Announcement: Our new CyberLink Feedback Forum has arrived! Please transfer to our new forum to provide your feedback or to start a new discussion. The content on this CyberLink Community forum is now read only, but will continue to be available as a user resource. Thanks!
CyberLink Community Forum
where the experts meet
| Advanced Search >
Thanks all - this has been extremely helpful. I'll try my video processing to my son's much more current and powerful gaming box and see how things go. DxDiag is attached

But I still don't understand why any encoding is even necessary when processing a bare H.265 to produce a H.265 with same specs that the profile analyzer finds. And why can my GoPro pull off H.265 compression in real time while my Desktop PC is brought to it's knees?

Is HVEC even worth it in this age of cheap drive space? I can set the GoPro to record in H.264 - will that result in longer battery life and lower energy cost to process?

There is much I do not know.....

Quote

Not really crazy, H.265 is a much more compression efficient algorithm vs H.264, hence to achieve that higher level of video compression, it takes substantially more compute cycles. That's why my reference prior that one needs a good GPU with hardware encoding. Or, a very high end CPU to achieve anything similar time wise results with CPU encoding. Your GTX 760 does not support H.265 hardware encoding and i5-7400 is very weak for CPU based encoding. That's why you see your 4+ hrs for 15 minutes.

As for the reason SVRT is not functional, probably my item 2 referenced prior. This is a very small duration clip but the video only GOP based bitrate ranges from 10-55Mbps with a overall average of 38Mbps. That variable bitrate is high for functional SVRT based on my experience.

Jeff
Update - I see a similar discrepancy in PD 15. Take a bare H265 file and it takes almost 10 times longer to produce in its native format as it does to produce it as an H264 - that seems crazy. I'm looking at 4+ hours for a 15 minutes video.

Quote Fascinating -

I used your settings (I get the same with the profile analyzer) to make an H265 clip and it took 76 seconds to produce (4.3 second clip). Using the H264 setting from my post below, it took 7 seconds - one tenth the time. I did this three times just to verify. I can't really tell the difference between in the H264 and the H265. What the hell?!?

I have OpenCL turned on and hardware acceleration off. Turned on hardware acceleration and it took 75 seconds.

I have a bit of a minimalist system:

Windows 10, 64-bit
i5-7400, 16G RAM, 2Tb HD
GeForce GTX 760

p.s., I've only downloaded the Essentials content pack - that should not make a difference, right?





Fascinating -

I used your settings (I get the same with the profile analyzer) to make an H265 clip and it took 76 seconds to produce (4.3 second clip). Using the H264 setting from my post below, it took 7 seconds - one tenth the time. I did this three times just to verify. I can't really tell the difference between in the H264 and the H265. What the hell?!?

I have OpenCL turned on and hardware acceleration off. Turned on hardware acceleration and it took 75 seconds.

I have a bit of a minimalist system:

Windows 10, 64-bit
i5-7400, 16G RAM, 2Tb HD
GeForce GTX 760

p.s., I've only downloaded the Essentials content pack - that should not make a difference, right?




Quote Thanks for sharing your clip. I also don't have SVRT available, but by using PD's built-in Profile Analyzer I was able to have PD create a custom profile and then produce the clip in <1 second. I can't tell the two clips apart when playing them one after another in VLC. Here are the settings I used:"""""" I've attached the produced clip along with the custom profile (in case you have trouble getting the Profile Analyzer to work for any reason)
Thanks for the reply - all my testing has been done using only a short clip - to no avail. On reflection, I think variable bitrate is more likely to be the culprit than variable framerate, but that is a decidedly uninformed opinion.

I've done some more testing and found that if I render in to h.264 instead of HVEC using the default 1080/60p 40Mbps setting, it renders quite quickly, but there is still some quality loss. The attached clip is a good example - snow scenes in flat light look really good in the native gopro file, but get blocky if any kind of rendering is done in PD, even with the highest quality settings in the profile.

