Topic: M-Audio MicroTracker details **PART III**

[Nick's Picks]

Or go with a microdrive 6gb drive. 3hrs at 24/96. I just recieved one from an eBay auction for 100$. It's an Apple branded Hitachi 6gb drive sealed from OEM.

but this is a microdrive, and not CF, correct?
and who knows it if will actualy record relaibly when under the stress of 24/96 I/O
the larger toshiba PCMCIA 5gb drives do not.

[rdvdijk]

but this is a microdrive, and not CF, correct?
and who knows it if will actualy record relaibly when under the stress of 24/96 I/O

Stress? 24/96 is 576kb/s .. That shouldn't be a problem for a microdrive? That's hardly "4x speed". But I agree with you, we still have no "real" performance test results on this device.

Roel

[1st set only]

if it doesnt work w/ the microtracker, a usb 2.0 card reader would do the trick

explain?

[John Kelly]

if it doesnt work w/ the microtracker, a usb 2.0 card reader would do the trick

explain?

You connect a USB Card Reader to the OTG drive instead of the MicroTrack.

but this is a microdrive, and not CF, correct?
and who knows it if will actualy record relaibly when under the stress of 24/96 I/O

Stress? 24/96 is 576kb/s .. That shouldn't be a problem for a microdrive? That's hardly "4x speed". But I agree with you, we still have no "real" performance test results on this device.

Roel

I don't think it's a question of whether or not it can keep up with the I/O, I think it's more of a question of how much tolerance the microdrives have for heat and vibration.  Those are things you don't have to worry about with the CF Cards.

[1st set only]

ahhh that makes sense

thanks

[ford prefect]

Even at the oldest standard, USB 1.0 can transfer data at a rate of 1.5 MBit/sec (which is 187.5 MB/sec). So, at this rate the transfer of 4 GB of data should take approximately 21.3 seconds.

Your math is wrong - 1.5Megabit/sec = 187.5 kilobytes per second.

4GB = 4,000,000,000 bytes.  (roughly)

4,000,000,000 bytes / 187,500 bytes per sec = 21,333 seconds = 355 minutes = 5.9 hours (that sounds about right for my JB3 at this speed)

For USB2:
480Mbit/sec = 60MBytes/sec

4,000,000,000 bytes / 60,000,000 bytes per sec = 66.6 seconds = 1.11 minutes.  (but most IDE hard drives can't write that fast, so it'll be quite a bit slower)

[Sebastian]

Your math is wrong - 1.5Megabit/sec = 187.5 kilobytes per second.

You're right. My mistake was that I thought 1.5 MBit = 1,500 Bit, but in fact it's 1,500,000 Bit (I was just off by a factor of a thousand *g*). I guess that's what you get when trying to solve a simple maths problem in a hurry

(but most IDE hard drives can't write that fast, so it'll be quite a bit slower)

The transfer speed will not be limited by the destination hard drive - the read speed of the CF card will be the bottleneck here! The few cheap ones I've found on eBay can be read at a speed of 4.7 MB/sec (= 0.0047 GB/sec). So, a 4 GB transfer will take 4/0.0047 = 851.06 seconds (about 14 minutes). With a 10 MB/sec card, the same transfer will need 400 seconds (6.66 minutes).

[ford prefect]

The transfer speed will not be limited by the destination hard drive - the read speed of the CF card will be the bottleneck here! The few cheap ones I've found on eBay can be read at a speed of 4.7 MB/sec (= 0.0047 GB/sec).

Until I read this thread I had always thought CF was like RAM - too fast to be a bottleneck.  Thanks for pointing that out.

[neutrino]

FWIW, I just got off the phone with my local salesman of M-Audio products who spoke with his rep from M-Audio today and was told the MicroTrack is expected to ship to retailers on Tuesday the 6th. We'll see...
dB-

[chitaper]

Your math is wrong - 1.5Megabit/sec = 187.5 kilobytes per second.

You're right. My mistake was that I thought 1.5 MBit = 1,500 Bit, but in fact it's 1,500,000 Bit (I was just off by a factor of a thousand *g*). I guess that's what you get when trying to solve a simple maths problem in a hurry

(but most IDE hard drives can't write that fast, so it'll be quite a bit slower)

The transfer speed will not be limited by the destination hard drive - the read speed of the CF card will be the bottleneck here! The few cheap ones I've found on eBay can be read at a speed of 4.7 MB/sec (= 0.0047 GB/sec). So, a 4 GB transfer will take 4/0.0047 = 851.06 seconds (about 14 minutes). With a 10 MB/sec card, the same transfer will need 400 seconds (6.66 minutes).

Ah, thanks! Now I have a formula to use. I have now opted for a slightly more expensive CF card, a SanDisk which should read at 7.9 MB/sec, so even a full card should take less than 9 min. to transfer to the OTG drive. This is less time than it takes between sets/shows, so I should be ok just having 1 CF card at festivals. And most times there will be far less than the full 4GB on the card, so it will take even less time than that. BTW, it seems that SanDisk CF cards read faster than most other brands. But you can save money (read/write speed seems to be the main difference between cheap and expensive cards) if read speed isn't that important to you, and any card is more than adequate for audio as far as write speeds go.

Now if I can just get my MicroTrack to arrive before the NMAS show on Sep. 16! 

[jbraveman]

difference between cheap and expensive cards) if read speed isn't that important to you, and any card is more than adequate for audio as far as write speeds go.

Now if I can just get my MicroTrack to arrive before the NMAS show on Sep. 16! 

Is it true that the write speeds won't matter for this device?  For the PDAudio the requirement was that the card could write at least 576 kb/second for 24/96.  Some of the microdrives apparently couldn't do 24/96 very well.  Since nobody has the device for testing, is it premature to go out and buy some cards?

[John Kelly]

Theoretically all you need is a 4x CF card to write 24/96 (and I'm not sure if you can even get CF cards that slow anymore). 

We'll have to wait for the MicroTrack to be released to see what it *really* needs.

[chitaper]

Is it true that the write speeds won't matter for this device?  For the PDAudio the requirement was that the card could write at least 576 kb/second for 24/96.  Some of the microdrives apparently couldn't do 24/96 very well.  Since nobody has the device for testing, is it premature to go out and buy some cards?

Even the cheapest CF cards can write several times the necessary speed. This one from Costco has a sustained write speed of 3.4 MB/sec.

[rdvdijk]

Even the cheapest CF cards can write several times the necessary speed. This one from Costco has a sustained write speed of 3.4 MB/sec.

I would like to point out again that these speeds are often very dependent on algorithms in the client device. We'll have to wait and see if this 'sustained speed' is indeed feasible in the MicroTrack. As can be read on the PDAudio discussion list (yahoogroups), the programmers have CF-writing problems on PDAs, which do not implement those algorithms. They had some real troubles with some 'high speed' cards. (for a CS-reference, see here)

This being M-Audio, they probably tested the hell out of the MicroTrack, so I'm sure this all won't be a problem.

Roel

[Nick's Picks]

but this is a microdrive, and not CF, correct?
and who knows it if will actualy record relaibly when under the stress of 24/96 I/O

Stress? 24/96 is 576kb/s .. That shouldn't be a problem for a microdrive? That's hardly "4x speed". But I agree with you, we still have no "real" performance test results on this device.

Roel

I've got lots of real experience with the Toshiba 5gb PCMCIA drive..which is larger..and may be more robust than these which are 1/3rd the size.
and these toshiba drives suck.