Video choppy durring panning scenes

ondoho · 01-18-2022, 12:14 AM

Quote:

Originally Posted by Red Squirrel

For example if I pause one of the videos and then drag the window at the same speed it was panning, I don't get the choppiness. so I don't think it's necessarily a display issue but a playback issue specifically.

I thought we established that already.

Quote:

Originally Posted by Red Squirrel

Not sure if playing video touches different parts of the hardware than just GUI actions

It does, or at least it's better if it does.

See my previous post, still a few questions to answer.

Red Squirrel · 01-23-2022, 11:53 AM

Not sure which questions I missed, but yeah it does it with almost all videos. The 60fps video downloaded and played directly seemed to be better but in Youtube it was bad. But the other ones are bad regardless of being played in a player locally or not. Tried some movies too.

Not really sure if it's codec specific, not sure what codecs every video is running or how to check, there are tons of different ones.

Would getting a higher refresh rate display fix this issue? At this point I am willing to splurge on a higher end display if that's what it takes. Any videos that have panning or someone free handing a camera make me sick and I can't even watch them.

I guess worse case scenario I can just watch videos on my windows machine but I want to not need to rely on windows for anything.

enigma9o7 · 01-23-2022, 06:47 PM

There are browser extensions for firefox like h264ify that will tell youtube that you don't support VP8/VP9 so they feed you h264, which likely your video card can decode in hardware.

SW64 · 01-24-2022, 11:21 AM

Quote:

Originally Posted by enigma9o7

There are browser extensions for firefox like h264ify that will tell youtube that you don't support VP8/VP9 so they feed you h264, which likely your video card can decode in hardware.

Well, that gave me a new direction to look into. I was wondering if the graphic card can support VP9 hardware decoding. I looked up the graphic card's specs (https://www.amd.com/en/products/graphics/radeon-rx-580) and I find these 'Unified Video Decoder (UVD)' and 'Video Code Engine (VCE)' in the Features sections. Wikipedia had their information in details.

https://en.wikipedia.org/wiki/Video_Coding_Engine#GPUs

https://en.wikipedia.org/wiki/Unifie...Format_support

In the first link, I see that the RX 580 graphic card is 'UVD 6.3' and in the second link, that fall in the 'No' section of the 'VP9 decoding' column for the 'UVD 6.3' row. Red Squirrel, if you convert a "problematic" VP9-encoded video to H.264 codec or playback only H.264 videos on youtube in your browser as suggested by enigma9o7, does your problem go away? If yes, then we may have found the source of the problem.

Code:

ffmpeg -i video.webm -c:v libx264 -pix_fmt yuv420p video.mp4

ondoho · 01-24-2022, 11:48 PM

Adding to what SW64 said:

Quote:

Originally Posted by Red Squirrel

Not really sure if it's codec specific, not sure what codecs every video is running or how to check, there are tons of different ones.

At this point in the thread, this answer really isn't sufficient.
Depending on your media player, it's just a hotkey ('i' on mpv) or somewhere in the menu (sth like "media info" I guess).
Otherwise you could use something like mediainfo, a separate gui/tui app.

And yes, I think we should concentrate on local files & media players for now. The browser brings in additional complexity that should be solved later, IMO.

SW64 · 01-25-2022, 11:26 AM

I didn't know about the 'i' shortcut in mpv. It shows all the information I wish to see during a playback. Thanks for that useful tip!

Red Squirrel · 01-25-2022, 08:07 PM

Ok so I tried to convert a video with that ffmpeg command but it did not seem to change anything, the resulting file still has lot of jitter. Playing it with VLC.

I also tried that h264ify extension and it does not seem to do a difference. Is there a way to confirm it's active/working when watching a video? But yeah for now maybe need to focus on getting media player playback to look normal first.

I'm not too familiar with any Linux games, any games I could also test with? This would rule out the issue being specific to playback.

ondoho · 01-29-2022, 04:07 AM

I feel like you're not hearing me. ¯\_(ツ)_/¯

Red Squirrel · 01-29-2022, 04:27 PM

I'm not sure what part I might be missing. Anything else to try that I missed? Reread through the thread but think I tried everything suggested so far.

SW64 · 01-29-2022, 06:18 PM

glxgears is the closet thing you can to testing the 3D side of the graphic cards without installing a game. Can you try #30 and post the result?

Quote:

Reread through the thread but think I tried everything suggested so far.

