Pops & Clicks - Hard Drive Too Slow??

ChrisBo wrote on 11/4/2000, 1:33 AM
Howdy!

I've been using Vegas since the first Beta came out, and
have recently worked my system up to be able to
theoretically record 24 tracks at at time... unfortunately,
I've been having some trouble with it - specifically getting
a substantial number of pops and clicks.

The project I've been working on is choral concert. The
concert was a little over two hours long... an hour per
half. I had about 16 tracks coming in: 8 through a Fronier
Designs Tango and another 8 coming in through a Tascam
MA-AD8 (chained to a IF-TAD) - both through the light pipe
inputs on a Frontier Designs Dakota. The word-sync was
provided by the internal sync of the Tango.

Anyway.. the first night, I solved most of the clicks
between the first and second night by ditching the routing I
had set up on the Dakota to monitor the input tracks... I
assume that the pathing was taking too much from the
processor?

So... the second night. The first half was great -
virtually no pops... the second half, however, the problem
came back. Any guesses what might have caused this?
I'm thinking that the drive was working close to it's limit
and as it got closer to being full, the performance
diminished? Will installing an IDE Raid help?

One additional annoyance: After I finished recording, it
took an EXTREMELY long time to give me control back of my
machine.

Does anyone else out there have any similar problems?? Or
exprience with extended record times? (45 - 60 minutes)

The Beast:
Celeron 333
256MB RAM
4 Gig "System" drive (C:)
Maxtor 40Gig 7200RPM DMA/100 Drive (Data Only)
Win 98
Frontier Designs Dakota

Comments

MJim wrote on 11/5/2000, 3:07 PM
Sounds to me like you're asking a lot out of that system. Consider
that each of your 16 mono inputs creates one wav file. Each wav file
is stereo (just a fact about wav files). So this means you're
actually recording the equivalent of 32 tracks simultaneously. 16
tracks @ aprox 10mb per track, per min, adds up to a lot of gigs for
an hour long continuous file. Windows may not be able to handle this.

The best advice for recording 16 tracks live is to go digital tape,
either Tascam or ADAT then dump it into the computer for editing.
The new dedicated hard drive recorders, Tascam mx2424 and Mackie
hdr2496, are also built for this application. If you must use a
computer, I think a lot more muscle would help. Like genuine PIII
chip on an ASUS or Intel mobo, 133 fsb, 800mhz or more clock speed,
or dual processors. Personally, I prefer to use scsi hard drives at
10k, although many claim the 7200 drives will indeed cut the pace.
You may not be getting dma 100 if your other drive is lower, due to a
default factor. Simply put, both drives have to be up to the task.
Other issues arise with conflicts and background apps running in
hidden places. The reason it takes a long time to recover at the end
is because your system is crashing. Little bit surprising that it
made it that long and was able to regain control. After all that I
still do not feel that computers are capable of 100% reliability on
this kind of live recording. I'll stay with the tape for now.
Frontier products are good, so no problems there. The 256mb is nice,
just enough.

A more traditional approach would be to mix live to DAT tape. But it
sounds like you're really wanting to multitrack.

Just my thoughts,
~ Jim

Christopher Boyd wrote:
>>Howdy!
>>
>>I've been using Vegas since the first Beta came out, and
>>have recently worked my system up to be able to
>>theoretically record 24 tracks at at time... unfortunately,
>>I've been having some trouble with it - specifically getting
>>a substantial number of pops and clicks.
>>
>>The project I've been working on is choral concert. The
>>concert was a little over two hours long... an hour per
>>half. I had about 16 tracks coming in: 8 through a Fronier
>>Designs Tango and another 8 coming in through a Tascam
>>MA-AD8 (chained to a IF-TAD) - both through the light pipe
>>inputs on a Frontier Designs Dakota. The word-sync was
>>provided by the internal sync of the Tango.
>>
>>Anyway.. the first night, I solved most of the clicks
>>between the first and second night by ditching the routing I
>>had set up on the Dakota to monitor the input tracks... I
>>assume that the pathing was taking too much from the
>>processor?
>>
>>So... the second night. The first half was great -
>>virtually no pops... the second half, however, the problem
>>came back. Any guesses what might have caused this?
>>I'm thinking that the drive was working close to it's limit
>>and as it got closer to being full, the performance
>>diminished? Will installing an IDE Raid help?
>>
>>One additional annoyance: After I finished recording, it
>>took an EXTREMELY long time to give me control back of my
>>machine.
>>
>>Does anyone else out there have any similar problems?? Or
>>exprience with extended record times? (45 - 60 minutes)
>>
>>The Beast:
>>Celeron 333
>>256MB RAM
>>4 Gig "System" drive (C:)
>>Maxtor 40Gig 7200RPM DMA/100 Drive (Data Only)
>>Win 98
>>Frontier Designs Dakota
ChrisBo wrote on 11/5/2000, 9:36 PM


jimm wrote:
>>Sounds to me like you're asking a lot out of that system. Consider
>>that each of your 16 mono inputs creates one wav file. Each wav
file
>>is stereo (just a fact about wav files).

