Well, it's a coarse estimation but I would say you need about at least 6 dsp's for mixing comfortably with the STM 2448, when you mix and use some stock EQ's en compressors. But then you can't do much else.
Keep in mind that the built in channel compressors have a bug when using the sidechain function: they don't remember the sidechain source after saving the project. So every time you recall your project, you'll manually have to reassign your sidechain source. Not a real showstopper, but quite annoying. And it slows the workflow a bit down. Certainly when you have a lot of channels using sidechain.
In that case what I do, is that I use a standalone compressor module and connect it to the input of a mixerchannel. The downside is that you'll get a little more "cableclutter" in your projectwindow, but the sidechain settings will be recalled correctly.
If I remember well, the channel EQ's also have a bug, but I can't remember what that was. I'm not sure but you should use all bands of the parametric eq in advancing frequency-order otherwise some parametric bands won't be recalled on next startup.
e.g. if you boost a freq on 4 kHz at the first band, and afterwards you add another band (nr.2) at 400 Hz something won't be recalled correctly. Maybe somebody can confirm this or correct this?
Anyway, I never use the mixer's built in eq's and compressors. I use the PEQ as an insert instead. (or heavier/better eq's if you have them). It's just as DSP-light as the mixer EQ, and bugfree.
