Artifacts in WAV's

Hello all,

So I'm finally ready to export some music as WAV's (rather than MP3's, which is what I normally do to listen to it), and when I export the songs I'm hearing random artifacts. Perhaps one per song, just a note that all the sudden gets louder at the end, or a couple notes that do. Then I don't change a thing, just re-export again, and now it's a different note that's bad. I'm using Project 5, with Virtual Grand Piano, which runs on the Kontact player, on a CPU with 1 gig RAM. Obviously there's some sort of glitch I don't understand.

So now it looks like I have to choose between exporting WAVs (and listening very carefully to every song, and then re-exporting ad-nauseum until they're perfect) or exporting to MP3's (256?) and then burning that way. I know I will lose something by going to MP3's but am not sure what.

Any help or advice is appreciated,

Sam

Whoa, you just lost me somewhere. I can hit "freeze track" and then export and that should fix it? And I don't follow this at all: "You want to start the mix with just audio tracks so the sources are "fixed" in place."

Thanks,

Sam

ORIGINAL: samhoff

Whoa, you just lost me somewhere. I can hit "freeze track" and then export and that should fix it? And I don't follow this at all: "You want to start the mix with just audio tracks so the sources are "fixed" in place."

Thanks,

Sam

The idea is that software synths driven by midi don't always produce the same performance every single time. In fast that type of slightly random output is often a "feature" to make the insturment seem more real or human played. Even hardware synths might do that. A software synth might even make different output based on the CPU power available at any given moment or simply have a bug that would produce a "mistake". The reason for the change is no important because there is nothing you can do about it and it's easy to avoid.

To make sure this can't happen you simple make a recording of the output as a wav file or "clip" so that it can't change from one playback or export to another. There are two main ways this is done. For example you can use the freeze function. That does a sort of in place bounce on the same track. This is quick and easy but not always the best way to do it. What I like to do is bounce (with the midi and synth selected) to a new stereo track. At that point I can archive and hide the "source" tracks and work with just audio tracks as if I had recorded a real band. You know for sure that a wav file will not change on you so if you get it just right in the mix it will stay that way and play back the same every time.

Be aware this is only one thing that might cause the export to differ. Your choice of pan law, sample rate, bit level, dither, etc can affect the output wav file. You should export the same way every time to make sure it's the same. For example I always export at the sample rate of the project and at 24bit. This is because I do mastering in Sound Forge and will be doing the resample (if needed) and bit reduction to 16bit with dither as the final steps after I do any mastering on the song like compression, EQ, trim and fade ends, etc.

if you freeze the kontakt synth track and then solo it and play that track back can you hear the artifact? if so it is certainly a glitch in the plugin. you can also try bouncing to tracks with "fast bounce" disabled, i've had a few soft synths that behave strangely when fast bounce is enabled, BFD for one.

What, exactly, would I lose by saving to MP3's? I mean, that's ALL I've ever done in the past, I think around 256 KB (is that a choice?) and I certainly can't hear the difference between that and a WAV....

Thanks,

Sam

ORIGINAL: samhoff
What, exactly, would I lose by saving to MP3's? I mean, that's ALL I've ever done in the past, I think around 256 KB (is that a choice?) and I certainly can't hear the difference between that and a WAV....

mp3s get pretty close to WAV quality but it depends on the encoding engine and the settings. a few years ago i did a test between a 320kbps mp3 and a WAV, the difference was immediately apparent on my studio headphones, listening back to back. since then i've heard some mp3s that i honestly could not tell the difference unless i listened to them back to back, but there is still audio information that mp3s get rid of to save space.

as dean said, you can try wma lossless. i use FLAC (free lossless audio codec) which is less popular but open source i believe. doesn't work in windows media player without some contortions but most other players can handle it.