hi,
I bought used graphics card "MSI GTX 580 Twin Frozr III PE OC 1536MB", to speed preview and render in sony Vegas.
After first tests the results was not satisfacted.
I've decided to do some rendering tests with 3 different version of Sony Vegas, with the same entry file, the same project and render settings and the same result file.
Results below:
Entry file: *.mts, 1080-50i, AVCHD, 60second (without FX) (Sony AX2000)
Project settings:
HD 1080-50i (1920x1080; 25,000 fps),
Field order: Upper field first,
pixel format 8bit,
Full-resolution rendering quality: Best
Render settings:
Sony AVC/MVC (*.mp4; *.m2ts; *.avc)
Template: AVCHD 1920x1080-50i
Bitrate:16Mbps
Hardware:
Processor: i7-4790
mainboard: ASROCK B85M Pro3
RAM: 12Gb DDR3 (2x4Gb 1600 CL9, 2x2Gb 1600 CL7)
HDD: SSD 250GB (only system and Sony Vegas), 2x750GB 7200obr. (dane)
Graphics card 1: MSI GTX 580 Twin Frozr III PE OC 1536MB (driver: 296.10)
Graphics card 1: HD4600 (intergated with processor)
Power supple: Chieftec (CFT-600-14CS) 600W
System: Windows 7 Home Premium 64bit, SP1
Sony Vegas Pro 10 (oryginal)
i7-4790: 163 sec.
i7-4790+GTX580: 131 sec.
i7-4790+Intel Quick Sync Video, HD4600: no support
Sony Vegas Pro 12 (trial)
i7-4790: 82 sec.
i7-4790+GTX580: 68 sec.
i7-4790+Intel Quick Sync Video, HD4600: 33 sec.
Sony Vegas Pro 13 (trial)
i7-4790: 92 sec.
i7-4790+GTX580: 74 sec.
i7-4790+Intel Quick Sync Video, HD4600: 46 sec.
In render GTX580 was used about 20-30%
Conclusion is one.
The fastest rendering was in Sony Vegas Pro 12 with card graphic integrated with processor with IntelQuick Sync (33 seconds).
The questions:
1) Anybody have the same configuration and the results are better then my results (with GTX)?
2) Is my card efficient in 100%? I try changes drivers (on CD with card; latest from nvidia site; ), and the results are worst. With driver version 359.06 the results is difference about 2-3 seconds
3) In this case, processor is strong enough, that any graphics card 5xx series can't match them?
After tests I installed Cuda Toolkit 7.5 too. Improvement is only in preview (Best-Full - fluently preview 25fps). Without Cuda Toolkit the preview was lag in transitions (even with gpu transition - 10-15 fps)
Maybe other observations?
Thanks for answer
I bought used graphics card "MSI GTX 580 Twin Frozr III PE OC 1536MB", to speed preview and render in sony Vegas.
After first tests the results was not satisfacted.
I've decided to do some rendering tests with 3 different version of Sony Vegas, with the same entry file, the same project and render settings and the same result file.
Results below:
Entry file: *.mts, 1080-50i, AVCHD, 60second (without FX) (Sony AX2000)
Project settings:
HD 1080-50i (1920x1080; 25,000 fps),
Field order: Upper field first,
pixel format 8bit,
Full-resolution rendering quality: Best
Render settings:
Sony AVC/MVC (*.mp4; *.m2ts; *.avc)
Template: AVCHD 1920x1080-50i
Bitrate:16Mbps
Hardware:
Processor: i7-4790
mainboard: ASROCK B85M Pro3
RAM: 12Gb DDR3 (2x4Gb 1600 CL9, 2x2Gb 1600 CL7)
HDD: SSD 250GB (only system and Sony Vegas), 2x750GB 7200obr. (dane)
Graphics card 1: MSI GTX 580 Twin Frozr III PE OC 1536MB (driver: 296.10)
Graphics card 1: HD4600 (intergated with processor)
Power supple: Chieftec (CFT-600-14CS) 600W
System: Windows 7 Home Premium 64bit, SP1
Sony Vegas Pro 10 (oryginal)
i7-4790: 163 sec.
i7-4790+GTX580: 131 sec.
i7-4790+Intel Quick Sync Video, HD4600: no support
Sony Vegas Pro 12 (trial)
i7-4790: 82 sec.
i7-4790+GTX580: 68 sec.
i7-4790+Intel Quick Sync Video, HD4600: 33 sec.
Sony Vegas Pro 13 (trial)
i7-4790: 92 sec.
i7-4790+GTX580: 74 sec.
i7-4790+Intel Quick Sync Video, HD4600: 46 sec.
In render GTX580 was used about 20-30%
Conclusion is one.
The fastest rendering was in Sony Vegas Pro 12 with card graphic integrated with processor with IntelQuick Sync (33 seconds).
The questions:
1) Anybody have the same configuration and the results are better then my results (with GTX)?
2) Is my card efficient in 100%? I try changes drivers (on CD with card; latest from nvidia site; ), and the results are worst. With driver version 359.06 the results is difference about 2-3 seconds
3) In this case, processor is strong enough, that any graphics card 5xx series can't match them?
After tests I installed Cuda Toolkit 7.5 too. Improvement is only in preview (Best-Full - fluently preview 25fps). Without Cuda Toolkit the preview was lag in transitions (even with gpu transition - 10-15 fps)
Maybe other observations?
Thanks for answer