real time vs recorded time using amarec

2014-04-16 10:16:12 pm

DO THE MEATEO

Hi, I recorded a run using Amarec 2.10b that was timed (realtime) at 2:04:24 using wsplit but playing back the recording is 6 seconds slower, timed at 2:04:30. The Amarec capture settings were at *w= 640, h= 480, fps=29.97, fcc=YUY2, bit=16 (i.e. 29.97 fps) while the recording settings were at 30 fps. Just a couple of questions:

1 - I'd like to make sure that recordings using Amarec are the same time as my actual runs in the future. How do I do this - can I just change the recording setting to 29.97 fps? I looked at the 'Mysterious gained seconds' sticky and am having trouble understanding it. Is VirtualDub more reliable?

2 - I recorded this offline so the only recording I have is that of the 6 second slower version via Amarec - I have no video evidence linking the wsplit time to the time of the actual run (i.e. no Twitch or OBS recording). Can I legitimately state that this run is 2:04:24 or just bite the bullet and take the slower 2:04:30? This run happens to be a current WR so I'd like to be as honest as possible here :/

Thanks

Thread title:

Hi, I recorded a run using Amarec 2.10b that was timed (realtime) at 2:04:24 using wsplit but playing back the recording is 6 seconds slower, timed at 2:04:30. The Amarec capture settings were at *w= 640, h= 480, fps=29.97,  fcc=YUY2, bit=16 (i.e. 29.97 fps) while the recording settings were at 30 fps.  Just a couple of questions:

1 - I'd like to make sure that recordings using Amarec are the same time as my actual runs in the future. How do I do this - can I just change the recording setting to 29.97 fps? I looked at the 'Mysterious gained seconds' sticky and am having trouble understanding it. Is VirtualDub more reliable?

2 - I recorded this offline so the only recording I have is that of the 6 second slower version via Amarec - I have no video evidence linking the wsplit time to the time of the actual run (i.e. no Twitch or OBS recording). Can I legitimately state that this run is 2:04:24 or just bite the bullet and take the slower 2:04:30? This run happens to be a current WR so I'd like to be as honest as possible here :/

Thanks

UraniumAnchor

2014-04-17 02:32:16 am

Not a walrus

To answer your first question VirtualDub has some settings you can use to make sure you get every frame at the cost of temporarily losing audio sync. (I believe "Do not resync between audio and video streams" and "Correct video timing for fewer frame drops/inserts" under "Capture -> Timing" is what you want) Once you've recorded it, there's an option that lets you adjust the video speed to match the length of the audio track, which should give you the correct result (Video -> Framerate -> "Change so video and audio durations match"). I haven't used Amarec to any significant degree so I can't answer you there, unfortunately.

In a run that long I doubt anybody's going to quibble that much over 6 seconds, so just stick a note somewhere if you do plan on uploading it to YouTube or whatever.

That said, SDA would probably keep the 2:04:30 if that's what the video looked like, but I don't know if you were planning on submitting whatever you recorded or not.

the_roth

2014-04-17 06:39:10 am

DO THE MEATEO

Thanks for your help, will look at vdub over the weekend. I restreamed the run on twitch (with wsplit which shows the time desyncing) so at least its out there, wasn't planning on submitting as there is some sloppy execution which doesn't look too good... Hopefully the next improvement by someone beats both times!

honorableJay

2014-04-17 04:28:03 pm

The Dork Knight himself.

Try this for Amarec: set your recording framerate to 31 and see if the video playback speed matches the real-time. It sounds like your console is outputting a framerate that is higher than the standard of 29.97 and for some users setting the recording framerate in Amarec to 1 frame higher than normal can fix the issue. Just be wary that you might run into an audio desync (just like using vDub).

the_roth

2014-04-17 09:43:38 pm

DO THE MEATEO

No worries will try that, thanks. Will let you know if it works!