I fear I\'ve made a mistake assembling all my drum samples for GS160 and selling my hardware samplers.
I\'ve never had good timing.
ASUS CUBX @133mhz
256MB legend RAM
Soundscape Mixtreme v2.3 (latest)
Fujitsu 20GB (programs), Seagate 30GB ATA100 (for gigs on onboard promise IDE)
MIDIMAN 2X2 ISA MIDI card
Intel Bus Master drivers/DMA checked
VM off, vchache 8192, pci latency 128
Win98SE lite, PCI steering off/manual IRQ
All latest drivers.
Basically I get sludgy timing (not horrobly but enough to ruin the groove) and I get this MIDI choking every few minutes or so where everything hiccups/stutters for a bit. No clicks or pops. I went thru all that, all devices live happily together.
Does anyone get ROCK SOLID timing for drums, especially when lots of single hits in 16ths and bass and keys are running.
For me CPU and voices aren\'t anywhere near maxed out when this happens and there is no PU or PO on the mixtreme. It really shows when I have loops on GS as well as many oneshots. The loops are triggered so inconsistently.
Is there any specific settings on the CUBX that are worth trying? What should my vcache be at. Does ISA cause polling and interuptions im streaming? Would I be better off with a com port interface or a PCI one like thru SBlive or? Am I better off with ME or windows IDE drivers instead of Intel?
My guess is that it is the fact that you have the gigs drive connected to the promise controller.
I have found that drives connected via a plug-in controller card, can have extra 20ms or so tagged onto the access time - and others have reported exactly the same when using the promise (etc) controllers built onto some motherboards.
An easy way of checking this is to download sisoft sandra lite from zdnet or so, and run the disk benchmark test. You should see something like 5ms access time on the fujitsu drive, but you will probably see something like 23ms access on the seagate.
If so, you will experience some problems with giga. You can try putting the seagate onto one of the \"native\" onboard controllers (i.e. not the promise one), otherwise you may have to try and rather use the Fujitsu drive for the gigs. While I am not familiar with the particular ATA100 drive, I do understand from our in-house hardware guys that many current ATA100 drives have problems working nicely with ATA66 or 33 controllers, so I am afraid you will have to play around a bit, just connecting the seagate to a different controller may not solve all the problems.
I have a timing issue too, that I noticed when I did some dance music. When I stacked three bassdrums on top of eachother, they don\'t hit at the same time. Not much, we\'re talking maybe 1ms difference, but it\'s big enough to cause weird phasing problems. Sometimes a hit sounds good, sometimes it does not, you just get very different sounds depending on how tight those 3 bd\'s are. I can\'t say for sure that it\'s a GS problem though, it might be the connection from my sequencer computer to GigaStudio...
I have a dedicated computer (P3 900mhz, 2 drives, 1 for programs, 1 for .gig\'s) running GS with no timing problems at all. My other computer runs Pro Audio 9. I use GS for my \"virtual rock band\", drums, bass, guitars all going via GS. I have seperate channels for bass guitar, rythm, lead, chords, bass drum, snare, hihat, cymbals, toms,etc. All rocking away very nicely I might add.
I have to be honest here, I don\'t do techo or dance music so I really don\'t understand the need for a mutli-kicked bass drum or playing two bass drums at the same time doing the same thing. I would think though if I needed this kind of effect I would try to have a sample of these two drums already playing at the same time and then trigger this sample when needed. By creating the mix ahead of time as 1 sample, you bypass any timing problems that might occur when playing two samples that need to occur at the exact same time.
All I\'m saying is that putting a kick on two or three tracks is an easy way to show up timing blur. It\'s easie to hear when the fronts of a few kicks line up (or don\'t), than compare timing between a guitar sample and a kick, or even a kick and snare.
Firstly, don\'t get me wrong, but you don\'t hear 1ms differences unless in phasing if it appears.
Moreover, if you layer 2 identical kicks, you can just as well raise level of the kicks by 6db. This is more exact as you get the maximum raise in volume and no phasing can appear.
Also, you can get rid of the sequencer midi clock instabilities by using an external clock source and having the sequencer synch to that.
I see no musical use in doubling two identical kicks, you\'ll just get phasing problems.
I have just found that doubling a very transient sound like a kick with a similar kick on another track is an easy way to hear timing slop:
\'Dum Dum Dum dDum Dum\' etc.,
I noticed in our old Cubase/Atari days that every time I ran the sequencer, even off external code, the exact position of midi notes in the tracks shifted ever so slightly. Whether this was due to midi, Atai, Cubase, S770 (doubt it), HD recorder or whatever - it was just easier to hear the slippage when you used a couple of short transient sounds as guinea pigs.