Hi everyone,
I have to admit that for the past year I've have struggled finding a Logic/VSL configuration that was solid and promoted an optimal workflow using my new MacPro. After many attempts and numerous templates, I have found that using VE as a stand-alone versus using it as a plug-in, by far, provides the most stable development environment.
I feel compelled in writting this in case anyone else in the user community is using a Mac Pro (8-core), maybe they can benifit from my trials. Considering that we in the Mac world are still 32-bit, for me, memory limitations not CPU has been the real culprit. I hate freezing tracks and purging samples!!! By implementing the stand-alone approach not only provides the use of memory that has been sitting idle but is also proving to be the most reliable and stable. I would have to re-boot on the average of 3 time per day during a 12 hour session. For the past week since I went with the stand-alone configuration, re-booting is something of the past.
I'm currently using 4 stand-lone instances of VE, (Winds, Brass/overflow Perc, Percussion/Keys & Strings/Choir) each having 16 instruments, a wet and dry bus to control debth placement, with an Altiverb loaded in each. Additionally, I have each channel with the new VSL EQ, some with Exciters, the masters with Master EQ & Multiband and using the built in power-panner. I have all my samples loaded and so far have not needed to purge anything - just concentrate on composing.
I can see where having slaves is a real time-saver but for now, knowing that I am able to fully utilizing my machine investment is a real joy.
The real test was using the same sequence I wrote in Logic 8 last year using individual Vienna Instrument plug-ins. That approach was stable but required me to purge samples and freeze most of the tracks in order to obtain a smooth play-back. Even then, I was only using a single Altiverb , none of the VSL Suite tools, and high latency settings (Apogee ensemble). Using four stand-alone instances of VE requires no freezing or purging for the same sequence. With that configuration and in the parts of my sequence where most instruments are playing at the same time, the highest I've seen the CPU read is 35%.
My hats off to the VSL developers. Your team have provided a very CPU efficient tool. Keep up the good work.