AudioMasters
 
  User Info & Key Stats   
Welcome, Guest. Please login or register.

Login with username, password and session length
June 21, 2015, 12:04:05 AM
75909 Posts in 8048 Topics by 2824 Members
Latest Member: toiletrolltube
News:       Buy Adobe Audition:
+  AudioMasters
|-+  Audio Related
| |-+  Hardware and Soundcards
| | |-+  More laptop buying questions - processor and hard drive
  « previous next »
Pages: [1] 2 3 ... 5 Print
Author
Topic: More laptop buying questions - processor and hard drive  (Read 27234 times)
« on: June 20, 2005, 05:49:42 AM »
blurk Offline
Member
*****
Posts: 408



In another thread we ending up discussing various things to look out for in buying a laptop.  I still have (yet) more questions, partly based on this posting from SteveG somewhat earlier in the thread than when I joined in.  I'll quote Steve's post extensively, simply because he happens to mention the various items I have questions about.

Quote
the internal drive on mine is a toshiba 5400rpm one, but this isn't actually much of a disadvantage, because the drive is physically smaller and the seek times aren't significantly compromised.

Some of the ThinkPad models I looked at (particular the ones with larger capacity drives) only had 4200 rpm drives.  Is there an easy way to determine a minimum rotation speed required for drives?  I expect the issue becomes significant for multitrack recording, but I'll only ever need to do at most a stereo recording, so I suspect a 7200 rpm model is not strictly required.  Or is it?

Quote
I have a small and very cheap 50GB USB2 external drive

Now I'm curious that SteveG confirmed my bias against USB audio, but it seems that USB is fine to run a hard drive.  So I'm guessing this means that USB2 has sufficient bandwidth, but latency is a problem for audio (though not recording audio - hmmm).  I'd just like to understand the issues involved here.

Quote
I set it up to record all of the analogue inputs from the Traveler and the Behringer simultaneously (8 pairs = sixteen individual channels) - that's 32-bit 48k recording - and it didn't bat an eyelid. In theory the internal drive should cope with this on its own, and I'll try that later.

So, even a (relatively) slow 5400 rpm drive should be able to record 16 channels at 32-bit 48k, so I'd imagine a measly 4200 rpm drive should be able to cope with a stereo recording.  I just want to know if drive rotation speed is something I should specifically look out for, or whether other things should be higher on my requirements list.

Quote
I quite like the Pavilion, even though it doesn't have an M-series processor

I had initially assumed that any mobile processor would be inferior to a raw P4, but I've since heard elsewhere that the M-series processors can actually outperform P4s.  Can anyone shed light on such claims.  I'm also curious to know whether the Pentium-M processors suffer from the P4 denormalisation problem - because if they don't, that could also sell me on a Pentium-M processor.

BTW, I'm probably looking at ruling out the ThinkPads at the moment because of the ones available in Australia, I'm finding it difficult to find ones that have a Firewire interface and the screen resolution I'm after.  It's also unclear as to whether any of them support dual head use.  I'm still looking at the Toshibas and the HP/Compaqs.  I'm still nervous about the Vaios, and certainly Steve's point about their software delivery is a concern, as likely the first thing I'll do with the thing is repartition and reformat.
Logged
Reply #1
« on: June 20, 2005, 07:39:48 AM »
Euphony Offline
Member
*****
Posts: 371



Quote
I'd imagine a measly 4200 rpm drive should be able to cope with a stereo recording.


I'm sure any computer drive made in the last decade can handle a stereo recording of 48khz/32bit.  Any recent year laptop will not have a problem with this.

Quote
it seems that USB is fine to run a hard drive


As a storage unit, yes.  The maximum usb2 can transfer is 60mb/s (theoretically), which is plenty fast for backing up files.

Quote
I've since heard elsewhere that the M-series processors can actually outperform P4s.


It is true - at Tom's Hardware, they were able to outfit a desktop computer with a Pentium-M, concluding it outperforms P4's on the majority of tests, runs cooler and consumes much less power.

Quote
the first thing I'll do with the thing is repartition and reformat.


