I researched a lot for this but I cannot figure it out.
I am a complete newbie in this area so I need help!
I have a 23.976fps Video and a 29.97fps video with my local language audio track.
I want to extract 29.97fps video's AUDIO track and put it in my 23.976fps video (as this is a superior video quality and audio is only available in 29.97fps file).
What can I do? I researched but most say use eac3to, but they only give instructions for 25->23.976 not 29.97->23.97.
I read on one of the pages that 23.976FPS and 29.97FPS can easily be done (if same length because of some duplicate frame or something, not an expect haha, but my case is that my 29FPS is SMALLER than the 23.976FPS video, so I think I wanna slowdown my 29fps to 23).
23.976FPS Video Length: 01:31:25
29.97FPS Video Length: 01:28:22
Is this even possible? I am a complete newbie so if another thread has answers it'll be awesome!
+ Reply to Thread
Results 1 to 20 of 20
Please study what "telecine" does: It changes the structure of the video without changing the audio or the playtime. There is neither speedup nor slowdown.
The reason why these two videos have a different playtime is possibly that there are different scene cuts...
Sure, but I dont think there are scene cuts. I played both videos side by side and the smaller length (29fps), by 1 minute into the video, it was AHEAD of the 23fps film, no scene was left out.
So what can this be?
I'll look into telecine as well.
Try my clever FFmpeg-GUI (read the HowTo.txt on first use).
Load your 29.97fps video, click on Main, click on Modify Audio Stream, click on Continue, set AAC or AC3, set Bitrate, on Change Audio Lenght input 29.97 (as Source FPS) and 23.976 (as Target FPS), set Lenght & Pitch, click on convert.
Click on Multiplex, set your 23.976 video (as Video Source) and the new created Audio file (as Audio Source) and click on Multiplex. Done.
Check if video and audio are synchronous, e.g. in VLC. If there is an A/V delay, remux the two tracks with this value.
If the difference isn't caused by different cuts (and I doubt he's right about this) all the OP has to do is change the length of the shorter audio to match the length of the longer video: 5302 seconds to 5485 seconds.
Just a hunch, but could one of them be being misinterpreted? E.g. 23.976 vs. 24, or 29.97 vs 30?. That would only account for <1% difference, however. 4% is usually due to 25 vs 29.97 differences. Barring the obvious - scene cut differences.
The difference most likely isn't a cut scene as you think but the duration of scene changes can be so small that you dont see it but a lot of them will add up a time difference.I think,therefore i am a hamster.
It's a combination of both PAL/NTSC differences plus different frame count/scenes. He said, "I played both videos side by side and the smaller length (29fps), by 1 minute into the video, it was AHEAD of the 23fps film, no scene was left out." So, that almost guarantees a PAL/NTSC difference (with the PAL one being field blended). Since one isn't 4.16% slower, that would mean different cuts of the film. This is probably a major undertaking, even if he did understand how to fix it.
So been bit busy with this,
I just took a look, my fault I am clearly blind but yes, indeed the shorter film has some small (nearly un-noticable) scene cuts. So I can see why its less. The scene cut is so fast and well done that it missed my eyes the first fiew times.
So what are the best choices for me to merge audios? Should I manually for each scene cut, add audio from un-cut video to replace the cut points? (so all scene cuts for at musical scenes, so for me, dialogue wont be an issue Is this the simplest solution?
Also since the smaller cut version is 29fps and while uncut is 23, will it make any difference, can I manually fill in the cut points and the 29fps audio can then merge with 23?
Load the file into audacity and choose select all and then effect change speed,in the selection length put in the times you want to adjust and file export as audio of your choice.
Then mux the audio and see if the audio matches close enough.I think,therefore i am a hamster.
Wait so if I just change speed how will that work?
Like I said there are 5 6 very hard to notice scene cuts, so in that 10-15 seconds of where in the shorter video Scene N does not exist, what will extending audio do? Wouldn't it just mess it up? What will play in the 10-15 seconds of uncut scene then?
I was thinking to manually find scene cuts and load audio from uncut video and manually fill gaps. Wouldn't that work?
Just try what i posted,i had the same issue and what did fixed it,might help with your situation.I think,therefore i am a hamster.
Sure worth a try, I'll let you guys know.
Just tried it, did not work. Everything is more out of sync haha.
So what's left? Manually doing it every scene cut by scene cut using something like Audition/Audacity right?
Is it okay to merge 23FPS audio into the missing cuts in the 29FPS english audio? (I know audio has no FPS explicitly but you get the idea).
Audio has no "frames per second", it is considered continuous. So you can try mixing it (although I fear it might have an audible difference in the channel mix).
Haha yes I know it has no frames, but I just wanted to describe it to differentiate the video source.
Ill try manual cutting. Hopefully it works.
Thanks for the help everyone, appreciated.
Its easier to do the conversion with an audio/video container format than with the audio stream by itself, because it loses some info about the framerate, and in any audio manipulating software you will end up having to deal with BIT RATE, BIT RATE MODE, SAMPLING RATE, FRAME RATE, and even if you will be basing your work ins frames or seconds. So, if you don't know what you are doing, you may end up with a source material with the same exact lenght, but which sounds funny because you unintentionally messed up the pitch, not by doing so, but by inducing the audio app to do it as a way to achieve your target.
Please don't confuse people more than what they are already with unbased statements, because instead of helping them, you are creating more confusion. If you aren't 100% sure of what you are saying, is better not to say it.
"Norm conversions" between PAL and NTSC are not pure Telecine (like the 3:2 Pulldown between NTSC-Film and NTSC-TV), but contain either an audio speedup/slowdown (which of course changes the duration) or pretty much a destruction of the video (field operations with or without blending).
Verbosely explained in the guide "Exotic Interlacing" (originally in German by scharfis_brain, translated by StainlessS on doom9). Including AviSynth examples how to visualize the issues and how to fix them (if possible at all, which is less probable for a norm conversion which keeps the audio but converts the video instead).
PS: VOLUME does not increase relevance.
Wow, that is complete ignorance.
Take a Region 4 DVD (NTSC 29.97fps) and a Region 2 (PAL25fps), from THE SAME MOVIE, and you will be able to verify that the length is DIFFERENT.
We were never shown a sample so there's no real way to know what's going on. But one thing's for sure and that is if anyone is confused or is causing confusion, it's you.