Do I have flawed reasoning?

  • Thread starter Thread starter thebigcheese
  • Start date Start date
thebigcheese

thebigcheese

"Hi, I'm in Delaware."
I currently record at 24/48k. Originally, I did that because I had done a lot of video work and thought that if I wanted to stick my stuff in a video, it made sense to record at 48k. But lately I've been thinking that really, it's just going to go on CDs, and it would make sense to do 44.1k and not have to worry about downsampling. On the other hand, when I get it mastered, they will probably have good converters anyway, so it won't matter too much, except that my computer might perform slightly better at 44.1k. I'm not going to go any higher than 48, so if I was to choose between 44.1 and 48, which makes more sense? If someone called me one day and said, "Hey, we want to use your music in our movie," would it actually matter that I recorded at 44.1?
 
I'm not going to go any higher than 48, so if I was to choose between 44.1 and 48, which makes more sense? If someone called me one day and said, "Hey, we want to use your music in our movie," would it actually matter that I recorded at 44.1?

No . . . it wouldn't make any appreciable difference. Most video editing programs will accept quite happily a sound source at 44. They will convert it to 48 when you do the final rendering. You should feel confident at doing all your recordings at 44.
 
No . . . it wouldn't make any appreciable difference. Most video editing programs will accept quite happily a sound source at 44. They will convert it to 48 when you do the final rendering. You should feel confident at doing all your recordings at 44.
I thought so. I mean, I've done lots of editing and added mp3s, I just wasn't sure if "real" editors would whine about that sorta thing. Alright, then I might as well give 44.1 a try, see if I notice any difference.
 
The only time you NEED to worry about 48k is if you want a composition to SYNC up to the video. Drop frame synchronization requires 48k to sit properly without getting off the lock to the frames.

If you're just going to dump a background song on a movie, then dont worry about it.
 
The only time you NEED to worry about 48k is if you want a composition to SYNC up to the video. Drop frame synchronization requires 48k to sit properly without getting off the lock to the frames.

If you're just going to dump a background song on a movie, then dont worry about it.


Jumping in here...

Do you mean for stuff like lip synchronization?? Not lip syncing, but lining up the sounds with the movements. Like if I video myself recording a part then put that recorded track in the video, it will be out of sync if I don't use 48khz??

Thanks,
 
Jumping in here...

Do you mean for stuff like lip synchronization?? Not lip syncing, but lining up the sounds with the movements. Like if I video myself recording a part then put that recorded track in the video, it will be out of sync if I don't use 48khz??

Thanks,

I think its to do with the fact video is broadcast at 24fps. 48k audio broadcast on a carrier wave with 24fps video gives a nice round figure of 2000 samples per frame of video. Audio sampled at 44.1k sent out on a carrier with video gives 1837.5 samples per frame, which obviously means it will be rounded to 37 or 38 so it will drop out of sync after a while. Can someone please confirm I am right in this? Its just a guess.

In response to your question, if the above logic is correct, you will have to re-sample at 48k to produce the right number of samples for the video.
 
I think its to do with the fact video is broadcast at 24fps. 48k audio broadcast on a carrier wave with 24fps video gives a nice round figure of 2000 samples per frame of video. Audio sampled at 44.1k sent out on a carrier with video gives 1837.5 samples per frame, which obviously means it will be rounded to 37 or 38 so it will drop out of sync after a while. Can someone please confirm I am right in this? Its just a guess.

In response to your question, if the above logic is correct, you will have to re-sample at 48k to produce the right number of samples for the video.
Most TV, at least in the states, is still shot at 29.97 fps, actually. I think they've started moving more TV dramas to 24 to make them more like movies, though. Music videos might also be 24. So, there's that to consider, just to make the math more complicated.

They can do the resampling fairly easily though, right? Will I lose any quality by resampling to a higher sample rate?

Edit: Since the above poster was talking about drop frame, I'm assuming he was thinking 29.97 fps, since that's drop frame and 24 usually is not, as far as I know. So your math may not be relevant to his point.
 
And 29.97 standard NTSC video drops a frame every minute, except those minutes that are divisible by 10.
(Hey I don't make up these screwy standards...)
 
Back
Top