When I bought my computer, thats the first thing I did.  I don't like the miscellaneous junk every computer brand loads on their system - ends up wasting space and wasting memory to run apps that I don't want.
Logged
Reply #2
« on: June 20, 2005, 08:46:13 AM »
ozpeter Offline
Member
*****
Posts: 2377



I've now got the budget but I'm still in a quandary about the laptop itself, really entirely on the grounds of noise (not affecting the recording because I'm always too far away, but affecting the audience if I'm working in the auditorium, which tends to be the norm sadly).  I' thinking about a 1.7+ Centrino rather than a P4 on those grounds, but I suspect that Audition at least will be happier on the P4.

I've not found much on the net relating up-to-date experiences with laptop recording, which is odd - maybe I'm not using the right search terms.  Another source of advice is to look at the configuration of laptops customised for this purpose although specs vary pretty widely.  A common theme however seems to be the provision of one or more 7200rpm drives.

There's some interesting stuff about running centrinos cooler at this forum

There is some general advice at this SOS link

See also this SOS article but note the date.

http://www.silentpcreview.com/ was an interesting find but not particularly helpful in this context.  See also http://www.silent.se/ and http://www.endpcnoise.com/ where there is a laptop "heat pipe" cooling device.

See http://www.rainrecording.com/livebook for the spec of a custom audio laptop.





Blurk, presumably you don't want simply to record at 44.1/16 bit in stereo, because if so, personally I'd much rather use a CD recorder than a PC for that.  I'm only interested in the Traveler/laptop combo in order to be able to record more than a stereo pair and to remix afterwards.
Logged
Reply #3
« on: June 20, 2005, 08:51:47 AM »
pwhodges Offline
Member
*****
Posts: 1386

WWW

Quote from: blurk
I've since heard elsewhere that the M-series processors can actually outperform P4s.  Can anyone shed light on such claims.  I'm also curious to know whether the Pentium-M processors suffer from the P4 denormalisation problem - because if they don't, that could also sell me on a Pentium-M processor.

I may be out of date, because some names (like Celeron) have carried over to fundamentally different processors, but last time I looked the Pentium-Ms were a development of the Pentium-III range not the Pentium-IV.  In any case, they run so much cooler than P-4s that they have a huge advantage in laptops, and are likely to be much quieter running even if a fan is required.

Paul
Logged
Reply #4
« on: June 20, 2005, 09:05:10 AM »
blurk Offline
Member
*****
Posts: 408



Quote from: ozpeter
Blurk, presumably you don't want simply to record at 44.1/16 bit in stereo, because if so, personally I'd much rather use a CD recorder than a PC for that.

Well, I'd consider recording at a higher bit depth, but basically I do most of my composition with soft synths, so I would only be occasionally recording external inputs.  This would mainly be the old analog synth, but also perhaps bass guitar/electric guitar.  Or the radio.  Hmm, forgot about that.  But hey, I don't see a benefit in higher bit depths and sampling frequency just for the FM radio.  I'm not really interested in a separate CD recorder - I have no real use for it given the multitude of PC-based recorders at home.  smiley

But back to recording, I guess with a better audio interface I should really consider recording at higher bit depths.  On the desktop PC I have an M-Audio AP2496, but I never remember to actually record at anything other than 16-bit.  Old habits and all that, not to mention not actually being an audio professional.

Thanks for the link, BTW, I'm checking them out now.

Quote from: pwhodges
...last time I looked the Pentium-Ms were a development of the Pentium-III range not the Pentium-IV.

Yes, that's what I've also heard.  In this context I think it is likely to be a good thing.  You've pointed out one reason: heat.  I'm hoping another is an absence of the denormalisation bug/problem/feature.
Logged
Reply #5
« on: June 20, 2005, 11:02:28 AM »
SteveG Offline
Administrator
Member
*****
Posts: 10547



Quote from: Euphony


Quote
it seems that USB is fine to run a hard drive

As a storage unit, yes.  The maximum usb2 can transfer is 60mb/s (theoretically), which is plenty fast for backing up files.

I think that this is the maximum bi-directional transfer rate, judging from the performance of a real drive - and the results of that are quite interesting. The first thing to note is that every external drive you can buy appears to be formatted as a FAT32 drive - both of the drives that I've now got have performed the benchmark tests miles better when reformatted as NTFS.

