Besweet and virtualdub to change audio to suit converted 25 fps avi

[AceRimmer667]

I am changing some videos i created to 25 fps, as they seem to work better in my divx compatible dvd player. I have used besweet to change the audio, and it works absolutely fine, and in synch. The problem is that the 25 fps compatible audio is slightly higher in pitch than the original, which is very odd. It happens with anything i try and convert to 25 fps. Is there better software out there to do it, or even software that will lower or higher the pitch of my audio, without changing its length - therefore making it easy to add to the video? thanks for any help

[Adub]

Unfortunately, it is something you are going to have to live with I believe. Even professionally done NTSC -> PAL conversions have the increased pitch.

[sneaker_ger]

Avisynth has a plug-in built in that can do what you want:
http://avisynth.org/mediawiki/TimeStretch

[AceRimmer667]

its not a major issue since its a very slight increase in pitch, i just wondered why the program i used would do it. I will try avi synth out. Good to know professional programs or ways to stretch audio may also change the pitch, at least i know the tool i used wasnt playing up. i thought perhaps if the slight change in pitch wasnt there then perhaps quality would be lost, or something along those lines whilst stretching or shrinking its length to suit the fps? i will check avisynth out to see if it can stretch/shrink audio without it changing. thanks to both of you

[Skelsgard]

When converting 23.976-->25 fps, the audio can be dealt with in many ways. One of them is what I believe you´re doing: adjusting the sample rate to fit the fps change.
So 48000 Hz at 23.975 becomes 50050 Hz at 25 (48000 x 25 / 23.976 = 50050) , which keeps perfect audio/video sync. As the sample rate increases, so does the pitch. This way is lossless, as you just change the amount of samples played per second, not the amount of samples in total. If you set the sample rate back to 48000 Hz, the pitch will be restored to original.

Another way of changing the speed alters the total number of samples, thus allowing to retain the sample rate and the pitch, but this is not lossless as it usually works with methods like splitting the audio every N number of samples, displacing, then repeating samples in a given position to fill the gaps. If done properly and within a threshold of length, the result is almost indistinguishable from original.
If done pass that threshold of length or with low qualty tools, is highly noticeable and annoying (speech sounds like speaking while gargling water).

This last method might work for you if you're interested in keeping the pitch while stretching audio.
Some tools are Avisynth's, Audacity and Audition's stretch filters.

[AceRimmer667]

thanks for the info :) i guess id rather that than restoring to the original pitch and having to lose quality, although it would be good to try out to compare to the original