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 >
I know this is an old thread and for v17 but...

I just built a new PC using v18.

CPU: I9-9900KF
RAM: 32GB DDR4 2666
GPU: 5GB Nvidia P2200
Storage: Samsung M.2 SSD (plus SATA drives on NAS).

On initial rendering over the last few days, it seems that:


  1. Not a huge difference between rendering files on system SSD and NAS SATA drives.

  2. When rendering, CPU maxes out at around 50% consistently.

  3. When rendering, GPU maxes out at around 40% consistently.



Now if I can get maximum utilisation for faster rendering that would be great, but I'm not sure where the bottleneck is. It seems PD18 cannot fully utilise my CPU/GPU.

This machine is AWESOME because I no longer require any proxy files when editing 4K. It is a dream!

But still leaves me pondering if PD18 limits the ability of my PC to produce faster. Thoughts?
The pros for using ffmpeg:


  • Versatility in proxy file creation - see the man page, documentation, etc. for options available

  • Does not require use of PD (multiple benefits)

  • Can be done on another machine or multiple machines if with multiple files

  • Provides completion status updates (and you can pipe standard output/error to a log file in linux easily, etc.)

  • Is available on other platforms (i.e. Linux)

  • Free/open source

  • Easy to install/use once you know the options



The cons for using this workflow:


  • Having to create proxy files (though is negated by the need to create shadow files)

  • Having to rename files, enter/exit PD, etc.



I don't have benchmark statistics on whether ffmpeg is faster. This is by no means the only workflow or necessarily the best. I just find it's much easier and doesn't tie up the CPU on my main computer.

It's a good point that you make about one's use case. I edit live performances (mainly music performances) so don't have the same requirements of others. But with ffmpeg you can recode to 1080p, lower frame rate, etc. so still have a high quality proxy file that is faster to load than an original 4K file.
Always use proxy files when editing - do not use shadow files created by PowerDirector (PD). My current workflow involves using ffmpeg. Here's a cheat sheet to get you going.

1. Download ffmpeg (open source) from www.ffmpeg.org .

2. Use ffmpeg on the command line to create a proxy file (in the \bin folder), for example if ffmpeg is downloaded to the root of your C drive:


  • Start > Run > cmd (for access to the command line)

  • cd \my_videos (where your videos are located)

  • "c:\ffmpeg-4.1.4-win64-static\bin\ffmpeg.exe" -i sample_video.MP4 -vf scale="640:360" -preset ultrafast proxy-sample_video.MP4 (folder may be different depending on the version)


This will create a proxy file at 640x360 resolution. I find it's roughly 2% of the file size of 4K files. You can read up on ffmpeg on other settings, but this tells ffmpeg the input file ( -i ), the resolution ( -vf scale="" ), the quality ( -preset ) and the output file name. Note: If this resolution is too low just change the settings while keeping the horizontal/vertical ratios the same.

3. Run PD and edit your video using the proxy file (in this case proxy-sample_video.MP4).

4. Save your project and exit.

5. Rename proxy-sample_video.MP4 (e.g., to real_proxy-sample_video.MP4).

6. Rename sample_video.MP4 to proxy-sample_video.MP4 (PD will not know the difference).

7. Open your PD project and render your edited video.

8. When exiting, you can rename the files back to the original names (e.g., sample_video.MP4/proxy-sample_video.MP4).


From my experience this is far better than using PD shadow files and you don't get the same type of lag. I regularly edit four 40-50GB 4K video files in the Multicam Designer, sometimes files are over 90GB each. I can't use PD without creating proxy files.
Quote

To my knowledge, no Cyberlink editing product supports either. I would guess that 4:2:2 is more likely to become available in a future version, maybe even as soon as PD18/CD8, but that's pure speculation on my part.

Thanks for the response.
I shoot with GH5/GH5S with latest firmware versions.

Does ColorDirector support color grading/editing 10-bit color (4:2:2)?

Does it support the ALL-I codec (4K/4:2:2/50fps/400Mbps)?

I have an important shoot and would love to film and edit at a higher quality, but not looking to switch from PowerDirector/ColorDirector.

I'll happily edit with proxy files in PowerDirector (using ffmpeg to generate them) but would like to color correct/render using higher end codecs.

Thanks!
Hi,

I shoot with GH5S and GH5 - does PowerDirector or ColorDirector support 10-bit color (4:2:2)? Does it support the ALL-I for imported files?


I use PowerDirector 15 and it seems not at all.

I would like to shoot with the ALL-I codec (4K/4:2:2/50fps/400Mbps) but need to be able to color grade. I always edit with proxies using ffmpeg but would like to import highest possible quality and render at highest possible quality.

Thanks!
I had this happen to me about a week ago. Lost literally a day of work as you cannot save during an edit. First time it has happened to me but it was not pleasant.

Any suggestions on workflow so that you can mitigate this risk?
Go to:   
Powered by JForum 2.1.8 © JForum Team