But what is interesting is that I also managed to acquire a PIKAone external disk, and that supports both FireWire and USB2 connectivity. Now, the drive fitted to it is no slouch - it's a Seagate Barracuda 7200 200Gig. Yes, I know that there are faster drives about, but this is more than adequate for this purpose, and the results are as follows:

USB2:
Sequential Read: 28.9MB/s
Sequential Write 24.1MB/s
Random seek @ RW 4.3MB/s

FireWire:
Sequential Read: 22.2MB/s
Sequential Write 13.6MB/s
Random seek @ RW 4.0MB/s

No real attempt at optimisation (other than the reformat) has taken place at all - this is a straight replug, and nothing else was altered. Simply for comparison purposes, the results from the WD courier (USB2) at the same time were:

Sequential Read: 27.5MB/s
Sequential Write 20.6MB/s
Random seek @ RW 4.6MB/s

I can't really comment on the M-series processor benchmarks, except to say that I've heard that for some perhaps rather critical operations in terms of audio processing, that the P4 still outperforms the M processor - but I'd guess that it's not by much, and that all of these tests depend to a large degree on the external components as much as anything - a test in a non-laptop MOBO might not actually reveal everything there is to say about this. But it's pretty clear that with a reasonably fast version of either series, the results with audio should be more than adequate for most purposes. Which means of course that in a laptop, the M series wins hands down!
Logged

Reply #6
« on: June 20, 2005, 11:47:16 AM »
SteveG Offline
Administrator
Member
*****
Posts: 10547



Quote from: ozpeter
I've now got the budget but I'm still in a quandary about the laptop itself, really entirely on the grounds of noise (not affecting the recording because I'm always too far away, but affecting the audience if I'm working in the auditorium, which tends to be the norm sadly).  I' thinking about a 1.7+ Centrino rather than a P4 on those grounds, but I suspect that Audition at least will be happier on the P4.

I had a feeling that it was going to turn out like this. Currently, I'm considering some other acoustic remedies to reduce the effect of the fan noise without affecting the cooling - there are in fact a number of options here, but the precise implementation of them varies according to the laptop you've got, and how intrusive you mind them being.

Unfortunately the only way that you can really determine how much noise a laptop actually makes is to do an accurate intensity measurement of it - and that's not the easiest thing in the world to do, even when you have access to the relatively expensive kit necessary to do it. This way you end up with not only an effective radiated acoustic power figure, but also a footprint of the noise radiation pattern - this will vary around the machine. But if you want a clue as to what might prove to be effective (this applies to any laptop), then just cup your hand around the fan exit point, and see how much your perception of the fan noise alters when you move  your hand about a bit... there's quite a lot you can do with this. There's another option involving a plexiglass fold-down screen as well, but this would be a little harder to engineer, despite probably being remarkably effective. We've done tests with these on HF fan noise before, and discovered that you can reduce it significantly. This particular solution has the advantage of reducing the actual fan noise, as well as the airflow noise, and that can often make all the difference. But all of these things are going to require one to make up a special rig, and with the top cover approach, you lose immediate access to the keyboard. With a Red Rover though, this might not be such a problem.
Logged

Reply #7
« on: June 20, 2005, 02:44:24 PM »
Wildduck Offline
Member
*****
Posts: 881



I guess this is the place to try to ask this question again, and also sow more doubts.

I have a bunch of laptops and a Soundblaster mp3+ usb audio interface. The SB works acceptably on all the older laptops.
On the latest 1.5GHz Pentium M machine, I get a message from the OS saying something like "A power surge has occurred on usb channelxxxx, and the device has been shut down". The usb channel has to be reset before it will work again for anything. After a bit of googling, I tried disabling the warning message, but the SB still doesn't work. If I plug in a powered hub, it works fine on the same machine.

I've tried on a friend's Centrino machine, and the same effect occurs. Both these machines are usb2, all the older machines are usb1.
Another usb audio interface that I have works fine. It doesn't have the nice blue LED of the SB one though Smiley

I understand about high power usb devices powering up at low power and then switching to high power mode when everything is ready. What I can't find is any real info or test software to monitor usb power usage. The limited info in Device Manager doesn't really help much.

I don't have any 'real' usb audio device to test with, and this might well just be some quirk of the design of the SB unit.
Logged
Reply #8
« on: June 20, 2005, 03:41:59 PM »
Jester700 Offline
Member
*****
Posts: 599



A couple thoughts...

The theoretical max for USB2 is 60 MB/s, not Mb/s (that's 480mb/s).  as Steve found, it's not really that fast in reality, but is usually good enough.  I found his results interesting, as they show an advantage of USB2 that I've not seen in other test results.  Usually the two are on par, and his 13.6MB/s result is way under what I've seen.  I wonder if the particular chipset used in the interface device affected this?

