If they could reach x264 superfast quality or very close with the speed they got (or at least realtime 60fps speeds) that would be amazing, doubt we will see that though. I do think encoding gameplay with Quicksync etc is quite common with streamers, now are they the vast majority, most likely not, but it's far from small.īarely anyone uses x264, even though they should in my opinion if they Can, even x264 superfast beats them by a lot. In complex scenes they will perform even worse than x264, so this is more of a "fair chance", but honestly i just took this video randomly.Ī good video would be some game recording in with lots of tiny details (maybe grass and stuff like that), and why Gameplay, cause it's crystal clear giving away the quality loss much easier. This is not meant to be a very thorough, it's meant to give you an idea of it.Įverything currently compared to x264 is basically losing big time.įrom the Quicksync it seems that after Ivy Bridge there has been little if any quality improvements, probably there has been speed improvements.Īs for NVENC, no clue there, i would assume the same, just speed. It's not Animé, but it's a fairly simple video to encode, yes. I haven't searched yet, but there must be a freely available high quality source one can use for testing without infringing any copyrights.
#QUICKSYNC H.264 SOFTWARE#
Personally, I would like to know the difference in quality and speed between the generations of Quick Sync compared to software x/h264 encoding. Like I said, I'd be willing to help out with real testing to find good results. I'm willing to participate in testing if it's actually going to be realistic.Īt my disposal, I have an Intel HD 4600 GPU as well as a computer with an Ivy Bridge Extreme 4930K OC'd to 4.6GHz, liquid cooled.Īlso, I have and EVGA GTX 780 ti Classified for GPU testing.
#QUICKSYNC H.264 MOVIE#
Try transcoding at least an hour of a high action movie at different bitrates/rate factors, using different hardware then compare results. When doing comparisons, WATCH the entire clip at it's rated framerate.
#QUICKSYNC H.264 1080P#
1080p 10bit is also do-able, but I wouldn't have more than one stream going.I hate to burst anyones bubble, but, the source file isn't really something that should be used for testing for many reasons:Ģ - it's anime - the vast majority of people using Quick Sync aren't going to encode Anime or game playģ - really, are you going to rate or grade an entire paper/test based on a few frames which takes mere moments to complete?
#QUICKSYNC H.264 720P#
Once it catches up, it drops to idle, then reads some more, thus resulting in the fluctuations.I would say 720p 10bit should be fine for the QNAP 451+ with the HW transcoding on. I believe this is due to the system catching up and buffering ahead.
Sometimes the CPU usage is consistent, but after a while the CPU values fluctuate wildly. Here are my results (keep an eye on the pink and green lines).ġ080p 10bit = No HW 100% -> With HW 70-50% (When stable 20% with regular spikes of 50%)ħ20p 10bit = 80% -> With HW 70-50% (When stable 20% with regular spikes of 50%). Enabling the hardware transcoding on Emby (no Intel quick-sync option unfortunatly) only helped with various functions of the transcoding (maybe audio or some other aspect).
While the QNAP 451+ can't do hardware 10bit, it is capable of doing it in software, which is why it maxes out the processors. Okay, I went ahead and got a 1 month subscription for hardware transcoding, just to see how it works out.
0 kommentar(er)
