I am having trouble recording my final mix-down of a song (all midi track audio outputs and pre-recorded audio tracks from live instruments and voices) directly to one stereo track of Logic or Cubase.
The problem seems to center around GigaStudio 96. When I mix-down several midi tracks of GigaSampler audio output within Cubase or Logic (either with or without pre-recorded live instrument and vocal audio tracks) I experience very small drop-outs in the mix-down audio track under certain circumstances. Mix-down of GigaSampler audio output can be accomplished without problem with some songs (those which seem to use smaller Giga samples and therefore require less streaming from the hard drive at one time I guess) but doesn\'t seem to work with other songs (which are using larger Giga samples).
The number of pre-recorded instrument and vocal audio tracks mixed-down at once does not seem to matter (I have mixed down as many as 25 pre-recorded stereo audio tracks to one mix-down track with no drop-outs, pops, clicks, etc. However, no Giga audio output was recorded at the same time). The drop-outs seems to be from GigaStudio audio output only. The confusing thing is that I don\'t hear the drop-outs in the GigaStudio audio output as I am recording the mix-down track; it sounds great. I only hear it later as I listen to the mix-down track. If I record the same mix-down track and leave out the Giga audio output, everything is fine no matter how many pre-recorded instrument and vocal audio tracks are present. Also, if I pre-record the audio from the Giga output one or two tracks at a time, creating several audio tracks from the GigaSampler midi tracks output, and then mix-down all the audio tracks (both pre-recorded audio from live instruments and vocals and pre-recorded Giga audio output, everything is fine.
My computer is home built and consists of a Ausus CUSL2-C Black Pearl Motherboard, a PIII-1000 processor, 512 MB of PC 133 RAM, two Maxtor 60 GB ATA 100 hard drives operating at ATA 100 and an M-Audio Audiophile 2496 audio card with the latest multi-client drivers. It is really a very high performance machine. I have also optimized my machine by making all the normal system changes recommended by the Tascam system optimization paper found on their website.
Anyone have any ideas on what I may be doing wrong? Maybe this is normal procedure and I\'m not aware of it. I thought the computer system I built had more then enough power, guess I may have been wrong.
Steve, I think you are correct. If I record only one instrument at a time, I have no drop-out problems. Also, if I am recording only smaller (less number of voices) instruments, I have no problem with several at a time.
I have also discovered your work-around and record the large (i.e. GigaPiano, etc.) one at a time and then mix down. That works fine.
Then strange thing is that when I am recording several large instruments and the recorded audio has drop-outs I did not hear the drop-outs in the Giga audio as it was being recorded. Also, if I record all the Giga instruments within GigaStudio (unlike GigaSampler, GigaStudio has a built in audio recorder) I also get no drop outs.
Well, as long as I have a work-around everything is fine. Thanks for your help Steve, I appreciate it.
I don\'t believe your system is the problem, it sounds like a polyphony issue to me. Since I\'m using GOS and Giga LE this is a problem I know well. Do you nget dropouts if you only play one instrument at a time or is it only when you try to do the whole mix?
It\'s easy to run out of polyphony. I don\'t know the exact details, but I seem to recall that velocity layers count as a polyphonic part whether that layer is called for or not. (somebody correct me if I\'m wrong) If you have Logic why don\'t you record the individual instruments into Logic as audio and mix from there? That\'s what I do. I\'ll take all the 1rst violins (short and long bows) then the 2nd violins, then violas, etc.
We don\'t think that there is any wrongs with your system and setup. Otherwise, you could get clicks and pops in audio recording.
However, in order to do the audio-mixing-down tasks, you probably need faster CPU/MEM system, like 1.5GHz-CPU or better. That\'s all.
[This message has been edited by LHong (edited 11-08-2001).]