Support |
At 8:24 AM -0700 3/14/05, Kris Hartung wrote: > >I have a friend who bought a notebook, but decided to cut corners on >the processor....I noticed a huge difference in latency and performance >with ProTools. We both had notebooks with similar RAM, XP, and hard >drive configurations. The only difference was our processors. It was >very interesting. On ProTools in the Playback Engine settings, I can >run my hardware buffer size at 256 samples with no latency. He had to >run his at 128, but then when he started to record it would crash >ProTools once in a while....lower buffer size, less latency, but more >risk of the crash. I found that 256 yielded no latency and no crashes >for me, unless I'm playing back a lot of tracks and mixing down...then >I bump my bufffer up to 512 where latency is not an issue because I'm >not recording. > This seems odd to me. While I am a Mac guy (love my PB 1.5 Ghz!), I don't think this would be different on PC. From what I understand, hardware buffer settings introduce various amounts of latency independent of anything else. If I set a buffer size of 256, that means that there will be a latency of 256 samples. I started using Protools LE a while back on various processors and every one of them has given me (admittedly not scientifically tested) the same amount of latency for each setting. I may have described my situation above incorrectly. I can't recall exactly if he had more latency on 256 than me on my processor. I just know that his system produced the buffer error constantly on 128 and occasionally on 256. For some reason, I think his 256 buffer setting did produce more latency than mine....now I want to go over and do the test again. > Is there any chance that you had the Low Latency Monitoring option turned on? If this operates in a different fashion, I'd really like to know, because whenever I select a buffer size over 128, the latency becomes very noticeable. Where is that setting? I haven't seen it in my version of ProTools LE. What do you have under Setup, Preference, and Processing? > It would be very interesting to me to know that if I just got a fast enough processor, I could set the buffer as large as I wanted without introducing latency! I would not be able to test what we're asking here without getting a better notebook and processor than what I have now. I don't know if the buffer size to latency ratio is relative or absolutute. One thing I do know, based on the test I did with a processor that wasn't as robust as mine, is that buffer errors are relative to processor efficency. Your guess is as good as mine! I'm going to go over to the ProTools user group on Digidesign's site and see if this is discussed anywhere. All I know at this point is that the buffer error frequency changes from system to system with the same buffer settings. What can we conclude from this? Kris All the best, Edwin -- Edwin Hurwitz Boulder CO http://www.indra.com/~edwin http://www.cafemontalban.com Location Recording Services