Hi guys! I'm a long-time MVP consumer, but my first post. I'm working with a musical partner where we are creating recorded music for a live professional theatre production, much in the style of Pirates of the Caribbean, etc. I'm on a 4GB 2.66 GHz MacBook Pro ( Snow Leopard, Logic 9.1.3) and my libraries are all on an external 7200 FW800 Raid. I'm anticipating that at some point I'm going to choke, so I'd love some suggestions on maximizing my workflow. I'm expecting the cues to be no more than a minute or so, but I can see them easily being 30-40 tracks. Do you recommend stems? Does freeze work well in Logic (I've actually never had the need to use it!)? I just picked up the EW Complete Composers Collection before Christmas, so I anticipate I'll have to play (PLAY!) with the buffer and DFD settings to maximize CPU, but anyone have any suggestions (primarily Platinum Strings, BTW). Lastly, for some of the cues that will be too tough to play, the composer will be giving me Finale midi files; any suggestions on musical humanizing tips? Many thanks, Marcel

From what I've read by people whose opinions I trust, using a RAID system for sample libraries won't give you any speed advantage. In fact, it might be best if you dispensed with the RAID system and used individual FW800 drives (good) or eSATA drives (best). Being that you're using a MPB, though, I'm not sure what options you have for eSATA. That said, computers are notoriously flaky for playback of live virtual tracks. And Logic can sometimes be a bit flaky in the playback department, where one day a project will play back fine, but the next day it will spit out "disk too slow" error messages (stemming from playback of virtual instruments). So my emphatic advice would be to print stems for playback during the show. In fact, there are no advantages to running virtual (MIDI) tracks during a live show over running audio unless you plan to, say, sweep a filter of a synth part live or something along those lines. You're not going to be tweaking (say) the velocity of parts while the show is going on. Playback of audio tracks is so much less CPU intensive that you'll find safety and stability in playback of audio tracks. So the steps to take would be: 1) run your virtual tracks during production rehearsals and work with the FOH to adjust the balances, EQ, and so on. Once you've settled on a mix and satisfied with playback of the parts, print stems. To that effect... 2) Converse with the FOH mixer as to what he/she needs most in terms of flexibility without the number of stems becoming overwhelming. For example, it might be absolutely sufficient for FOH to have simple stereo stems of (say) strings, woodwinds, brass, percussion, and so on. They might want all solo instruments (regardless of the orchestral family) to be routed out of a single output, or particularly boomy percussion (gran cassa, bass drum) to come out of another. The rule for live playback is "keep it simple", in all respects, including... 3) Your computer. It should be a dedicated computer that has no access to the internet (Airport turned off), no active Widgets, nothing that will sap computing power. Energy Saver prefs should be set to [i]never[/i] let the computer or hard drives sleep, and the screen should not be allowed to sleep either. Finally, you should absolutely get yourself a good UPS (uninterruptible power supply) and run your computer and all hard drives from that at all times. Surge protectors are definitely insufficient because they don't provide backup power. A great place to get UPS's is www.refurbups.com. Go with something that outputs a pure sine wave, like the various APC "Smart UPS Tower" series. As far as humanizing parts that are too difficult to play, my sincere advice is to get players who can play them. You can spend inordinate amounts of time trying to humanize quantized MIDI data and still not have it sound good. Human comes from humans! :-) HTH, Ski

Thanks for the detailed and thorough reply Ski. Unfortunately, you've misunderstood my post, in that I wasn't clear in that what I was looking for was a way to help the workflow, that is, as I'm sequencing. I have no intentions on playing the multiple-track sequences live; I've giving them a stereo mix of each cue and they'll trigger them from their computer system. So I was looking for things to do to build up complex arrangements. We talk about giving the engineer stems, but decided in the end it would be better to alter a mix if they need more or less of something in a cue than to risk "screwing" up a mix. Also, one thing that I wasn't clear on is that I'm using a Drobo, which is LIKE a Raid but not really. So it's enclosure is holding four terabyte drives and holds the contents of all my sample libraries, including close to 800 GB of the East West collection. So fear, it's performance seems great. But you ARE right about humanizing. I am a professional pianist, but I admit that Alto and Tenor coeds slow me down (!), and I can probably step input the real tough runs if I need to. We're just going to be on a tight deadline, so playing in one part at a time will obviously take longer. But thanks for the input. Marcel

(I shouldn't try to post from my iPhone! Sorry for the typos!!) Marcel

OK, seemed to me at first like you wanted tips for playback. Anyway, let's talk about workflow then. You may find that your approach has to change on a piece by piece basis, depending on the orchestration of each one. Let's say your overture starts with a trumpet and horn fanfare and percussion, but then strings and winds come in and continue through the end of the piece. Certainly start with your brass and perc, and then record the strings from either the violins downward, or, the bass upward (top down being the way I usually approach this). Then add your woodwinds. At some point you may find yourself concerned with levels. Naturally, each instrument you add is going to drive the output hotter and hotter. At some point you may find yourself peaking on the output, at which point you can simply turn down the stereo output. Leave the master fader at zero. Let's say the piece ends with timps pounding, cymbals crashing, horns blaring. While it may seem tempting to write in those parts first to assess the peak level, it's just not practical to work that way. Assuming you write in the rest from the top, you'll likely find that the volume change "seam" between the the bulk of the piece and the ending will be a nightmare to fix. So work through each piece from bar 1 onwards. In terms of setting levels for each instrument, it's impossible to say where to set them without knowing the orchestration. In terms of starting out with enough headroom on channels, you'll probably only figure this out after you've gotten through the first few pieces. And in the event you find yourself running out of headroom on channels, post about that, as there are numerous ways to bring the levels down without damaging whatever mix you have in progress. HTH, Ski