The wav standad allows for an number of tracks interleved with a wide
variety of bit-rates and sampling freqency. Mono files are not a
stereo file - unless the software doesn't know any better (Vegas does
know better) I'm creating 6 stereo files and 1 mono file for one
specific aplication. All told, 44.1k, 24 bit means a little better
than 2MB/min for 16 tracks... nowhere near even DMA/33
As near as I can figure, 7200 RPM drives can sustain around 5 - 7
MB/sec, depending on block size.

>>SNIP.....
>>The best advice for recording 16 tracks live is to go digital tape,
>>either Tascam or ADAT then dump it into the computer for editing.

Again, recording to a DA-88 doesn't solve the problem, because you
still have to dump the tape to the computer. In fact, that specific
application is probably more in line with why my specific problem

>>The new dedicated hard drive recorders, Tascam mx2424 and Mackie
>>hdr2496, are also built for this application. If you must use a
>>computer, I think a lot more muscle would help. Like genuine PIII
>>chip on an ASUS or Intel mobo, 133 fsb, 800mhz or more clock speed,
>>or dual processors. Personally, I prefer to use scsi hard drives at
>>10k, although many claim the 7200 drives will indeed cut the pace.
>>You may not be getting dma 100 if your other drive is lower, due to
>>default factor. Simply put, both drives have to be up to the task.
I have an ASUS MB with genuine Celeron (I'll be upgrading eventually)
Yes, I'm hitting the processor hard, but according to the system
monitor, I'm not quite maxing it out.

>>Other issues arise with conflicts and background apps running in
>>hidden places.
Of course there are things that are conflicts in the background....
but which ones??

>>is because your system is crashing. Little bit surprising that it
>>made it that long and was able to regain control. After all that I
>>still do not feel that computers are capable of 100% reliability on
>>this kind of live recording. I'll stay with the tape for now.
>>Frontier products are good, so no problems there. The 256mb is
nice,
>>just enough.
>>
>>A more traditional approach would be to mix live to DAT tape.

I do mix live to DAT as well, but the ability mulitrack allows a lot
more flexibility (duh)


So.. the bottom line.. who's fault is it? The hard disk drivers?
Vegas? the Dakota Drivers? Video? Why does it take 30-45 minutes to
start??

BTW.. since my first post, I added a Promise IDE raid with 2 identical
Maxtor 7200 RPM 20 gig drives - same problem.

Chris
SamHardy wrote on 11/5/2000, 10:40 PM
A faster processor would probably help you the most; although I'm a
real believer in fast drives; when i switched from DMA 33 to UW SCSI
on Ultra160 drives, the throughput increased dramatically. The SCSI
card was only $160 (Tekram DC-390U2W) and the drive about $300 (IBM
Ultra160 18.6 GB), so not too bad.

I'm very impressed you've gotten as much stability out of your PC as
you have, but you could really use a jump in CPU power. As jimm said,
a faster fsb would be nice too; although at that point you're making
a significant investment: new mobo (probably), new RAM (PC133 is
expensive) and a new processor.

It might also be this Win98SE thing; VV apparently really wants that
OS (according to some replies on this board); I have a post up here
about audio dropouts and I'm dreading rebuilding my system. But for
you, it would be certainly the cheapest way to fix your problem if
that's the issue.

