My system is a fairly beefy system. Core I7, 64 GB RAM, 2 SSD drives (one for samples, other for audio), RME Fireface 800 and Cubase 9. Lately I’ve been getting a lot of disk caching and spiking where my audio will drop out for a few seconds, even on songs with low track counts. It’s especially noticeable when I resurrect old projects (some of them 15 years old from Cubase SX days). It was driving me insane because I knew it wasn’t a hardware issue and it seemed to be getting worse. For example, I would write a song with roughly 10 audio tracks and everything would be fine. 4 months later, I would open the song up and play and I would get audio track dropouts due to disk caching.
So… I did what everyone says you don’t have to do, I defragged my audio drive which is an SSD drive. Windows 7 reported it was 47% fragged. After it ran overnight, I loaded my song back up and everything worked great.
The only thing I can think of is my audio drive has data from 2 decades and it’s been trough 3 or 4 different drives, so maybe that really did scatter everything around… but for whatever reason, defragging it fixed it.
You shouldn’t really defrag a ssd drive…
If windows 7 or above it should automatically detect it and then use the built in defragger which will run a trim on the drive.
That’s a weird one, there’s no techical reason why a defragged ssd should outperform a fragmented one as there is no seek time associated with solid state drives.
You can benchmark your drives with CrystalDiskMark and see if your drives indeed underperform, should be around 500MB/s for most 2,5" SSD’s over SATA600.
You can also check CrystalDiskInfo for it’s health status, temperature, properties if TRIM is supported, transfermode (sata300 or 600) and such. These are freeware.
The lifecycle of a SSD is determined by a limited amount of writes of the sectors, when you defragment a SSD files will be moved around eating up this limited amount of writes. So don’t defragment a SSD.
Sort of, SSDs have a limited read and (particularly) write cycle limit, defragging is just burning SSD life for no reason. The only way I can think defragging could help an SSD is if over-provisioning wasn’t turned on, and some data is sitting on “bad” sectors until it’s moved by the defrag.
I like AusLogics Disk Defrag giving me graphical interface like XP defrag had.
It has an option to optimize SSD with special algorithm, but don’t know how that works.
But since random access on SSD has the same access time to any part of disk, there is less need than normal hdd.
There was nothing different I did between the time I had audio dropouts and the time I defragged, 0 dropouts now on the same song. not saying you should do it or its not a false positive, just sharing the experience.
In my experience some SSD drives are simply terrible for streaming applications. Maybe it’s the drivers, and not the hardware, but none the less, I’ve had some that are plain unusable for ‘streaming’, despite passing all of the latency and interrupt tests I could find to throw at it.
I had a batch of PNY drives that just will not stream continuously (neither long audio tracks, nor VSTi sample content) without glitching…and I’ve tried them on different host controllers, in different PCs (Intel and AMD), and with many different A/V streaming applications.
I’ve had good luck with Samsung’s mid-range drives for hosting samples/vsti content. For longer audio tracks I’ve gone back to using much less expensive platters.
