Hard Drive Preference: Lacie or Maxtor?

[qo]

I just tested the Maxtor 5000XT FW400 drive with Xbench. It's sequential writes are about 12-13MB/s. FW400 bandwidth is (theoretically) 50MB/s.

(X * 82000 * 24)/8 = 13MB

Where X is number of tracks. Solving for X gives 9 tracks. Extrapolating to 44.1 gives 18 tracks, give or take a track or two.

If you RAID-0 two of these, you might expect somewhat less than 2X improvement, so you'd be looking at 26MB out of 50MB FW400 bandwidth. The above doesn't take into account drive read performance for track playback. Xbench's read results (for sequential reads) were 5.32MB for 4K blocks and 14.98 for 256K blocks. You CAN set the RAID block size in Apple's Disk Utility, so I'd probably go for a larger block size. The tradeoff being that, for small files, large blocks are inefficient and you waste a lot of disk space. But, for audio, you're going to be utilizing 100% of each block (for the most part) when recording to a newly formatted RAID array.

[pcm]

Originally posted by qo:
I just tested the Maxtor 5000XT FW400 drive with Xbench. It's sequential writes are about 12-13MB/s. FW400 bandwidth is (theoretically) 50MB/s.

(X * 82000 * 24)/8 = 13MB

Where X is number of tracks. Solving for X gives 9 tracks. Extrapolating to 44.1 gives 18 tracks, give or take a track or two.

I gotta wonder about your math, and that ANY drive made today would perform that poorly. Even the 5400 rpm internal drive in my POWERBOOK let me write 32 tracks for 1 to 1.5 hours at a time, all night long, at a recent concert.

If you are getting rates that slow, I suspect your fw interface, as the older computers had rates that slow. With a new drive and a modern computer or fw card, you should see rates of at least 30/megs per second.

On a side note, I swore off Maxtors a long time ago after several failures. I have many Western Digital 8 meg cache drives that have done well for me, and of course Seagate is a top drive as well.

For audio, I see no reason to do raid arrays, especially considering that if ONE drive goes down, you lose the data on both. I had that happen once, and ain't going there again.

[qo]

Hi pcm. The math is correct. At least I'm sure it is. Maybe someone else can run the same calculation and correct me.

The performance figure is using Xbench, so it's empirical data. Not sure what else to say about that.

The drive was purchased two years ago. I'm sure there are faster drives since I use faster drives. I only tested the Maxtor since it was laying around and was similar to what the original poster, Slickie, was asking about. I wasn't recommending it, or any other drive.

With respect to RAID-0, I've been running tests in another thread where it does appear to be overkill with 10K SATA drives. But for slower drives, I'd think it would help. But, I don't plan on testing since I've finally got a decent working setup and so can spend my time tracking now

As far as risk of data loss, I just don't see it. I format the RAID-0 array, copy a project from a RAID-1 (mirrored) array (used for storing versions of works in progress) to the RAID-0 array. Do my tracking, and then copy the new version back to the RAID-1 array. I agree that, if one of the drives in the RAID-0 array goes south, I'll lose a few hours work. Not much different from using a single drive though, right? I agree that using RAID-0 for long-term storage isn't a good idea at all.

[dannyjm]

I like my maxtors. I have some firewire external and a 250G internal dedicated to recording and VI's.

Dan

[qo]

OK, not being one to put blind faith in math, even if it IS correct math , I just ran a test recording to this FW400 Maxtor.

I was able to record 18 tracks at 88.2/24 easily. I then played these back while recording to another 18 tracks. Again, this went well.

So, obviously, pcm has a point. I'd really like to see a way to take a benchmark like Xbench and gain a reasonable understanding of what a drive is capable of. But, it does appear that this isn't possible using my methodology. Anyone have any ideas as to why not?

Thanks for questioning this pcm. I wouldn't have run this test had you not.

<small>[ August 06, 2005, 08:04 PM: Message edited by: qo ]</small>

[qo]

Arrrrggg. No, the math is not correct. Actually, it IS correct for the way the parens are laid out. But, the equation should REALLY be:

X * ((88200 * 24)/8) = 13MB

That is, 88200 and 24 are bits, and we want to convert them to bytes. X isn't bits, it's tracks, heh. If we solve the modified equation for X, we get 52.8 tracks.

I'll try to run this many tracks now and post a reply in about 20 minutes. My apologies pcm!

[qo]

Recording 52 tracks at 88.2/24 and after about a minute, the Record buffer hits 100% and DP throws up the following dialog:

File buffer not ready (Disk is too slow? Too many tracks?)
MotuAudioSystem Mas MasRecorder.cpp: 347 [1:0])

I then dropped down to 45 tracks and tried again. This time, recording was fine for 2:54 seconds and then Record buffers finally hit the wall. DP also crashed in PerfFoundation.dynlib.

