Successfully using a 4200 rpm HD in notebook for multitrack after doing all
the tweaks, partitioning etc (at least at 44/16!). Have I gone overboard in
making the main data partition 64K cluster size (NTFS)? I've seen elsewhere
32K is used eg in Mackie's units, but I doubtless these have faster spinning
drives. With careful management I have enough space, but if the improvement
is likely to be minimal @ 64 is 32 a better compromise?

"Karl Engel" karlengel-at-excite-dot-com wrote in message

Successfully using a 4200 rpm HD in notebook for multitrack after
doing all the tweaks, partitioning etc (at least at 44/16!). Have I
gone overboard in making the main data partition 64K cluster size
(NTFS)? I've seen elsewhere 32K is used eg in Mackie's units, but I
doubtless these have faster spinning drives. With careful management
I have enough space, but if the improvement is likely to be minimal @
64 is 32 a better compromise?

The major downsides of large allocation units are wasted disk space and
wasted I/O for files that are a lot smaller than an allocation unit. Given
that almost all audio files, even MP3 files run multiple megabytes, that
isn't too much of a threat.

Too-small allocation units lead to problems with fragmentation and excess
CPU use. Modern CPUs are fast enough that its easy to get into diminishing
returns as far as CPU bottlenecks go.

I wouldn't waste a lot of time worrying about this issue given the size of
allocation units you are using.

On Mon, 8 Nov 2004 14:50:34 +1100, "Karl Engel"
karlengel-at-excite-dot-com wrote:

Successfully using a 4200 rpm HD in notebook for multitrack after doing all
the tweaks, partitioning etc (at least at 44/16!). Have I gone overboard in
making the main data partition 64K cluster size (NTFS)? I've seen elsewhere
32K is used eg in Mackie's units, but I doubtless these have faster spinning
drives. With careful management I have enough space, but if the improvement
is likely to be minimal @ 64 is 32 a better compromise?

When we were all trying to run DAWs on under-powered hardware a few
years ago there was much discussion over cluster size. It never
actually seemed to make much difference.

CubaseFAQ www.laurencepayne.co.uk/CubaseFAQ.htm
"Possibly the world's least impressive web site": George Perfect

Laurence Payne wrote:

When we were all trying to run DAWs on under-powered hardware a few
years ago there was much discussion over cluster size. It never
actually seemed to make much difference.

It still matters on handheld PDAs.

CPU clocks on Pocket PCs (e.g., HP iPAQs, Dell Axims) just passed 600
MHz. The lions share of PDAs that are already out in the field are
clocking at 206 and 400 MHz. Their CPUs don't have floating point units
or DMA for mass storage transfers, so the larger the cluster size the
more efficient the CPU's transfer of 24/96 audio data.

So we recommend formatting the flash storage and microdrives to 64 KB
clusters.

--
Len Moskowitz PDAudio, Binaural Mics, Cables, DPA, M-Audio
Core Sound http://www.stealthmicrophones.com
Teaneck, New Jersey USA http://www.core-sound.com
Tel: 201-801-0812, FAX: 201-801-0912

Len Moskowitz wrote:


Laurence Payne wrote:

When we were all trying to run DAWs on under-powered hardware a few
years ago there was much discussion over cluster size. It never
actually seemed to make much difference.

It still matters on handheld PDAs.

CPU clocks on Pocket PCs (e.g., HP iPAQs, Dell Axims) just passed 600
MHz. The lions share of PDAs that are already out in the field are
clocking at 206 and 400 MHz. Their CPUs don't have floating point units
or DMA for mass storage transfers, so the larger the cluster size the
more efficient the CPU's transfer of 24/96 audio data.

Are people REALLY trying to run DAWs on PDAs? We're not talking about
dictation or stealth recording of concerts, but real multitrack 44.1/16 or
better (hell, you said 24/96) with editing, DSP & summing??!!??!!??!!

Why ?????????