Best,
Sam Hardy

Christopher Boyd wrote:
>>
>>
>>jimm wrote:
>>>>Sounds to me like you're asking a lot out of that system.
Consider
>>>>that each of your 16 mono inputs creates one wav file. Each wav
>>file
>>>>is stereo (just a fact about wav files).
>>
>>The wav standad allows for an number of tracks interleved with a
wide
>>variety of bit-rates and sampling freqency. Mono files are not a
>>stereo file - unless the software doesn't know any better (Vegas
does
>>know better) I'm creating 6 stereo files and 1 mono file for one
>>specific aplication. All told, 44.1k, 24 bit means a little better
>>than 2MB/min for 16 tracks... nowhere near even DMA/33
>>As near as I can figure, 7200 RPM drives can sustain around 5 - 7
>>MB/sec, depending on block size.
>>
>>>>SNIP.....
>>>>The best advice for recording 16 tracks live is to go digital
tape,
>>>>either Tascam or ADAT then dump it into the computer for
editing.
>>
>>Again, recording to a DA-88 doesn't solve the problem, because you
>>still have to dump the tape to the computer. In fact, that
specific
>>application is probably more in line with why my specific problem
>>
>>>>The new dedicated hard drive recorders, Tascam mx2424 and Mackie
>>>>hdr2496, are also built for this application. If you must use a
>>>>computer, I think a lot more muscle would help. Like genuine
PIII
>>>>chip on an ASUS or Intel mobo, 133 fsb, 800mhz or more clock
speed,
>>>>or dual processors. Personally, I prefer to use scsi hard drives
at
>>>>10k, although many claim the 7200 drives will indeed cut the
pace.
>>>>You may not be getting dma 100 if your other drive is lower, due
to
>> >>default factor. Simply put, both drives have to be up to the
task.
>>I have an ASUS MB with genuine Celeron (I'll be upgrading
eventually)
>>Yes, I'm hitting the processor hard, but according to the system
>>monitor, I'm not quite maxing it out.
>>
>>>>Other issues arise with conflicts and background apps running in
>>>>hidden places.
>>Of course there are things that are conflicts in the background....
>>but which ones??
>>
>>>>is because your system is crashing. Little bit surprising that
it
>>>>made it that long and was able to regain control. After all that
I
>>>>still do not feel that computers are capable of 100% reliability
on
>>>>this kind of live recording. I'll stay with the tape for now.
>>>>Frontier products are good, so no problems there. The 256mb is
>>nice,
>>>>just enough.
>>>>
>>>>A more traditional approach would be to mix live to DAT tape.
>>
>>I do mix live to DAT as well, but the ability mulitrack allows a
lot
>>more flexibility (duh)
>>
>>
>>So.. the bottom line.. who's fault is it? The hard disk drivers?
>>Vegas? the Dakota Drivers? Video? Why does it take 30-45 minutes to
>>start??
>>
>>BTW.. since my first post, I added a Promise IDE raid with 2
identical
>>Maxtor 7200 RPM 20 gig drives - same problem.
>>
>>Chris
PipelineAudio wrote on 11/6/2000, 2:50 PM


Christopher Boyd wrote:
>>Howdy!
>>
>>I've been using Vegas since the first Beta came out, and
>>have recently worked my system up to be able to
>>theoretically record 24 tracks at at time... unfortunately,
>>I've been having some trouble with it - specifically getting
>>a substantial number of pops and clicks.


You must have balls the size of asteroids to be carrying a PC around
and using Vegas as a tape recorder...I commend you man
PipelineAudio wrote on 11/6/2000, 2:56 PM


>>Again, recording to a DA-88 doesn't solve the problem, because you
>>still have to dump the tape to the computer. In fact, that
specific
>>application is probably more in line with why my specific problem
>>


yes but you can dump it into the computer in smaller incrememnts...i
dump my DA-x8's into the computer 16 tracks at a time, with vegas
chasing MTC, then do the rest of the tracks in subsequent passes

I am happy when this sucker can record 16 tracks at a time...never
mind 24 :)

but of course, we all know Nuendo can record 999 tracks at a time
right? :)
Rednroll wrote on 11/6/2000, 10:43 PM
Yeah...hahah...basically you guys are right. You're asking too much
out of that system. I'm really surprised that you can even get 16
Tracks simultaneously recording. Just as an extra note, make sure
you uncheck "simultaneous record and playback", this will greatly
improve system overhead, and also uncheck "playback video tracks". I
tested my system out and I'm running an Athlon 700Mhz with a Cheetah
10K rpm SCSI drive and I was only able to record 30 Tracks
simultaneously before getting stuttering. So if you're attempting 24
Tracks on a 333Mhz Celeron with an 7200 rpm IDE drive, you should be
thankful that it's not crashing. Other things you can do to reduce
overhead on your system is to set it in your preferences to not draw
waveforms while recording. Also make sure that hard drive is
Defragmented before attempting to record like that again.

Aaron Carey wrote:
>>
>>
>>>>Again, recording to a DA-88 doesn't solve the problem, because
you
>>>>still have to dump the tape to the computer. In fact, that
>>specific
>>>>application is probably more in line with why my specific problem
>>>>
>>
>>
>>yes but you can dump it into the computer in smaller
incrememnts...i
>>dump my DA-x8's into the computer 16 tracks at a time, with vegas
>>chasing MTC, then do the rest of the tracks in subsequent passes
>>
>>I am happy when this sucker can record 16 tracks at a time...never
>>mind 24 :)
>>
>>but of course, we all know Nuendo can record 999 tracks at a time
>>right? :)
>>
ChrisBo wrote on 11/7/2000, 12:10 AM


Aaron Carey wrote:
>>
>>
>>Christopher Boyd wrote:
>>>>Howdy!
>>>>
>>>>I've been using Vegas since the first Beta came out, and
>>>>have recently worked my system up to be able to
>>>>theoretically record 24 tracks at at time... unfortunately,
>>>>I've been having some trouble with it - specifically getting
>>>>a substantial number of pops and clicks.
>>
>>
>>You must have balls the size of asteroids to be carrying a PC around
>>and using Vegas as a tape recorder...I commend you man