After restarting DP and dropping down to 40 tracks, things all went smoothly for a 5 minute recording.

So, what to make of this? I have a theory (well, I always do, right or wrong, hee). XBench gives stats for two different block sizes (4K and 256K). The drives I've been testing likely have a block size somewhere in between these two values (I don't know what it is though). So, this would explain the descrepancy. Right?

[BobK]

I'm using an OWC Mercury Elite 160 GB I bought last year. For an extra $10-20, I got the one with Firewire 400 and 800 and USB 2.0. Maximum load during tracking has been 13 tracks going in, and playback of 30-something tracks. Not a huge load, but I've had NO problems with this drive at all. It was just plug and play. And it's quiet.

I recall reading an article in EQ that spoke well of OWC drives. One advantage they cited was that you didn't need to load any special drivers, in contrast to whatever other brand(s) they were discussing. I have a friend who does a lot more pro-level recording, and he's had problems with La Cie and Western Digital FW drives, though he's also used a lot more drives overall. When I need another drive, I'll probably go back to OWC.

My $0.02 worth.

[XYZ]

Don't worry aobut it. Make sure you have a TESTED backup system. As you can see, everyone has good and bad experiences with drives. They are all very competitive. Ive had good luck with all three of the brands mentioned. Typically, the newer they are the quieter. I have a DeskStar, Maxtor, and LaCie500 for backup. All work quite well, but I would not depend on any of them. Also, Run Disk Warrior on them once in a while. It is amazing how it can find and fix problems that no other package can. That is about all it does, but it does it great.

[bweston]

I've had good luck with the Pacific Pro Audio FW400 drives. They're very quiet (no fan, convection cooling):

http://www.pacificproaudio.com/drives.asp

Also, good luck with the EZQuest Cobra.

Bob

[Shooshie]

Originally posted by Bob Weston:
I've had good luck with the Pacific Pro Audio FW400 drives. They're very quiet (no fan, convection cooling):

http://www.pacificproaudio.com/drives.asp

Also, good luck with the EZQuest Cobra.

Bob

I wish I'd known about this company before I bought this Maxtor drive about a month ago. Basically the same prices as other drives, but looks much quieter, and takes up less space. I like convection cooled things. No fans. Fans have a habit of going squirrely in the middle of projects where you really need them to sneak around on tiptoes. All of a sudden, they just can't take the pressure of being told "ssshhhhhh..." and they just burst out sceaming a High C with flutter-tongue. It's almost a syndrome among fans. Should be called Trumpette Syndrome. Sort of like they want to audition for the part, not realizing that they're only good for one note.

So, I look for drives without fans, but I only own two (by Que and Western Digital). The Glyph and Maxtor both have quiet fans, but that LaCie is begging for a bullet in the vent one of these nights.

Anyway, those Pacific Pro Audio look decent. I'd be concerned that the company might quit making them or go out of business, but who ever calls a hard drive company for tech support?

Shooshie

[dgrussell]

I have 2 LaCie Triple Interface D2 drives. They are VERY quiet. Thus far, excellent and no problems. One is hooked to my Mac and one is my main library drive for my Giga. I will say, though, that LaCie has caused me huge problems in the past. There was a little utility on an old LaCie SCSI deive. Running it caused the directory on my internal hard drive to be destroyed. The only way to get the drive back was to reformat thus losing quite a bit of data. Now if I see any LaCie utility I trash it or at least never run it. I recommend the same.

[mikebeckmotu]

I make my own external firewire drives - I like Seagate drives, and I buy the IDE drives and put them in a 3rd party Firewire case (a five-minute task). I started doing it just to store and transport graphics files prior to permanently archiving them, but I once used one for audio on my Powerbook, and I've been using it ever since. An inexpensive way to go, but I am not in a pro studio recording for clients, which might change the way I do these things...

[Tim]

Originally posted by Bob Weston:
I've had good luck with the Pacific Pro Audio FW400 drives. They're very quiet (no fan, convection cooling):

http://www.pacificproaudio.com/drives.asp

Those look like the Mercury Elite cases OWC sells, which I've used (with WD and Seagate drives) for years with no problems.
I am curious about their "uncommonly fast version of the Oxford 911 chipset" though.

[Aramis]

Hey!
I think it would be better to precise the model # of a piece of equipment when we criticise those ....For example , I own a FW800 Lacie Big disk extreme and it does not make any noise ....I do record on it all the time .....It is one of the best thing I have ever bought .....And I am certain that a lot of persons can say the same for some other brand too ....
This disk
Aramis