On 9 Nov 2004 19:21:03 -0500, (Len Moskowitz)
wrote:

When we were all trying to run DAWs on under-powered hardware a few
years ago there was much discussion over cluster size. It never
actually seemed to make much difference.

It still matters on handheld PDAs.

CPU clocks on Pocket PCs (e.g., HP iPAQs, Dell Axims) just passed 600
MHz. The lions share of PDAs that are already out in the field are
clocking at 206 and 400 MHz. Their CPUs don't have floating point units
or DMA for mass storage transfers, so the larger the cluster size the
more efficient the CPU's transfer of 24/96 audio data.

So we recommend formatting the flash storage and microdrives to 64 KB
clusters.

What sort of audio applications are running on Pocket PCs?

CubaseFAQ www.laurencepayne.co.uk/CubaseFAQ.htm
"Possibly the world's least impressive web site": George Perfect

agent86 wrote:

Are people REALLY trying to run DAWs on PDAs? We're not talking about
dictation or stealth recording of concerts, but real multitrack 44.1/16 or
better (hell, you said 24/96) with editing, DSP & summing??!!??!!??!!

As I mentioned, even the most recently released PDAs have limited
processing power and I/O efficiency, so for now we're limited to two
tracks of 24/96 WAV files. I'm certain that PDAs could handle four
tracks of 16/44.1 with no problems (or perhaps even 24/48) if the front
end hardware (and drivers) were available.

For PDAudio there are three recording applications. All of them can
function as players and one of them can be used as an editor too.

In addition there are at least two non-PDAudio editing applications, and
many fine player applications that rival any of the dedicated MP3
players in ease-of-use.

DSP-based effect and summing will have to wait until PDAs have Floating
Point Processors, and more than four tracks of 16/44.1 will have to wait
for I/O DMA.

--
Len Moskowitz PDAudio, Binaural Mics, Cables, DPA, M-Audio
Core Sound http://www.stealthmicrophones.com
Teaneck, New Jersey USA http://www.core-sound.com
Tel: 201-801-0812, FAX: 201-801-0912

Laurence Payne wrote:

What sort of audio applications are running on Pocket PCs?

There are recorders (up to two tracks of 24/96 WAV, MP3, Ogg Vorbis and
other formats), players and rudimentary editors. There are also a
variety of musical instrument applications not necessarily related to
recording.

For more information about PDAudio, please see our Web site.

--
Len Moskowitz PDAudio, Binaural Mics, Cables, DPA, M-Audio
Core Sound http://www.stealthmicrophones.com
Teaneck, New Jersey USA http://www.core-sound.com
Tel: 201-801-0812, FAX: 201-801-0912

On Wed, 10 Nov 2004 16:46:18 -0500, agent86
wrote:

Well, whodathunkit?

My first impressions upon being informed of that information (not trying to
give you a hard time).

1. If you're stuck with 2 tracks & no DSP, anything over 44.1/16 is
probably overkill (although I know some people will argue this point).

2. You'd have to be a real glutton for punishment to try & edit audio on
such a tiny little screen. That would have me WISHING for a razor blade.

Then.....
Audio history is littered with useful things achieved by equipment
that was No Way Capable. Like 4-track on cassette. Like any audio
at all on cassette, come to that :-)

CubaseFAQ www.laurencepayne.co.uk/CubaseFAQ.htm
"Possibly the world's least impressive web site": George Perfect

Len Moskowitz wrote:

agent86 wrote:

Are people REALLY trying to run DAWs on PDAs? We're not talking about
dictation or stealth recording of concerts, but real multitrack 44.1/16 or
better (hell, you said 24/96) with editing, DSP & summing??!!??!!??!!

As I mentioned, even the most recently released PDAs have limited
processing power and I/O efficiency, so for now we're limited to two
tracks of 24/96 WAV files. I'm certain that PDAs could handle four
tracks of 16/44.1 with no problems (or perhaps even 24/48) if the front
end hardware (and drivers) were available.