I don't know why this would be a problem for PD - I'm down to just asking PD to string a bunch of HVEC video together taken at identical settings with no effects or transitions and I can't do it.
I recently started shooting with a Hero 10 that uses HVEC compression. My prior Hero7 used h.264. I always shoot 1080 video at 60 fps. The bit rate for the h.264 files is 40000 and the bitrates for the HVEC files are nominally 44000, but they vary slightly. SVRT analysis of my h.264 video give me a usable profile that requires no rendering and results in super fast renders and no loss in quality. But the SVRT for the HVEC video is not usable and the renders produced by the profile analyer (custom profile of 1920x1080, 59.94 fps and a 44400 rate are excuciatingly slow and are lower quality than the original. I upgraded from PD15 to PD21 thinking that would fix the problem, but it did not. In fact, PD15 renders my HVEC files faster, although no better.

I'm no expert in video compression and I'm beginning to suspect that his might have something to do with variable frame rates. If so, is it possible to get lossless renders with HVEC in PowerDirector? Any thoughts on what, if anything, I might be doing wrong?
Thanks to both of you for this -I was having the same issue - did not realize that "apply to all" only applied to the active timeline. I've never used multiple video timelines - I guess I now know what they are useful for.

I'd like to follow up with a workflow question - I find that applying a color preset really bogs down the editing process - as a newbie at this, I was wondering what you recommend for a workflow. Right now, I import all my desired clips (usually around 50), trim them to size, add transitions, add music, adjust clips to match music transitions, and then color correct. I do all of this in PowerDirector and never really invoke ColorDirector except to create a preset. But by the time I'm done, I can't really see a smooth preview of a 6 minute video. Is this a reasonable workflow? And is it true that color correction/effects is really hard on the processor?

Roger
Following up after the purchase - everything is pretty awesome so far. But I have to say, I would not have purchased the software based on the demo - Cyberlink has this community to thank for that. They should just put a big watermark on the demo rather than crippling critical features IMHO.

I'm sure to be posting more noob questions...
Thanks for the feedback - I'll give it a whirl.
Thanks for the info on the 30 day return - just hoping to save myself some hassles.

I've attached the DxDiag - thanks for that - nice summary of things I'm always wondering about or forgetting.

Re: the WMV, I'm using WMM to encode the same clips and it's smooth and crisp compared to the wmvs I'm encoding on PD. Both are highly compressed , but I guess that's to be expected. I'm only trying out wmv in PD because I can't do mp4 or mts in the trial version. I thought you could encode wmv files at more than 10 Mbs, but if not, my mistake - I'm trying to get up to speed here.

I'm trying to make high quality videos of surfing and skiing with the family for burning to blu-ray, or playing on a big screen or projector off USB. The quality is of paramount importance to me and it's disappointing that I can't assess that in the trial version. I guess I'll be buying it and reporting back.
I'm am having trouble producing quality HD video in the trial version of PowerDirector. My hardware is as follows:

AMD Radeon HD 6700 Series
Intel i5 quad core
16 G RAM

My media is GoPro mp4 files and mts files from my Canon HF200 (which AVCHD 60i). The Canon is older, so the default setting was interlaced - I have now learned its better to set it to progressive and I have done so, but I have tons of video in 60i.

My test video is 4 alternating clips from the Canon and the GoPro.

In the trial, mp4 encoding is not allowed at all, so I can't test that. The AVC look bad, but I can only test it with hardware acceleration and I understand that can cause issues with AMD. The output is choppy as is the display in the editor - compared to the originals, it looks horrible. I tried the mpg, which is supposedly fully supported in the trial, and it too is choppy. The only thing producing smooth results is the wmv format with hardware acceleration off, but it has too much compression. I find it really irritating that I can't up the quality of the HD WMV since it should be native (maybe I'm missing something).

From reading the forums, I understand that the trial version may not support decoding and encoding of my formats. If so, how will I know if it works for my system and media? Right now, Windows Movie Maker is doing a much better job, but I'd really like to take my video editing to the next level and I have no confidence at all based on the trial. Is anyone else having issues (or not) with the full version?
Go to:   
Powered by JForum 2.1.8 © JForum Team