[Tim Kolb]

I think the question is indicating that there are shots with multiple camera settings that need to be adjusted.

The question is whether you can somehow cross-assign “results” instead of adjustment input.

Is the “Color Match” effect what answers the question?

TimK,
Director, Consultant
Kolb Productions,

[Tim Kolb]

On the multiple card question:

I have one CUDA capable card and one NVIDIA Quadro non-CUDA card, but I run four displays, so the CUDA card doesn’t get to do its work without running the displays, so I don’t know the ultimate answer to your question as to whether the preview acceleration will work better, or even work with a non-display coupled CUDA card. The issue would be once the frames are ready, you’d have to expend some bandwidth to move them to the other display card vs just sending them out the DVI/Display port on the card that processed them. In the case of PPro, I guess I’d question whether you’d truly gain anything this way.

[Erik Lindahl] “Also, how crippled is the Mercury Engine when it comes to output over a video-card like a Kona 3 or similar?”

“Crippled” is not really a good way to describe it in my opinion.

The purpose of PPro CS5’s CUDA utilization at this point is to speed feedback to the edit video overlay during the editing process and the main limitation would be what applications or sub-routines, or whatever you want to call them…what “functions” can be rewritten to both execute under NVIDIA’s CUDA architecture, and thread out widely enough to make that sort of insane parallel processing and the administrative overhead it takes to get the math in and out worth while.

Also…the benefit of Mercury’s ability (regardless of the existence of a CUDA card in the system or not) to drop to 1/2 or 1/4 res (which works on any material regardless of codec) for working preview playback cannot be underestimated. When you combine taking CUDA-capable effects off of the CPU’s plate with the ability of dropping to half-res and therefore cutting the pixel-processing load to 1/4 (half res is half res both vertically and horizontally), the editing experience gets very responsive.

I/O cards are a caveat at this point as far as I know. An I/O card resides in a system slot…the system slots take their orders from…the CPU. You also have to remember that those of us who use Kona cards, use them to view and analyze full resolution HD…so Kona (or I assume BlackMagic, BlueFish, etc) outputting to HDSDI is typically full resolution.

Of course, the catch is that your optimal dataflow under the Mercury w/CUDA option is

1. harddrive-video file read-to-
2. CPU-video decode-to-
3. NVIDIA CUDA card-effects processing-to-
4. DVI port-squirt out preview frames-
5. the end.

Adding an I/O card adds one backtrack to the flow:

1. harddrive-video file read-to-
2. CPU-video decode-to-
3. NVIDIA CUDA card-effects processing-BACK to-
4. CPU-send to PCIe I/O card-to-
5. HDSDI monitor.

If you need full res for your HDSDI I/O, then you also need to be handling full res all the way through…full res decode on the CPU…full res effects processing in CUDA…THEN shove each frame back down the pipe to the CPU, where it has to convey it to the I/O card…while the CPU is still simultaneously decoding video frames for the playback in the first place.

So…while I disagree with the word “crippled” as it suggests a defect in design, CUDA effects preview acceleration can only work as fast as the system can feed it frames, and my guess would be that most mortal desktop systems will bottleneck first at the CPU when in full res, multi-stream, on-the-fly video decode and back hauling simultaneous full res frame output to HDSDI vs a limitation in the ability of the NVIDIA card to process the effects.

I do think that the industry in general (not just Adobe or NVIDIA) tend to distill sales messages to a point where information gets simplified to the point of being in the least, vague, or at worst, misleading or even occasionally inaccurate.

Being at NAB in the AJA booth was interesting this last year as people would exit Adobe’s Mercury demos higher than a kite and they would come to me and expect me to run 4 layers of RED RAW 4K in RT out the HDSDI port on the Kona station I had there based on the demo they saw using multiple layers of 720p24 sub-raster DVCProHD at half res coming out the DVI port on a system with as many CPU cores as you could buy at the time. Explaining the technical caveats of what they saw vs their expectations didn’t seem to make them feel better.

I personally think that the companies involved just didn’t do enough out of the gate to really provide realistic detail about where this feature was designed to begin and end. You have guys like Todd K. trying to get those details out there, but there is a lot of ground to make up.

I will say this… Premiere Pro CS5 is the first NLE to really utilize CUDA processing (not GPU display processing…actual code executed on the GPU cores) at all, and they’ve been pretty ambitious.

