Avid Pro Audio Community

Avid Pro Audio Community

How to Join & Post  •  Community Terms of Use  •  Help Us Help You

Knowledge Base Search  •  Community Search  •  Learn & Support


Avid Home Page

Go Back   Avid Pro Audio Community > Legacy Products > 003, Mbox 2, Digi 002, original Mbox, Digi 001 (Win)

Reply
 
Thread Tools Search this Thread Display Modes
  #31  
Old 04-19-2001, 10:03 PM
davec davec is offline
Member
 
Join Date: Nov 2000
Location: Montreal, Quebec
Posts: 651
Default Re: PT Performance

I agree with you. I'm waiting to see more people post their results with the various MOBO setups but so far the Athlon boards are kicking @$$.

Makes me wonder when the P4 equiv. comes out from them how much more they'll rock. I love messing around with my sounds which means lots of LP Filters and Phasers for LFO and such with lots of delay and all these little things that I put together to make "my sound" cost lots of CPU power. So the more power I can get the more creative essentially I can be which is what making music is all about right -- pushing the limits of technology whether it's your instrument or the equipment you use to produce your music.

------------------
--D
__________________
Peace...

--davec
StudioFX Multimedia
SpellWell Software
Reply With Quote
  #32  
Old 04-22-2001, 04:11 PM
JPS JPS is offline
Member
 
Join Date: Feb 2001
Location: Ohio
Posts: 924
Default Re: PT Performance

Why don't you call it the "Davec CPU Torture Test"!!!
__________________
John-Q6600,GA-EP35-DS3P, Zigmatec S1283: 235 D-Verbs @ 2.4 ghz
Reply With Quote
  #33  
Old 04-22-2001, 08:43 PM
Mark_Knecht Mark_Knecht is offline
Member
 
Join Date: Apr 2001
Posts: 2,500
Default Re: PT Performance

Dave,
Sorry I'm late on this, but non the less here's my setup. Definately a bit low end.

PT Ver. 5.1
Dell motherboard - Dimension XPS T500
Intel PIII - 500MHz
256MB
1 ATA-66 12GB HD (7200) DMA enabled
1 SCSI controller for external CD-RW
(and soon an audio disk....)

9 Tracks

I will post a bit more information later. I'm a bit more interested in some of the PC architectural issues, and looking more into what's up with PCI Bus errors as opposed to this test which stresses the CPU.

Thanks for putting all this info together.

Mark
Reply With Quote
  #34  
Old 04-22-2001, 10:59 PM
davec davec is offline
Member
 
Join Date: Nov 2000
Location: Montreal, Quebec
Posts: 651
Default Re: PT Performance

JPS - yeah, I guess it is somewhat of a sadistic test , no reflection on me of course...

Mark - thanks for the post. It's never too late to gain knowledge, unless your in the stock market. Mark your numbers seem about right. I'm getting twice as many tracks as you with almost twice the CPU power. At least the Intel chips are consistent.

BTW - the one-pager is entitled:
Rig Configurations Benchmarks - Which MOBO combo do I buy?

I've also included some keywords at the top of the document so it'll come up on various word searches. I'll keep updating it as long as the community keeps posting in this thread.

------------------
--D

[This message has been edited by davec (edited April 22, 2001).]
__________________
Peace...

--davec
StudioFX Multimedia
SpellWell Software
Reply With Quote
  #35  
Old 04-23-2001, 01:52 AM
davec davec is offline
Member
 
Join Date: Nov 2000
Location: Montreal, Quebec
Posts: 651
Default Re: PT Performance

Well folks so far we only have about 6 different board configurations. Statistically, not a very large sample to draw conclusions from but such is the situation.

I guess from the data in the posts, about the only conclusion that can be drawn is that the Athlon (unsupported) rigs are the best choice. Though with the number of samples taken the margin for error is probably +/- 50%

Thanks again to all those who've participated and if you wish to contine posting by all means please do so; however, the topic, sadly, no longer will be refreshed at this point, even though people still post the infamous "What system should I buy?" question.

I will create a final post with the numbers posted by everyone in a single one pager with a relevant title for easy searching.

------------------
--D
__________________
Peace...

--davec
StudioFX Multimedia
SpellWell Software
Reply With Quote
  #36  
Old 04-23-2001, 10:35 AM
Mark_Knecht Mark_Knecht is offline
Member
 
Join Date: Apr 2001
Posts: 2,500
Default Re: PT Performance

Dave,
Yep, I thought the numbers looked about right also.

However, the problem I'm looking into is that these results are not translating into my ability to record longer, live tracks. I want to do some simple stereo or maybe 3 channel recording of a a vocal group. I'm passing your test, but I'm finding a statistal dropout at anywhere from 3 minutes to 10 minutes with PCI errors. Recording stops and I have to start over.

