I’ve run into a bug with the way WL 7.2.1 deals with gain info in AIFF and WAV headers. I do a lot of sample development and matching levels is normally done by changing the individual levels of each zone. When an instrument is pretty far along in development, it’s common to export the samples and have the hi/low key info, detune, etc. written into the sample headers so that the samples can easily be moved from one platform to another. The zone volumes are normally mapped to the sample Gain setting in the header. Unfortunately, WL not only doesn’t honor the setting of this parameter when it saves, it totally rewrites it to ridiculous values. For instance, I have samples that started with +5dB of gain that come out of WL with +41dB written in the header. Believe me, it’s a very bad moment when you import one of those samples into a soft sampler and play it back. I’m fairly new to WL and enjoy using it, but this bug makes it impossible for me to use it once I have info in the headers I want to keep. It’s especially bad when I want to use the batch processor to change all of the files. It’s obviously not worth the extra work of reassembling all the gain info, so I use other apps now.
The bug is slightly different between AIFF and WAV files. The AIFF headers are altered to bizarre values, but WL always saves WAV files with the Gain set back to 0. Unfortunately, this is a real show stopper for me, so I’d love to have it fixed.
I’ve attached a sample file with the gain set to +5. If you alter this file in WL and check the gain after you’ve saved, you’ll see the problem. Trombone test +5.aif.zip (240 KB)
Thanks…
WaveLab does not write any gain info in headers. But WaveLab tries to preserved the existing meta data that was present when the file was opened.
Where do you see these values, -5dB, + 41dB ?
Sample Manager, KeymapPro, Halion 4. In the last two cases, the gain value appears as the zone volume. Sample Manager is probably the most convenient way of quickly checking.
Yes, I assumed that WL was supposed to just preserve these values, but in this case, WL is actually changing them.
PG, did you ever get a chance to look at this? I’d really like to use WL7 more for this project I’m working on, but this is a deal breaker for me, since it actually ruins work I’ve already done.
It’s a little frustrating that I’m doing a project with Halion 4 and other third party editors I’m using seem to be more compatible with the Steinberg sampler than the Steinberg wave editor is. Plus, I have a feeling this is going to be very simple to fix…
Same programs I mentioned when you first asked the question.
Since Halion 4 is a Steinberg product, it should be easy to get a copy or, better yet, maybe you should talk to the H4 developers. Leaving sample header info intact is a fairly big deal in sample development world since editing will definitely have to take place and nobody wants to go back and repeat work that’s already been done in another application.