In my opinion, an investment in a display card for a video editing system has never furnished a return like it does now with NVIDIA’s qualifying cards. I know I spend a lot of effort on what probably seems to many as criticizing or diminishing the feature, but I use a qualified professional NVIDIA display card on my system and utilize the CUDA option for the Mercury playback engine and appreciate the feature very much.

I think that Adobe has shown some obvious initiative in this area and I think that users will find that this sort of thing will only get better and better, we’re at v1.0 with executing actual code on the GPU. Remember how long it’s taken to get most of our software applications to really utilize 2 cores, then 4…then 8, 16, and so on. Imagine rewriting the code for a process that breaks itself into 240 threads and reassembles the results…Adobe has done that with some staple effects in this first round.

Mercury is good. CUDA augmentation of Mercury is good. The ROI is definitely there. You just have to know what it’s all designed to do so your expectations of the feature’s effect on your workflow are based on reality.

Around here, as you can see, we’re trying our best to explain it as clearly as possible.

TimK,
Director, Consultant
Kolb Productions,

[Tim Kolb]

[Erik Lindahl] “You connect displays to the Radeon-card and is your “GUI” card, the Quadro is left unconnected to external displays and is only used for processing”

Er…how do you view what you’re doing? There has to be a reference display hooked to…something.

TimK,
Director, Consultant
Kolb Productions,

[Tim Kolb]

These days DaVinci technical info is a little challenging to find on the BlackMagic website…so I’m not sure what you’re referring to…

“GUI” stands for Graphic User Interface, which any piece of software with mouse interaction has, so I’m guessing you mean something else.

As I noted in several threads below, GPU support is the visual processing augmentation that is facilitated in Open GL and Open CL, whereas CUDA is a protocol for actually moving the executable software to the GPU.

TimK,
Director, Consultant
Kolb Productions,

[Tim Kolb]

Well…it does speed encoding and decoding and responsiveness and all those things…it’s just important to note that it does this by handling the effects computation.

The only time when this becomes an important distinction is when someone simply has a computer that can’t get the video decode work done…even without effects (we’ve had a couple of those in the last couple days here). In those cases, CUDA won’t be the difference between being able to play the video back at full frame rate on the timeline and not. It can make the difference between being able to play it back with those specific effects (color correction being a big one) and not as once CUDA picks up that work, more CPU goes to video decode.

The bottom line is that encoder/decoders (codecs) need to be rewritten to thread out into 200, 400, 600 threads and back to run on CUDA…at the moment all the video codecs that I know of are written to run on a CPU.

Adobe rewrote a quantity of effects to run on CUDA, which was the necessary step to getting all that to happen.

Also…GPU acceleration is not the same as CUDA (we have a person who asked about GPU acceleration when what they want to know about is CUDA). GPU acceleration takes care of accelerating visual display functions, computing pixels and texels and shadows and transparency, etc, etc. CUDA actually runs the -executable- application on the GPU cores. The GPU becomes a super-threaded CPU basically.

All of it’s cool and beneficial, but it’s good to understand the differences. For example, After Effects’ GPU accelerated previews or third party plugin makers with GPU accelerated effects are not using CUDA/Mercury…they’re using the traditional GPU functions.

…which actually isn’t all bad as GPU use in general uses any display card with a GPU you put in the computer…

TimK,
Director, Consultant
Kolb Productions,

[Tim Kolb]

[Kevin Dietrich] “Also the 4k files are scaled to fit the 1080 sequence so watching that at 1/8 is very low quality when I want to see how my color correction and effects are blending.”

Where the files shot at “HD for 4K” (3840×2160) or regular ole’ 4K (4096×2304) …or are they Mysterium clips shot at 4.5K (4480×1920)?

…and you are using a RED sequence preset so the frame interpretation is correct, yes?

TimK,
Director, Consultant
Kolb Productions,

[Tim Kolb]

It’s a dead-on match for me…

Are you certain there wasn’t any scaling of the text layer in the AE project?

