is a separate conversion application that requires rendering into another online codec. That means you have to jump out of your primary working application into another app to get at the RAW data, and then re-import that back into your primary working editing/compositing application-basically offline/online. Cineform RAW completely avoids that scenario.

[David Newman]

Thanks Jason, you're reading up. ;)

Nate, the beauty of the system is that it feels like you are editing like normal (i.e. DV-ish), yet the engine hides all the high dynamic range processing and bayer demoasicing filters, and bunch of other secret sauce we aren't discussing yet. Raw is supposed to be hard, that's why we have the Dalsa Origin, Arri D20 and Red workflows, each offering (publicly) variations on the same pre-processing workflow. To prove the flexibility we put our own money on the line (the entrance fee :) ) and used a SI prototype camera to shot in 48 Hour Film Project (http://48hourfilm.com/). It is a hard enough getting this done when shooting DV to tape, yet we did it with a prototype camera, a $50 lenses off ebay and rangling a toddler in the shoot (I couldn't get a baby-sitter, so she needed to star) -- we shoot to disk and onlined directly in 1920x1080p24, all in RAW, I think faster than if it was DV. We also did well because we had an awesome team -- thank guys! See how we did here http://www.cineform.com/48hour/index.htm

[Nate Weaver]

Aha. It seems I'm understanding it correctly. Jason, I wasn't talking about an offline/online situation. I was talking about how a Cineform-like workflow would be implemented in a Mac environment.

When I was describing was multiple ways to decode the same codec datastream. One is a quick/easy way for most QT apps, and one for really extracting the raw data.

Regardless, as unfortunate as it may be for Cineform, it looks like the same style of RAW workflow may be co-opted for other systems down the line. On the longest timeline, I'd say it's likely that that might be where it's all going.

[Jason Rodriguez]

Yes, doing the typical "conversion app" approach, the 48 hour film-festival would have been over by the time you'd be ready to begin editing . . .