For PDAudio there are three recording applications. All of them can
function as players and one of them can be used as an editor too.

In addition there are at least two non-PDAudio editing applications, and
many fine player applications that rival any of the dedicated MP3
players in ease-of-use.

DSP-based effect and summing will have to wait until PDAs have Floating
Point Processors, and more than four tracks of 16/44.1 will have to wait
for I/O DMA.


Well, whodathunkit?

My first impressions upon being informed of that information (not trying to
give you a hard time).

1. If you're stuck with 2 tracks & no DSP, anything over 44.1/16 is
probably overkill (although I know some people will argue this point).

2. You'd have to be a real glutton for punishment to try & edit audio on
such a tiny little screen. That would have me WISHING for a razor blade.

YMMV

Laurence Payne wrote:

On Wed, 10 Nov 2004 16:46:18 -0500, agent86
wrote:

Well, whodathunkit?

My first impressions upon being informed of that information (not trying
to give you a hard time).

1. If you're stuck with 2 tracks & no DSP, anything over 44.1/16 is
probably overkill (although I know some people will argue this point).

2. You'd have to be a real glutton for punishment to try & edit audio on
such a tiny little screen. That would have me WISHING for a razor blade.

Then.....
Audio history is littered with useful things achieved by equipment
that was No Way Capable. Like 4-track on cassette. Like any audio
at all on cassette, come to that :-)

My point was not that the equipment is not capable. Only that it seemed
like doing things the hard way on purpose.

Laurence Payne wrote:

On Wed, 10 Nov 2004 16:46:18 -0500, agent86
wrote:

Well, whodathunkit?

My first impressions upon being informed of that information (not trying
to give you a hard time).

1. If you're stuck with 2 tracks & no DSP, anything over 44.1/16 is
probably overkill (although I know some people will argue this point).

2. You'd have to be a real glutton for punishment to try & edit audio on
such a tiny little screen. That would have me WISHING for a razor blade.

Then.....
Audio history is littered with useful things achieved by equipment
that was No Way Capable. Like 4-track on cassette. Like any audio
at all on cassette, come to that :-)

My point was not that the equipment is not capable. Only that it seemed
like doing things the hard way on purpose.

On 2004-11-10, agent86 wrote:

Well, whodathunkit?

Jam Band tapers, of course.

Two channels, high quality, solid state, fits in your shirt pocket?
A New Jersey deadhead audiophile's dream. Or Len. :-P

On 2004-11-10, agent86 wrote:

Well, whodathunkit?

Jam Band tapers, of course.

Two channels, high quality, solid state, fits in your shirt pocket?
A New Jersey deadhead audiophile's dream. Or Len. :-P

james of tucson wrote:

Well, whodathunkit?

Jam Band tapers, of course.

Two channels, high quality, solid state, fits in your shirt pocket?
A New Jersey deadhead audiophile's dream. ...

Who else are using PDAudio? What else are they recording?

Ambient tapers, location sound recordists, dialog boom mic recording
without cables, DV and mini-DV videographers who need studio quality
sound, documentarians, oral historians, ENG (electronic news gathering),
sound effects libraries, audience tapers,

Who needs it?

Anyone who wants to record at 16/44.1 to 24/96 recording quality, who is
tired of lugging around laptop computers and recorders as big as the
Manhattan phone book, and who wants to spend under $1000 (or under $500
for the digital only PDAudio).

...Or Len. :-P

Yup!


--
Len Moskowitz PDAudio, Binaural Mics, Cables, DPA, M-Audio
Core Sound http://www.stealthmicrophones.com
Teaneck, New Jersey USA http://www.core-sound.com
Tel: 201-801-0812, FAX: 201-801-0912

james of tucson wrote:

Well, whodathunkit?

Jam Band tapers, of course.

