|
Forum List
>
Café LA
>
Topic
Why does ProRes choke my machine?Posted by chad haberstroh
"Chokes the machine" in what way? Beachballs? Crashes? Slow response to commands? Stuttering playback? How full are all your drives -- including drives that aren't being used in the project, as well as the system drive? (Don't say "I have plenty of space"; give numbers.)
www.derekmok.com
chad haberstroh Wrote:
------------------------------------------------------- > I'm mainly working off of a terabyte ethernet > server That would be my first look. Ethernet works fine as a way to transport video to editing systems, IF... Can you run something like AJA system test on the network attached drive. What is the average speed, and if you graph it, is the data delivery smooth, or does it have a lot of noticeable spikes in performance?
The only other thing that comes to mind is-- check to see if your material is 10-bit, not 8-bit. Everytime I load 10-bit material into the Mac Pro it stutters and craps, stutters and craps. It's like editing at an assisted living community. Check the clip Inof, your sequence settings info and the Video Processing pane in System Settings.
Transcoded the material to 8-bit ProRes, checked it out on the large consumer flat via Blackmagic Intensity, looked great. Proceeded to edit smoothly. And yes, I have an internal RAID 0 for this stuff. 10-bit....argh. It's for the next gen of machines. I even asked Grant Petty and BMD to include an 8-bit ProRes and DNxHD codecs for download from the highly promising Cinema Camera. (If enough people chime in, he'll listen). 10-bit is beautiful, make no mistake. But overkill for so many non broadcast corp/ed/industrial/PSA jobs where you generate nothing that's going to band in the colors. Best, as always, Loren S. Miller www.neotrondesign.com Home of KeyGuide Central
Really, 8-bit ProRes? Is there such an animal? The Apple ProRes White Paper of July 2009 makes no mention of an 8-bit version. I've assumed that ProRes treats 8-bit material by multiplying its values by four (or adding two binary zeros) making it 10-bit. When I transcode 8-bit material to ProRes 422 it has the file size expected for ProRes 422. When I examine the .mov file information with QT-Edit it doesn't mention 8-bit or 10-bit. When rendering in FCP7 there's a choice between 8-bit and 10-bit YUV rendering. I've assumed that this is a choice between computational precisions rather then coding. For an experiment I started with 8-bit material transcoded into ProRes 422. For an "effect" I set scale to 50%. This will average the values between pixels, which can result in 10-bit (really 9-bit) outcomes from 8-bit. Sure enough rendering with the 10-bit option took 1.5 times as long as rendering with the 8-bit option. But the first render file was a smidge smaller than the second one.
I don't know what to make of this statement from Apple. If the codec had an 8-bit version, or included a flag saying "keep it 8-bit", then why did the mentioned rendering experiment turn out as it did? If another application is used to do the 50% scale "effect" will the 8-bit become 10-bit or not? When an 8-bit image becomes 10-bit and returns finally to 8-bit through a chain of effects the result is slightly different from its staying 8-bit, so experiments can be devised to test the statement.
This statement in the more authoritative Apple source must be taken seriously. So is there 12-bit ProRes as well as 10-bit ProRes and maybe 8-bit ProRes too? Or is there some indefiniteness about bit depth in ProRes codecs? To answer the original questions about 8-bit ProRes it might help to understand how ProRes handles 12-bit. Dennis Couzin Berlin, Germany
@Josh B. What we've always understood might not be so. What do you make of quoted statements from the Apple ProRes White Paper?
Well in regards to the first quote i've always read it as "doesn't matter if its 8 or 10bit, its preserved" ie if your footage is 8bit then your not loosing anything as its a 10bit codec, and if your footage is 10bit your still not loosing anything - apples way of saying "hey you don't need to worry about it, either way we got you covered - but just don't ask us for any details"
The second quote i find extremely confusing! Its interesting that after several concerted efforts to find definitive information on the technical aspects of the codec the closest i can find is in the white paper - and that is far from comprehensive!
Sorry, only registered users may post in this forum.
|
|