[Sonic 67]

[John Rofrano] “DV uses 4:1:1 color sampling and DVD MPEG2 uses 4:2:0”
That’s what’s sad.
I thought that because DV uses 4:1:1 subsampling in NTSC/60 Hz variant and 4:2:0 subsampling in PAL/50 Hz variant, DVD MPEG2 will do the same.
But no, they know better 🙂

[Sonic 67]

Try to use the two-pass encoding.

[Sonic 67]

Well, after a lot of reading, eventually I had it made. But GTX480 remain power supply hogs, needed a bigger power supply for it anyway.

Also the GTX have only 1.5GB, not 6GB ECC like the real Quadro 6000, is not like memory will grow up magically 🙂

[Sonic 67]

If you care at all about the data, stop using RAID 0. It doesn’t help in editing speed and just multiplies chances of failure, especially on big drives.
WD Red drives are optimized for NAS and RAID in ventilated enclosures, but RAID0 in a PC is not really the proper use.

I am running now a RAID5 with three drives on Intel ICH10R and never had an issue. But those are not cheap “desktop” level drives.

However, I did start to see problems with both my WD and Seagate drives (not RAID, backup externals) just because… they lowered the price/GB. I think that in order to compete pricewise, they increased the density to unsustainable levels, especially on bigger hard drives, resulting in lower MTBF and shorter warranties. Heck, for some of <4TB drives they needed to lower the RPM from 7200 to 5400 just to be able to keep the errors low – for the brand new drives!!!

MTBF has also hidden connection to Non-recoverable read errors per bits read – WD RED has listed 1 per 10^14 bits.
If you have two drives, one 2TB and one 6TB, with same MTFB numbers, the 6TB one is three times more likely to fail in the same span of time. When you added two of those in RAID 0, you just doubled again the chance of failure.

https://www.raid-failure.com/

[Sonic 67]

Can you try using the GPU-Z (you don’t even have to install it) to see the actual GPU utilization during rendering.

I have a HD 7970 (re-badged now as R9 280X) and it is not used at all in MainConcept encoding. It detects “OpenCL” in system, runs at 25% at beginning and after that, just short bursts (probably because real-time display).

What it uses it slightly, is the Sony encoder.

0_mainconcept_hd7970.jpg
sony_hd7970.jpg

[Sonic 67]

Haha Sony supports OpenCL 2.0 – translation it is “it will work because is backwards compatible”.

About the transfers between GPU and CPU: I had modded a GTX480 into a Quadro 6000. That activated a hidden bi-directional DMA between system memory and card memory – as a gaming card it have it active only in one direction, from system to video memory. I think that helps with computing tasks slightly.

However, even on pure CPU encoding, my CPU (6 cores, HT, with 12MB shared L3 cache on board) is utilized only at 50%. More strange, from my 12 virtual cores, two are “parked” constantly and another one toggles between “parked” and “in use”.

I always thought that video encoding a highly parallel process and will scale perfectly. After all, you can process hundreds/thousands of segments at once, picked between I-frames (key frames).

[Sonic 67]

I just test a random file – render it with Sony encoder that supports OpenCL.
During this rendering I used Windows Resource Monitor and GPUZ to monitor usage of various parts of my PC.

1. HD7970 GPU utilization – maximum 24%, but it was variable (sometimes to zero). Average is maybe at 15%.
With my previous nVidia GTX480 (modded as Quadro6000) card and MainConcept it was at 35-40%, but very constant during rendering.
2. CPU – my Xeon 6 core with HT (12 logical) was used at 38%.
3. Memory – from my 15GB DDR3, Vegas was using only 1.2GB (it can be selected to show in Resource Monitor)
4. Disk utilization – below 0.5MB/s. My rendering HDD is a RAID5 from 3 HDD that can push 70-90 MB/s.

I am puzzled of the bottleneck…

[Sonic 67]

Yes John, I was aware, I just wanted the HD 7970 for other things too and reported my findings in Vegas…

I have one more x16 slot available, I might use another HD 6950-6970 just to use it for MainConcept…Even if Sony encoder seems to do a decent job.
I wonder when Sony will push DivX to update their encoder software or… drop it all together form Vegas.

However I am baffled why the utilization of GPU cannot go higher (CPU wasn’t used at max either).
Wonder if Windows 8.1 can use more efficient the video card (WDDM 1.3 versus WDDM 1.1 in Windows 7)?

[Sonic 67]

My experience is not that good. Just replaced my nVidia Quadro 6000 (actually was modded from a GTX480) card with an ATI HD 7970 one (equal to R9 280X).

Encoding times of my test file @ 1080-60p on a PC with CPU Xeon X5650, 15GB of memory:
1:25 min. GPU Quadro 2000 @50% MainConcept using CUDA
1:18 min. GPU Quadro 6000 @34% MainConcept using CUDA
1:16 min. GPU HD 7970 (GHz edition) @39% Sony Encoder
4:18 min. GPU HD 7970 (GHz edition) @0% MainConcept OpenCL (didn’t make use of GPU)

[Sonic 67]

Cyberlink’s PowerDirector accepts mkv files containing video encoded with avc/h264 and having multiple audio streams. And it can output them too.