attack theory and other nonsense

[not Cereal]

ok so i dont feel good at all today so i'll try and be clear, but bear with me hear it may just be jibberish i am setting up some samples on my akai for the new band we are starting. i need a bunch of different sounds than what my current gear makes so i thought i would go about putting the new sounds on my sampler (because it doesnt cost money). well i have a lot of the sounds i need from libraries and other sources, and i have gotten a lot of the sounds to work great. the one i am having a problem with is "real violins" i have a library of sounds of real violin. almost every note is sampled at least only a few notes arent, and you can hear the bow riding on the strings which is very cool. they really do sound like a violin. there are a few different dynamics on each note, so i have like a "P, MP, and F" for each note. this sounds great after i looped each sample in bias peak (that took about 4 hours!). the violins sound like a real section behind music and if played like a bed, and thats the problem. the attack is all wrong. i set it to a slow enough attack (55 - not sure what that number rep's) so that it would sound like a string player starting and sustaining a note. this works really well until i try to play a line on the violin. the attack is too slow. if i speed up the attack than playing the sample like a bed sounds wrong. when i play this sound live (been doing it with a roland jx-1 so far - not too great) its both a bed and a lead. i'm not sure how to comprimise this live. studio i would create a different program of each and simple use the attacks for each that sound good - but thats not efficient for live - i need it on one keyboard on the same keys. its an expression issue, i need to "play" the samples. either i have to add more multisamples with different attacks and use the velocity to change between them (hard to play i imagine) or there is something i am overlooking. there has to be something i am overlooking (?) i like it when i find something that is a challenge. and sampling is very challengine for me. i dont think this way! its not natural for me. help!

[not Cereal]

heres a sample of the strings used as a bed. its a small mp3 file 988k [url=http://24.116.43.40/stuff/string%20test.mp3]strings.mp3[/url]

[randy clay]

I think the way you deal with this is to play the note ahead of or on top of the beat, to compensate for the time it takes for the bow to excite the strings. If you are sequencing the string section, just nudge it ahead a bit. YMMV,

[Salyphus]

Maybe you could set the attack to be varied by a continuous controller, like the mod wheel, and adjust it while you are playing?

[Gtoledo3]

This is a weird way of handling it, but you could use a lowpass/high pass filter type of thing, w/ an envelope filter. That way you could "wah" one range or another. Not natural sounding, but it will give "attack" to only a specific range of the keyboard. Do you have any kind of keyboard split function?

[fuzz]

just get an expression pedal to control the volume or a volume pedal will to do....old tricks will sometimes work wonders.

[alcohol]

Don't know what kind of keyboard you use, but most keyboards can be split. Have one side for leads and the other for pads.

[not Cereal]

these are all good ideas. since i am still building, i am trying a few things. the split mentioned is a good idea. right now i have a string patch with sustained strings on the top half and pizz on the bottom. this coould be setup different obviously because you can put any sample on any key. the neat thing is you can put [i]as many samples on a key as you want[/i] by duplicating keygroups on the same key and assigning 4 samples per keygroup.