Two channels, high quality, solid state, fits in your shirt pocket?
A New Jersey deadhead audiophile's dream. ...

Who else are using PDAudio? What else are they recording?

Ambient tapers, location sound recordists, dialog boom mic recording
without cables, DV and mini-DV videographers who need studio quality
sound, documentarians, oral historians, ENG (electronic news gathering),
sound effects libraries, audience tapers,

Who needs it?

Anyone who wants to record at 16/44.1 to 24/96 recording quality, who is
tired of lugging around laptop computers and recorders as big as the
Manhattan phone book, and who wants to spend under $1000 (or under $500
for the digital only PDAudio).

...Or Len. :-P

Yup!


--
Len Moskowitz PDAudio, Binaural Mics, Cables, DPA, M-Audio
Core Sound http://www.stealthmicrophones.com
Teaneck, New Jersey USA http://www.core-sound.com
Tel: 201-801-0812, FAX: 201-801-0912

Karl Engel wrote:

Successfully using a 4200 rpm HD in notebook for multitrack
after doing all the tweaks, partitioning etc (at least at 44/16!).
Have I gone overboard in making the main data partition
64K cluster size (NTFS)?

Yes. I have yet to see a real speed advantage in using anything but
standard 4 kbyte blocks, and the system supplied defrag software
requires them. One of the advantages of using 4 kbyte blocks is that the
windows memory mapping also uses 4 kbyte blocks.

I've seen elsewhere 32K is used eg in Mackie's units,

Probably because they use FAT32. Which btw. may have a slight advantage
compared to ntfs for linear files.

but I doubtless these have faster spinning
drives. With careful management I have enough space, but if
the improvement is likely to be minimal @ 64 is 32 a better
compromise?

I don't think you have to worry about this magnitude of marginal speed
and overhead considerations, and if you need to go that far in
optimizing then you are way out of of real world "give" to allow for
less optimum conditions, such as a fragmented drive.

My understanding is that the 64 kbyte NTFS blocks are intestering for
database files on raid 5 ntfs drive sets and similar because the default
raid block size happens to be 64 kbyte.

Disk rotation speeds is not really relevant, what matters is the number
of magnetic domains - or physical disk blocks - pr. revolution, a disk
with a slower rotational speed but denser magnetic recording may be
faster than a disk with fast rotation and less dense recording.


Kind regards

Peter Larsen


--
*******************************************
* My site is at: http://www.muyiovatki.dk *
*******************************************

Karl Engel wrote:

Successfully using a 4200 rpm HD in notebook for multitrack
after doing all the tweaks, partitioning etc (at least at 44/16!).
Have I gone overboard in making the main data partition
64K cluster size (NTFS)?

Yes. I have yet to see a real speed advantage in using anything but
standard 4 kbyte blocks, and the system supplied defrag software
requires them. One of the advantages of using 4 kbyte blocks is that the
windows memory mapping also uses 4 kbyte blocks.

I've seen elsewhere 32K is used eg in Mackie's units,

Probably because they use FAT32. Which btw. may have a slight advantage
compared to ntfs for linear files.

but I doubtless these have faster spinning
drives. With careful management I have enough space, but if
the improvement is likely to be minimal @ 64 is 32 a better
compromise?

I don't think you have to worry about this magnitude of marginal speed
and overhead considerations, and if you need to go that far in
optimizing then you are way out of of real world "give" to allow for
less optimum conditions, such as a fragmented drive.

My understanding is that the 64 kbyte NTFS blocks are intestering for
database files on raid 5 ntfs drive sets and similar because the default
raid block size happens to be 64 kbyte.

Disk rotation speeds is not really relevant, what matters is the number
of magnetic domains - or physical disk blocks - pr. revolution, a disk
with a slower rotational speed but denser magnetic recording may be
faster than a disk with fast rotation and less dense recording.


Kind regards

Peter Larsen


--
*******************************************
* My site is at: http://www.muyiovatki.dk *
*******************************************