show first message show previous message Showing Logic-users Message 241290 of 246033 show next message show last message

Forum Index | Read LUG: Policy/Rules Messages Threads Digests | Post New Message | Search!

From: "Sascha Franck" <S.Franck@gmx.de>
Date: Wed, 30 Apr 2008 at 5:21:25 AM
Subject: [LUG] More drive tests, was: Ext. 2.5 USB drive test
Message #241290
This is a reply to #241233.
Ok, just stressed my system a bit more, to know what it might be capable of. System: Macbook 2.4, 4GB RAM, 250GB Samsung 5400RPM internal drive, OSX.5.2, Logic Studio, all updates installed. Audio interface used was an M-Audio FW-410. No OSX or Logic tweaks others than switching off backgrounds, screensavers and Bluetooth. And I'm using 128 samples buffersize. Airport permanently switched on and connected to my router. I used a 16 bar cycle at 120BPM. For these tests, all files were starting straight at the same position. To actually survive the test, at least 2-3 cycles should play through without any overload messages, plus I wanted to be able to jump to whatever locator point while Logic was playing back. First thing I tested was the internal drive, using stereo tracks at 44.1kHz, 24bit. Now, prepare yourself to pick up your jaws back up after they dropped to the floor! I managed to play back 137 tracks! Yes, that's right: One-f***ing-hundredthirtyseven! I couldn't even remotely believe this myself, so I double- and triple-checked whether I was doing something wrong (using multiple instances of the same audio object, using copies rather than originals, using objects that weren't actually routed to an output, etc etc.). I wasn't. Really, I have NEVER seen a laptop playing back even remotely as much tracks from the internal drive only. I have been pretty much pleased with both my old Windows laptop and the old Macbook already - those managed to play back around 50 tracks. I still think there's something wrong or some magic going on. But all too apparently there isn't (I'm not exactly new to stresstesting whatever computers). The next thing I did was running the same project from the external 2.5" buspowered USB drive. That one "only" managed to play back 128 tracks. Still amazing, though. I proceeded playing the project from an external PSU powered 3.5" USB drive, carrying a 500GB 7200RPM IDE Seagate disk. I could play back "only" 122 tracks. This might be related to several factors, no idea what it actually is: - The drive used is an IDE one. - The case used is from Trekstor, my bus powered USB cases are from ICY (which are wellknown to do pretty fine external drives). - The drive has 2 additional partitions on it, one NTFS, one Fat32 (so I can move files back and fourth between OSX and Windows). - The drive is made by Seagate whereas all my 2.5" drives come from Samsung. - A combination of all that. Anyways, pretty much stunning results so far, so I proceeded to some 96kHz, 24bit tests. From the internal drive, I managed to play back 51 tracks. Just about as much as I could use on the old Macbook - at 44.1. The external 2.5" drive allowed for 44 tracks. And the external 3.5" drive only allowed for 40 tracks. These are still mighty fine results, if you ask me. And along with the other results, it's really even more amazing, especially considering that OSX and Logic have been wellknown for their notoriously bad disk performance for quite a while (basically in the "G5 era", using Logic 6.x), at least compared to most Windows PCs. Right now, it seems to be more like the exact opposite. Ah well, finally I did a last test. Still at 96kHz, using the internal drive, I started adding Platinum Reverbs to the track inserts. I thought something like "now let's force that little bugger to create some swap files to finally bring it down to it's knees". No way. I could add a PlatVerb to each of those 51 tracks without any noticeable performance leaks. I could've probably even added two of them to each track, but well - enough is enough. Really, to me, these are all mindblowing results. And mind you, unless the disk load was really at maximum, I haven't even seen a single overload error. On my system, those "unforced errors" seem to be stuff from the past indeed, at least so far. Great! I am really wondering whether using FW would probably increase performance again (at least for the external drives), but so far I don't have a single FW disk case. But well, for my personal stuff, I probably won't even need one, apart from a little extra convenience sometimes. And I'd guess that at least FW400 won't show much improvements over USB 2 anymore. Regards Sascha ------------------------------------
Viewed 578 times, 3 replies, 36 messages in thread. Reply to this message. Read this thread.

Forum Index | Read LUG: Policy/Rules Messages Threads Digests | Post New Message | Search!


© 1994-2008, All Rights Reserved.