My project is about 1hr 15 mins long, covering 8 sequences.
3 days ago, I encoded using the MPG2-DVD preset and burnt a trial disk with an Encore project.
After reviewing the disk, I made a few trivial changes to the PP project and encoded again.
Exactly the same encode settings except I checked Maximum Render Quality.
This time, it took FIVE TIMES longer to encode, and I cannot see any difference in disk quality.
What does checking Max Render Quality do?
Would you expect it to take 5x?
And would you expect a better quality disk?
Thanks
An excerpt from this help section (although this help section is about creating sequences, section about rendering redirects here as well):
Maximum Render Quality Maintains sharp detail when scaling from large formats to smaller formats, or from high-definition to standard-definition formats. Maximum Render Quality maximizes the quality of motion in rendered clips and sequences. Selecting this option often renders moving assets more sharply.
At maximum quality, rendering takes more time, and uses more RAM than at the default normal quality. Select this option only on systems with sufficient RAM. The Maximum Render Quality option is not recommended for systems with the minimum required RAM.
Maximum Render Quality often makes highly compressed image formats, or those containing compression artifacts, look worse because of sharpening.
You may also want to check this discussion.
Jim,
MRQ does not force GPU processing back to the CPU. Try it out with hardware MPE on. There is no difference in GPU load or memory usage with MRQ turned on or off., and there is no difference in performance. All MRQ does in software MPE mode, is improving the quality of the export as if it were in hardware MPE mode, but of course it can only use the CPU. As a consequence it takes much longer.
Well I admit, I'm confused here.
I took a one minute 1080 clip in a matching sequence with no cuts, no transitions and no effects, and exported to MPEG2-DVD. Under those conditions, the only place for CUDA to come into play is the scaling. Here's what I saw:
CUDA On
MRQ Off - 1:00
MRQ On - 2:35
CUDA Off
MRQ Off - 1:30
MRQ On - 4:40
We know that CUDA uses the best scaling by default. MRQ doesn't 'improve' the scaling method with CUDA turned on. And we know that with CUDA off, a lesser scaling method is used by default, and MRQ does improve the scaling. So I'd have expected to see the second and fourth times come out fairly equal, since in this export, the only thing CUDA was doing was the scaling. Yet that is not the case. It seems that MRQ is doing something, even with CUDA turned on. I believed it was forcing the scaling back onto the CPU, but the times show that some other factor is at play here.
Whatever the explanation, turning MRQ on does cause a significant slow down, with or without CUDA.
Jim,
Bill and I are testing this extensively for the new PPBM6 test and on our current time-line, which is an AVCHD 1080i-29.97 source with numerous effects, fast color corrector, brightnes & contrast, gamma correction, gaussian blur and 3-way color corrector and speed slowed down by 50%, for a total duration of 2;39;04 exporting to MPEG2-DVD with a preset of NTSC 23.976 Widescreen High Quality I have just tested again the export times and it gave me the following results:
| System | Hardware MPE On, MRQ Off | Hardware MPE On, MRQ On | Software MPE, MRQ On |
|---|---|---|---|
i7-3930K, GTX 680 | 24 s | 24 s | 436 s (94 without MRQ) |
i7-2600K, GTX 680 | NA | 33 s | 870 s |
i7-980X, GTX 680 | NA | 30 s | 556 s |
Now this test is taxing on the GPU, because there is frame blending, scaling and blurring going on during export, but I can not see any difference with or without MRQ with hardware MPE turned on. I ran several runs and they are consistently between 23 and 24 seconds on this test. Several things of interest here, the advantage the 6-cores have over a 4-core when using software mode only, and the difference between MRQ settings in software mode MPE.
PS. You may be testing this with AME instead of Direct Export. AME is seriously handicapped in CS6 and that may be the cause of your strange results. How does it look with your same test when you use Direct Export, because exporting a 1 minute clip without effects taking 1 minute contrasts seriously with my export of 24 seconds for a clip 2.5 times longer and filled with effects. You know that both Bill and I have rather tuned systems and when our software MPE exports take this long (436 - 870 seconds) there is something going on here I would like to know more about.
Jim Simon wrote:
You may be testing this with AME instead of Direct Export.
I was, as I think that is how most people will export, and how I always export. (I like to see the video.)
I'll give it another go Direct.
What does this mean ? That we should use Direct export for simple transcoding jobs for better performance ? Start up the premiere instead of the easier AME ?
I think the answer could be YES, especially if scaling is also involved. Because as far as I know there is no CUDA in AME so scaling is slower and probably lower quality even with MRQ on.
And what if I just want to do some simple transcodes with no scaling, no frame rate conversions, no nothing ? Would I get a faster result with Direct export ? But for multiple small clips the operation would be cumbersome. And what about the Maximum bit depth and Maximum render quality ? Do those make any difference on transcoding jobs ?
as far as I know there is no CUDA in AME
That depends on how you get there. If you open only AME and import a clip for conversion, there is no hardware acceleration. However, if you send a sequence over from PP, AME will inherit the setting from PP. So if CUDA is on in PP, AME will use it for export.
North America
Europe, Middle East and Africa
Asia Pacific