When you have as much gear as I do to lug in anyway.. might as well
drag a PC in to.. I'm thinking that those plasma screens are lookin'
REALLY good tho' :)
ChrisBo wrote on 11/7/2000, 12:18 AM


Brian Franz wrote:
>>Yeah...hahah...basically you guys are right. You're asking too much
>>out of that system. I'm really surprised that you can even get 16
>>Tracks simultaneously recording. Just as an extra note, make sure
>>you uncheck "simultaneous record and playback", this will greatly
>>improve system overhead, and also uncheck "playback video tracks".
I
>>tested my system out and I'm running an Athlon 700Mhz with a Cheetah
>>10K rpm SCSI drive and I was only able to record 30 Tracks
>>simultaneously before getting stuttering. So if you're attempting
24
>>Tracks on a 333Mhz Celeron with an 7200 rpm IDE drive, you should be
>>thankful that it's not crashing. Other things you can do to reduce
>>overhead on your system is to set it in your preferences to not draw
>>waveforms while recording. Also make sure that hard drive is
>>Defragmented before attempting to record like that again.
>>
>>Aaron Carey wrote:

Again, I know the machine needs more processor power, but I want to
make sure that there aren't other things I need first... I'm only
actually doing about 16 tracks.. and most of my stuff is less than 8..
but I figure if I can do 16-24 tracks... then 8 should be safe in the
real world. I'm finding the system monitor very useful.. I'm hitting
the processor hard, but what exactly is hitting it?? the HD
Controller? the Dakota? Vegas? What? and why does it work fine for
half an hour to an hour at a shot, and THEN break down? What kind of
buffers would take that long to fill up? If it's not buffer
related... is it the hard drives that are heating up? or is it just
something about how Vegas writes to the HD. Anyone know what block
size Vegas Uses??

Chris
Rednroll wrote on 11/7/2000, 3:25 PM
I would suspect your hard drive or your video card. It's a lot of
overhead when Vegas has to draw the waveforms and scrolls from screen
to screen. That's why I suggested that you turn draw waveforms off
while recording. I know you said the specs of that IDE drive should
widthstand the prolonged recording, but it's a known fact that IDE
drives just don't have the throughput that a SCSI drive does for
sustained recording operations. My suggestion would be to buy a SCSI
card and a SCSI hard drive with 10K rpm. When I did this I was able
to playback an additional 18 tracks compared to my 7200rpm EIDE
drive. The nice thing about doing this, is that when you finally buy
a faster computer, you can transfer this hardware to your new
system. For a cheaper alternative I suggest a Tekram DC-390F Scsi
card over an Adaptec. You can get one for about $75 and it will give
you 40MB/Sec transfer speeds. If you have the cash, then go for the
new Adaptec cards, but I imagine that your SCSI bus will then run
faster than your CPU.

Christopher Boyd wrote:
>>
>>
>>Brian Franz wrote:
>>>>Yeah...hahah...basically you guys are right. You're asking too
much
>>>>out of that system. I'm really surprised that you can even get
16
>>>>Tracks simultaneously recording. Just as an extra note, make
sure
>>>>you uncheck "simultaneous record and playback", this will greatly
>>>>improve system overhead, and also uncheck "playback video
tracks".
>>I
>>>>tested my system out and I'm running an Athlon 700Mhz with a
Cheetah
>>>>10K rpm SCSI drive and I was only able to record 30 Tracks
>>>>simultaneously before getting stuttering. So if you're
attempting
>>24
>>>>Tracks on a 333Mhz Celeron with an 7200 rpm IDE drive, you should
be
>>>>thankful that it's not crashing. Other things you can do to
reduce
>>>>overhead on your system is to set it in your preferences to not
draw
>>>>waveforms while recording. Also make sure that hard drive is
>>>>Defragmented before attempting to record like that again.
>>>>
>>>>Aaron Carey wrote:
>>
>>Again, I know the machine needs more processor power, but I want to
>>make sure that there aren't other things I need first... I'm only
>>actually doing about 16 tracks.. and most of my stuff is less than
8..
>>but I figure if I can do 16-24 tracks... then 8 should be safe in
the
>>real world. I'm finding the system monitor very useful.. I'm
hitting
>>the processor hard, but what exactly is hitting it?? the HD
>>Controller? the Dakota? Vegas? What? and why does it work fine for
>>half an hour to an hour at a shot, and THEN break down? What kind
of
>>buffers would take that long to fill up? If it's not buffer
>>related... is it the hard drives that are heating up? or is it just
>>something about how Vegas writes to the HD. Anyone know what block
>>size Vegas Uses??
>>
>>Chris