The Pentium M is based on the PIII architecture, and PIIIs have always been faster than P4s - at the same clock speed.  I think that's what people are talking about.  Of course, a desktop P4 at 3 GHz will outdo a 1.5GHz Pentium M; the superior architecture only helps so much.
Logged

Jesse Greenawalt
Reply #9
« on: June 20, 2005, 03:54:31 PM »
SteveG Offline
Administrator
Member
*****
Posts: 10547



Quote from: Jester700


The theoretical max for USB2 is 60 MB/s, not Mb/s (that's 480mb/s).  as Steve found, it's not really that fast in reality, but is usually good enough.  I found his results interesting, as they show an advantage of USB2 that I've not seen in other test results.  Usually the two are on par, and his 13.6MB/s result is way under what I've seen.  I wonder if the particular chipset used in the interface device affected this?

At present, I honestly don't know what's affecting the results at all. What I'll do at some stage is the same test sequence (it's not of my making) with this drive on a desktop machine, and see what the difference is. I was assuming that Euphony actually meant MB/s not mb/s - but I'm still pretty sure that this is the bi-directional maximum speed - if the interface is going flat out duplex, each way would only support half of that rate. But I must confess to being slightly surprised by the results, anyway.
Logged

Reply #10
« on: June 20, 2005, 04:56:16 PM »
Euphony Offline
Member
*****
Posts: 371



Quote from: Jester700
The theoretical max for USB2 is 60 MB/s, not Mb/s (that's 480mb/s).  


Sorry, I did mean Megabyte (MB).

SteveG's results on Firewire vs. USB2 speed are interesting as well.  Although USB2 has a higher theoretical max speed, I've heard that firewire usually performs better.  

Funny enough, a Pentium-M 770 @ 2.5ghz outperforms a Pentium 4 Extreme Edition @ 3.4ghz.

http://www.tomshardware.com/cpu/20050525/pentium4-10.html
Logged
Reply #11
« on: June 20, 2005, 06:05:03 PM »
SteveG Offline
Administrator
Member
*****
Posts: 10547



Quote from: Euphony

SteveG's results on Firewire vs. USB2 speed are interesting as well.  Although USB2 has a higher theoretical max speed, I've heard that firewire usually performs better.  

That is exactly what I'd heard too.
Logged

Reply #12
« on: June 20, 2005, 08:28:47 PM »
Jester700 Offline
Member
*****
Posts: 599



Quote from: Euphony

Funny enough, a Pentium-M 770 @ 2.5ghz outperforms a Pentium 4 Extreme Edition @ 3.4ghz.

http://www.tomshardware.com/cpu/20050525/pentium4-10.html

I guess that stands to reason.  The old PIII 1.2 GHz was on par or slightly faster than the newer 1.6GHz P4 (IIRC), for a roughly 33% apparent speed increase.  I guess that still holds, and maybe then some.
Logged

Jesse Greenawalt
Reply #13
« on: June 20, 2005, 11:32:47 PM »
Blueturnswhite Offline
Member
*****
Posts: 96

WWW

Quote from: Wildduck
If I plug in a powered hub, it works fine on the same machine.
Quote


i have the same problem with my dmx controller (lights)....not sure why....but i just use a hub....why fight it
Logged

Reply #14
« on: June 21, 2005, 08:53:46 AM »
ozpeter Offline
Member
*****
Posts: 2377



My thinking today is that any laptop (or one-piece multitrack for that matter) is likely to be too noisy to use in amongst an audience.  I mean, I've even been worried by the noise from my camcorder transport under those circumstances (a biggie that uses full-size DV tapes).  So, forget the noise, buy the best all-round performer for the purpose, and either find a room aside even if the tedium of running longer mic cables is involved, or just don't use that rig on occasions when a noise-free rig is required.

I'm back with thinking that the nearest system I can implement to SteveG's the better - given that his works.... and he knows of what he speaks.

My thinking tomorrow might be different!  (Er, no, not about Steve!)
Logged
Pages: [1] 2 3 ... 5 Print 
« previous next »
Jump to:  

Powered by MySQL Powered by PHP Valid XHTML 1.0! Valid CSS! Ig-Oh Theme by koni.