Rendering: MP3 vs WAV

[x-angel]

Has anyone else noticed how AE sometimes has real problems rendering with MP3s versus straight wav files?

I racked my brain over a huge project continually crashing on me... :furious3:

Then I remember have AE 6.5 had problems just importing MP3s, so I just turned off the mp3 layer, and also told the render engine to not export the audio (just in case I had miss something in the other comps), and it's rendering just fine. (albeit too late for the project deadline)

Fortunatley, the project was a volunteer thing for my wifes school...so I didn't really disappoint a major client... it was just irking.

Anybody have any sort of explanation for this?

[Matt Grunau]

Firstly, I don't think you should be worrying about audio in AE in any way, other than as a reference for effects, or for things like cues from the director that you can follow. And most of that could easily be handled in you NLE.

There are some cool things AE can do with sounds, Time remapping for instance, but don't think of AE as an NLE. I've never had a problem with AE and mp3s, because I never export audio and rarely import it. When I have a clip that has audio on it, unless I need it for reference, I mute the layer's audio.

If you have to render out audio, render it uncompressed with your video. If you have to use an mp3 for audio, convert it to an uncompressed audio file with your NLE and use that new file in project.

Leave the sound implementation to your NLE, and even then, try not to use compressed audio. You are asking for trouble having you NLE decode and use mp3s in your project file. If you are set up for 16 bit 48k, import you mp3 into you NLE and render it out at those settings and use the render to work with. Less strain on the software and CPU (albiet minor unless you have multiple instances) in decoding, and less loss of quality.

[x-angel]

I kinda figure somthing like that (AE not really intended for final sound lay out)

What you suggested (using an NLE for Audio layout, and just having it as a reference in AE) is what I ended up doing, and it seems to work better that way.

I can see where using the compressed audio would present a "snag"

Thanks for the tips and insight.