It's what ondoho was trying to tell you, we can't do much until we hear of the results of your efforts. In details, lol. You're our proxy eyes to your problem.

Red Squirrel · 01-31-2022, 05:47 PM

Ok just tried it, I didn't realize I had to stop it manually, so here is the output for as long as it ran for:

Code:

glxgears
Running synchronized to the vertical refresh.  The framerate should be
approximately the same as the monitor refresh rate.
323 frames in 5.0 seconds = 64.442 FPS
300 frames in 5.0 seconds = 59.996 FPS
300 frames in 5.0 seconds = 59.996 FPS
300 frames in 5.0 seconds = 59.995 FPS
300 frames in 5.0 seconds = 59.997 FPS
300 frames in 5.0 seconds = 59.996 FPS
300 frames in 5.0 seconds = 59.996 FPS
300 frames in 5.0 seconds = 59.996 FPS
300 frames in 5.0 seconds = 59.996 FPS
300 frames in 5.0 seconds = 59.997 FPS
300 frames in 5.0 seconds = 59.995 FPS
300 frames in 5.0 seconds = 59.997 FPS
300 frames in 5.0 seconds = 59.995 FPS
300 frames in 5.0 seconds = 59.998 FPS
300 frames in 5.0 seconds = 59.995 FPS
300 frames in 5.0 seconds = 59.997 FPS
300 frames in 5.0 seconds = 59.996 FPS
300 frames in 5.0 seconds = 59.996 FPS
300 frames in 5.0 seconds = 59.997 FPS
300 frames in 5.0 seconds = 59.996 FPS
300 frames in 5.0 seconds = 59.995 FPS
300 frames in 5.0 seconds = 59.997 FPS
300 frames in 5.0 seconds = 59.997 FPS
300 frames in 5.0 seconds = 59.995 FPS
300 frames in 5.0 seconds = 59.997 FPS
300 frames in 5.0 seconds = 59.994 FPS
300 frames in 5.0 seconds = 59.996 FPS
300 frames in 5.0 seconds = 59.997 FPS
300 frames in 5.0 seconds = 59.997 FPS
300 frames in 5.0 seconds = 59.995 FPS
300 frames in 5.0 seconds = 59.996 FPS
300 frames in 5.0 seconds = 59.996 FPS
300 frames in 5.0 seconds = 59.998 FPS
300 frames in 5.0 seconds = 59.997 FPS
300 frames in 5.0 seconds = 59.995 FPS
300 frames in 5.0 seconds = 59.997 FPS
300 frames in 5.0 seconds = 59.989 FPS
301 frames in 5.0 seconds = 60.001 FPS
300 frames in 5.0 seconds = 59.999 FPS
300 frames in 5.0 seconds = 59.993 FPS
301 frames in 5.0 seconds = 60.000 FPS
300 frames in 5.0 seconds = 59.995 FPS
300 frames in 5.0 seconds = 59.994 FPS
300 frames in 5.0 seconds = 59.998 FPS
300 frames in 5.0 seconds = 59.997 FPS
300 frames in 5.0 seconds = 59.995 FPS
300 frames in 5.0 seconds = 59.996 FPS
300 frames in 5.0 seconds = 59.992 FPS
301 frames in 5.0 seconds = 59.999 FPS
300 frames in 5.0 seconds = 59.995 FPS
300 frames in 5.0 seconds = 59.999 FPS
300 frames in 5.0 seconds = 59.995 FPS
300 frames in 5.0 seconds = 59.997 FPS
300 frames in 5.0 seconds = 59.998 FPS
300 frames in 5.0 seconds = 59.994 FPS
300 frames in 5.0 seconds = 59.997 FPS
300 frames in 5.0 seconds = 59.997 FPS
300 frames in 5.0 seconds = 59.996 FPS
300 frames in 5.0 seconds = 59.995 FPS
300 frames in 5.0 seconds = 59.998 FPS
300 frames in 5.0 seconds = 59.997 FPS
300 frames in 5.0 seconds = 59.996 FPS
300 frames in 5.0 seconds = 59.996 FPS
300 frames in 5.0 seconds = 59.996 FPS
300 frames in 5.0 seconds = 59.996 FPS
300 frames in 5.0 seconds = 59.994 FPS
300 frames in 5.0 seconds = 59.999 FPS
300 frames in 5.0 seconds = 59.994 FPS
300 frames in 5.0 seconds = 59.997 FPS
300 frames in 5.0 seconds = 59.997 FPS
300 frames in 5.0 seconds = 59.992 FPS
300 frames in 5.0 seconds = 59.999 FPS
300 frames in 5.0 seconds = 59.998 FPS
300 frames in 5.0 seconds = 59.994 FPS
300 frames in 5.0 seconds = 59.998 FPS
300 frames in 5.0 seconds = 59.995 FPS
300 frames in 5.0 seconds = 59.998 FPS
300 frames in 5.0 seconds = 59.995 FPS
300 frames in 5.0 seconds = 59.997 FPS
300 frames in 5.0 seconds = 59.991 FPS
300 frames in 5.0 seconds = 59.999 FPS
300 frames in 5.0 seconds = 59.998 FPS
300 frames in 5.0 seconds = 59.997 FPS
300 frames in 5.0 seconds = 59.994 FPS
300 frames in 5.0 seconds = 59.997 FPS
300 frames in 5.0 seconds = 59.996 FPS
300 frames in 5.0 seconds = 59.996 FPS
300 frames in 5.0 seconds = 59.997 FPS
300 frames in 5.0 seconds = 59.994 FPS
300 frames in 5.0 seconds = 59.997 FPS
300 frames in 5.0 seconds = 59.997 FPS
300 frames in 5.0 seconds = 59.997 FPS
300 frames in 5.0 seconds = 59.996 FPS
300 frames in 5.0 seconds = 59.997 FPS
300 frames in 5.0 seconds = 59.996 FPS
300 frames in 5.0 seconds = 59.996 FPS
300 frames in 5.0 seconds = 59.993 FPS
300 frames in 5.0 seconds = 59.999 FPS
300 frames in 5.0 seconds = 59.997 FPS
300 frames in 5.0 seconds = 59.996 FPS
300 frames in 5.0 seconds = 59.995 FPS
300 frames in 5.0 seconds = 59.998 FPS
300 frames in 5.0 seconds = 59.996 FPS
300 frames in 5.0 seconds = 59.996 FPS
300 frames in 5.0 seconds = 59.996 FPS
300 frames in 5.0 seconds = 59.996 FPS
300 frames in 5.0 seconds = 59.995 FPS
300 frames in 5.0 seconds = 59.997 FPS
300 frames in 5.0 seconds = 59.993 FPS
300 frames in 5.0 seconds = 60.000 FPS