Maybe you could fill in all the details of what you’re doing and then figure out what’s up…

TimK,
Director, Consultant
Kolb Productions,

[Tim Kolb]

Yes…CUDA misinformation still abounds…(sigh)

Keep in mind that “HD” is less than 2K…and it isn’t RAW. You have RED RAW 4K…5X HD in pixel count, and faaaar more computation necessary to decode RAW than say DVCProHD (very easy to decode, sub-frame resolution, low bitrate, and what makes up the bulk of most Adobe demos), XDcamHD, or even 10 bit CineForm or ProRes.

Keep in mind that CUDA acceleration takes over ‘effects’ processing for CUDA designated effects…it DOES NOT (I’m going to make a notepad doc with all this typed out as it’s getting to be 2 or 3X a day for me lately)…it DOES NOT DO VIDEO DECODE.

Your CPUs are what needs to do the video decoding…CUDA will belp because now your CPU doesn’t have to do the math for that color correction or other effects and it can focus on video decoding…but the CUDA card will not accelerate straight, unaltered video playback.

I don’t know of anyone who can achieve full frame, full res, full rate 4K RED RAW playback on any desktop system…you can use full res on pause to check things like focus, etc, but even a 30″ DVI-driven LCD only does 2K widescreen pixel-for-pixel, so there really isn’t a way to view it.

The beauty of CS5’s ability to handle RED RAW is not that it plays back at the full frame size, it’s that it will run at 1/4 res and allow you to edit to cut down the scene, in motion and in context while it’s still in RAW and you haven’t had to bake anything in as far as a primary color correction before you convert to some other format like ProRes…or worse, converting, then color correcting on the flattened image format.

You get to cut it, use the Adobe interface or RED CINE to make those primary color corrections that are embedded into the metadata on decode on the -selects- used in your edit because you’ve had the ability to cut it down as RAW…and then you can send something downstream that’s ready for higher end finishing like DPX files with linear or log curves, over-range color values, etc… Which I do not believe is available on any other NLE option (though I will gladly accept correction on that if I’ve overlooked something…).

So…you have a very good workflow there…maybe the best for RED available at the moment…but 4K is still 4K, it’s just a LOT bigger than HD. That doesn’t mean you don’t have a very effective toolset in front of you, just that you have to know what to expect.

TimK,
Director, Consultant
Kolb Productions,

[Tim Kolb]

[Warren Morningstar] “Thanks Tim. Appreciate all the advice you freely give on this forum.”

Well…thanks for the thanks.

The heads you see up top are definitely the brain trust around here. I’m not around as much as I once was…

Andrew’s tutorial portfolio for PPro is extensive here at the Cow and he does a nice job of subdividing topics to keep the tutes relatively short and focused, so that’s a resource that shouldn’t be taken for granted.

So…lots of help freely given around here…by many…

TimK,
Director, Consultant
Kolb Productions,

[Tim Kolb]

[Steve Brame] “Couldn’t agree more. There is actually a selection, even if you have an approved card, to turn off the Mercury engine and force CS5 into ‘software mode’. The selections are mislabled. They should read…

1 – Get work done
2 – Watch paint dry.”

I’m not sure I understand…are you advocating shutting CUDA acceleration off?

“Mercury Playback Engine” is the technology used for playback 100% of the time whether you have a CUDA card installed or not.

I’ll just stress again, in case anyone still misunderstands, CUDA acceleration speeds previews by taking over the effects computations of the CUDA-indicated effects (you’ll see a designation for ‘YUV’ effects which don’t transcode the video to RGB to work, you’ll see ’32’ to indicate those effects that work in 32 bit float, not truncating your RED RAW footage or AVC-Intra 10 bit footage to 8 bit… and the little arrow symbol indicates CUDA acceleration).

It can help a computer decode video more effectively as it takes over those effects so the CPU can focus on video decoding…but CUDA accelerated operations do not help directly with video decode (nothing says it may not in the future…but it’s deployed for effects preview at the moment).

So in theory, if all someone was doing was cutting footage and doing nothing to it at all (color correction, etc.), a CUDA card will actually show very little effect…not that it’s likely that many of us work this way, but the point is that some beef in the CPU department is still helpful.

I love what CUDA acceleration does in PPro CS5 BTW, I’m not recommending against getting a CUDA card, I have one and I depend on it…I just want a person who’s going to spend money to understand what benefit they’re receiving.

TimK,
Director, Consultant
Kolb Productions,