Re: QS8 dead spot
Posted by sam on February 01, 1998 at 22:24:55:
In Reply to: QS8 dead spot posted by Steve on January 28, 1998 at 12:11:58:
Here's what's happening: The QS has about 4 megs of piano samples for both left and right channels. If I do my math right that's about 44 seconds of piano sample per side, and this is divided up into a multisample, with each sample covering about 7 half steps. The sample that sounds weak is centered at the F two and a half octaves above C3, so it covers the D to the G#.
Alesis gets some additional mileage from the piano sample by providing versions that are pitch shifted, or move the sample start point back to cut off some of the hammer part of the attack.
The sample that is the most interesting is the 3 versions that are not pitch shifted. The GrandPno sample is the complete version as recorded with a hard hammer strike. The same sample is used for the "NoHammer" and "SoftPiano" samples, each with the sample start point moved back a bit later. This works pretty well on most of the samples in the multi-sample, but for the one sample centered on F it gets rid of the hammer sound altogether, so when you hammer these notes the attack is just gassy.
These 3 samples are layered together in the VeloSample that is used in some of the better Alesis patches, and around F you can really notice when it velocity switches between the GrandPno sample with the attack and the NoHammer without it.
Possible fixes: Stay away from the VeloPiano, NoHammer, and SoftPiano samples (this is not a very good solution!) Use a stronger attack envelope on these samples so the envelope provides the percussion where the sample doesn't.
Check out the piano patches on my web site (www.best.com/~sams/alesis/) - I think they're better than the Alesis ones and I tried to improve this area of the sample.