And here's codec info for one sample video:

https://i.imgur.com/FRl3JNr.png

SW64 · 02-01-2022, 08:34 PM

Thanks for posting the results. Yeah, your glxgears is working just fine. Your screenshot is very close to mine (see attached). I pulled down the 137 and 140 in the https://www.youtube.com/watch?v=fVrcbBz0d4Q youtube link to get the h264/mp4 format. Not sure why yours is saying 'vaapi', though.

I googled vaapi. Is your monitor plugged into your motherboard's display port? It should be plugged into your graphic card's port instead.

Red Squirrel · 02-05-2022, 07:20 PM

Yeah the monitors are plugged into the GPU's HDMI ports. I purposely chose that GPU as I wanted to avoid using DP to HDMI adapters if I can just to rule out any issues. That and DP cables don't seem to come in 25 foot lenghts and HDMI do. The PCs are in the basement and the cables come to the KVM. Makes everything cleaner at my desk as I don't have room for two machines. The KVM supports 4k at 60hz.

SW64 · 02-06-2022, 06:18 PM

Wait, you have a KVM in between the monitor and the computer? That will have an effect, too. Try without KVM just to verify if this KVM is potentially the cause of your problem.

Red Squirrel · 02-09-2022, 11:15 PM

Hmm that's worth a shot at least to rule it out. Hopefully it's not that.

I have too much stuff going on right now to mess with that (unplugging monitors tends to cause lot of chaos with the window positions) but when I get a chance I will try it. I can use a HDMI coupler to just connect the two cables together to bypass it.

The KVM I have is one of few that is actually very good as it emulates everything, so there's no delay between switching (at least for keyboard/mouse) and no monitor disconnect/reconnect either, so it means the OS does not go haywire. I actually bought one for my 2nd monitor too just so that when I turn it off the OS does not know. When I work from home I also switch it to my work machine so I have the 2 4ks for work.

But yeah I will try to bypass it just to see if it makes a difference and report back.

If it does turn out to be that, I suppose I can just setup one of my separate monitors with a RPI or something and dedicate that towards watching videos.