should i resample SFX from 44.1k to 48k to reduce CPU load?

Tim Young

Tim Young

Active member
i'm working in soundtrack pro on a project set to 48k. i have a small sound effects library from cd that is 44.1k. i'm thinking that STP must be resampling the media i import on the fly during playback since the sample rates don't match, so i'm wondering if it would be a good idea to go through the library and resample the media to 48k so that when i import them STP won't have to resample. would this increase performance by decreasing CPU load? if so, is there an easy way to batch convert them?
 
you could create a preset in compressor to do a batch conversion. I wouldn't bother with it myself, but I suppose it depends on your hardware.
 
I doubt that converting will lighten the load much on your CPU. STP is pretty processor hungry from my limited experience.
As a general rule it is better to keep all your sample rates the same so you don't get in trouble down the line. But if your strictly in a FCP -> STP -> FCP loop it doesn't make much difference. Out of curiosity how are you finding STP to work with?
 
i am literally just beginning with it, and haven't done a real project yet, so it's hard to say. i love the interface, and it seems much more taylored for post sound than digital performer, which is what i've been using. it does seem processor hungry though - i've been getting playback overload messages (but this is on a G5 with only 1gb of ram which will soon be remedied). i wondered if this was partly because of the sample rate issue. but the way i see it, STP will only get better. i never used soundtrack pro 1, but from what i hear there were a several early bugs that apple did a good job of fixing. and if you compare early versions of final cut pro to what it can do now and project that to soundtrack pro, i think it's pretty exciting for a version 2 release.
 
It's very picture editor friendly. So if your coming from say FCP it should be a nice step up. It's certainly more post friendly than music applications that started off as sequencers. They always seem to come at audio as if it were a MIDI sequence..? Glad your liking it.
Things that should help performance would be a good graphics card (don't know what you have but if it's a 64meg card and you have 2 monitors it will have an impact) this is true for more than STP, BTW. Memory is good. I would probably look at 2 gigs or more. One thing that can have an impact is how you got from FCP to STP. If you did a sent to STP from FCP then all the production audio that was from video came over as video clips, so your playing video files and not audio files. This might have an impact if you have a lot and or your drive is not fast enough. It doesn't make as big a difference as one would think, so unless you have a really large number of production tracks (like 8+) I wouldn't worry about it. If you went out OMF or AAF this isn't an issue at all.
Track count can have an impact. This is true of most DAW's but is usually not a problem until your over say 20 tracks. Somewhere between 20 and 40 tracks is generally the limit for "native" (those with out dedicated DSP chips) DAW's. Everybody lists some huge number of possible tracks but practically speaking I have never come across a "native" DAW that was smooth over 40 tracks on a system that was not very optimized or had some MONSTER CPU available.

Also if your doing a lot of track processing with plugins you may be able to free up some CPU time by "printing" the FX. I forget how this works in STP, in PT you would "bounce to disk". There is some equivalent function in STP. This is only true for track plugins, the FX on clips are rendered so should have no impact on playback.

Good luck and have fun.
 
thanks for the tips! i checked out my graphics card and it's 256MB. i'm planning on adding 2 more GB of ram. so about the issue with sending from FCP as a multitrack file to STP, if i did output as an omf, how would i get the video in? would i just do a render from FCP of a .mov file without sound and bring that in? would there be sync issues if i did that?

i'll check into printing tracks. i have done that in digital performer to save CPU load.

thanks again for the help.
 
As I said unless you did a send to of a lot of tracks that were video in the first place it shouldn't be an issue. But for an OMF yes you would export out the video as a quicktime DV stream (I'm assuming the video is DV) and bring that into STP. A catch with STP is that the video ALWAYS starts at the head of the project so you have to make sure that you have video before any sound you want to put in. But you should already have the video if you did a send to STP.

One other thing that could cause some performance hit is that if in the send to dialog you sent the video as a reference video instead of a fully rendered video. If memory serves there are two choices one says something about being faster, that is not the one you want. You want the slower of the two options. That will give you a fully rendered movie. The faster option is a reference movie. Things have gotten better but historically the biggest reason not to use a reference movie was that they weren't very portable. They do, or did?, use absolute references so moving the files at all caused all kinds of problems. In theory there shouldn't be much of a hit with a reference movie but??? seems like having to jump around to find all the little bits is going to add up.
Anyway if you did a fully rendered movie in the send to you already have the QT to use with an OMF. In theory the OMF should come into the project in the right place since the movie and OMF were from the same time line. You can check by phasing the tracks against the guide track from the video.
 
You must log in or register to reply here.
Back
Top