Acid3- Disk streaming vs ram loops

[williamconifer]

Greetings,

I am running Acid pro 3. I have a 5:15 min project I am wrestling with. I live recorded a 4 min. Rhodes track directly into Acid. I wanted a real live feel.

My project has has grown to the point that I am getting dropouts while playing it back live. I am trying to optimize my tracks to reduce the overall computer load this project is creating so that I can do the final mixdown. Would it be better to split up the 4 min. Rhodes oneshot into 6 or 7 15 sec. loops? or not. I got two long oneshots (including the rhodes track) and a couple small (30 sec) oneshots and about 20 or so loops. Which is better 1 long one shot or 6/7 loops?

Many thanks
jack

btw I have already remixed loops/oneshots that had live fx on them to save on computer cycles.

[AlChuck]

It should be able to play that back without a problem. Have you tweaked your system for use as an audio box? There are many things that could affect it. Best thing to do is find the recommended tweaks for your OS and try them if you haven't already.

[Fangar]

What kind of system are you running? What are your system resource showing? Do you have all other programs and drivers that usually run in the background off?

Fangar

[williamconifer]

I am running a beige box Celeron 333 with an Asus AT motherboard, with 128 meg in ram, 5400 rpm harddrives, sblive with APS drivers and Win 98se.

I know my setup isn't really ideal for this but I'm not putting any more money into a computer with an AT form factor. I want to get a barebones AMD 1.6 XP system and build on to that. It's so cheap nowadays.

I peg my cpu utilization at 100% when things start dropping out. Acid shows my memory using 56 meg. I think I need to be more strict with what is running in the backround than I am now.

I am just wondering if there is a performance benefit in running lets say a 30-60 second clip as a beatmap or one shot or loop? Or is it better to chop up a 4 min one shot into 7 15 sec loops and build the track as loops?

Any thoughts or suggestions on performance tweaks is always appreciated.

Thanks
Jack

[barefoot]

I think splitting the event into loops will make things worse. Loops are stored in memory and require extra processing even if they aren't actually being stretched.

Two things you can try are in the Audio Preferences setup (Appendix A in the manual). Make sure you choose the driver for your specific soundcard rather than using the default Microsoft Sound Mapper. Also bump up the playback buffering. As the manual says, you want to increase it just to the point where the playback gaps disappear.

And yes, try to do as much pre-processing as possible. Also consider post-processing. For example, if you want to add a bit of reverb to everything to help gel the mix, do it after the final mixdown. Save the dry mix as a .wav file and add this sort of stuff later.

Oh, and one more thing. Try to keep all your sampling rates the same, and don't use anything higher than 48kHz. Converting different sample rates also eats up system resources. I do everything in 44.1kHz since I use a lot of samples from CD's. Convert all your events to the same (lower) sample rate as part of your pre-processing. And record all new tracks at this sample rate.

Good Luck
barefoot

[williamconifer]

Thanks for the tips! As far as post processing, should I use a Volume Maximizer or WaveHammer in post processing in Soundforge or on the mains as I am mixing down. I don't want to compress the life out of my projects but I want them to be loud enough for the listener.

Thanks
jack

[barefoot]

Definitely afterwards in Sound Forge.

In fact, I would forget that these tools even exist while you're mixing. Work hard to get your mixes as hot and punchy as possible without these (this might mean compressing individual tracks like bass and vocals). Mix down and save the file as 24 bit. Sparingly apply your maximizer in SF. Then dither down to 16 bit. I get much better results this way.

barefoot