2 channels works reliably for 10 minutes at a time. 3 channels will die at 5-10 minutes 1 out of 4 times. 4 channels will die at 2-6 minutes 4 out of 5 times, but never less than 1 minute which was your criteria for this data.

I think that there is room for us to look at more ways of testing, and I'll let you know if I discover anything substantial.

(I think QuikDraw was pointing this out earlier. I personally don't care much about processing power when I'm getting these sounds onto disk.)

Thanks,
Mark

"More processing power than the A-bomb team, more tracks than the the Beatles - a lot less brains than Enrico Fermi and George Martin"
Reply With Quote
  #37  
Old 04-23-2001, 04:58 PM
davec davec is offline
Member
 
Join Date: Nov 2000
Location: Montreal, Quebec
Posts: 651
Default Re: PT Performance

Mark,

I think your PCI issues could be two fold: either bus sharing with another card you have in your box or the single drive. Remember PT is constantly writing to disk as well as Windows virtual memory kicking-in. Once you setup your 2 drive system you may see the PCI errors disappear. As for the bus sharing, I find I can't have any other cards on board (except AGP of course -- different bus). This may be a factor of the MOBO's ability to handle bus-sharing. Some people can run a sound card and the 001 simultaneously -- I can't. For me to get the system running clean is to only have the 001, and try to dissable (in Control Panel) as many devices that are unused as possible. In this manner, the OS is working less on machine fundamentals and more on PT and I still can only get 18 which others can go well over the 24 tracks. My test really prooves that my MOBO sucks which I guess was one of the sub-objectives of the experiment

What would be really nice is if Microsoft came out with a shrunk version of Windows at contains only the fundamentals required for either audio processing (WinAudio2k) or (WinVideo2k) for video processing. If they do that, they'll blow Apple out of the water. IMHO.

------------------
--D

[This message has been edited by davec (edited April 23, 2001).]
__________________
Peace...

--davec
StudioFX Multimedia
SpellWell Software
Reply With Quote
  #38  
Old 04-23-2001, 06:53 PM
QuikDraw QuikDraw is offline
Member
 
Join Date: May 2000
Location: Yucaipa, California, USA
Posts: 2,107
Default Re: PT Performance

Acutally Dave, separate stripped-down versions of Windows for each intensive processing need would suck big time!

Having a single machine to do multiple tasks is the very reason I own a PC in the first place. I don't want a separate Internet appliance. I don't want a separate word processor. I don't want a game console. I don't want a stand-alone digital audio recorder/editor.

We don't need separate Windows versions. All we need is for Microsoft to write APIs that allow more hardware control to applications instead of asking Windows' permission for everything. DAW, video, and game programmers should be able to hog the system while performing these intensive processing tasks.

Please!... Not more versions of Windows! I've been waiting a decade for the two versions of Windows to merge. Now that that is finally about to become a reality, we're talking about splitting off even more versions? Please! No!...



Mike
__________________
-- Mike
- MacBook Pro 15.4" (9,1) - 2.3 GHz i7 16GB RAM
- OS X 10.14.1 Mojave / Windows 10 (Bootcamp) - PT running on Windows partition
- PreSonus Quantum 2
- PreSonus Studio 24c
- MBox 3 Mini
- Mackie Onyx 1640i
- Behringer X32 Producer
- PT 2019.12 (vanilla)
- Various USB2/3 and Firewire hard drives
Reply With Quote
  #39  
Old 04-23-2001, 10:15 PM
Mark_Knecht Mark_Knecht is offline
Member
 
Join Date: Apr 2001
Posts: 2,500
Default Re: PT Performance

Dave,
I had the same thoughts today. I made a breakthrough (for my machine!) and simply unplugged my Ethernet cable, based on the idea that if packets aren't arriving, then there won't be a conflict. This cleaned things up significantly. Now I'm doing 4 tracks for 8 minutes most of the time. While that's not enough to do a good drum kit, it should allow me to get the vocal/keyboard stuff I want to do immediately. (BTW - I'm not sure your data was clear about what buffer size people are supposed to use. I was doing 1024. Things get worse quickly at 512 and below, for me...)

I do have other boards in the system to experiment with turning off - Sound Blaster and 1394 OHCI Controller. However, I doubt they are contributing much to this specific problem, but could be. We'll see...

As for the second drive, while I'm sure that will help some, I want to understand the real issues first. I have worked on PC chipset design and understand the bus issues pretty well. As I see this architecture, we're really driving the PCI bus twice for every audio packet, once when the PT hardware writes a new packet to memory, and a second time when it's written back to disk for storage. My suspicion is that motherboards that allow some tuning of the PCI MAX_GRANT and MAX_LATENCY values would probably demonstrate this pretty quickly.

From another direction, my suspicion suggests that a motherboard that actually has two separate PCI buses might work really well. I have a Dell PowerEdge 1400 Server at work. It's got 6 PCI slots - 4 64-bit and 2 32-bit. My understanding from Dell is that this machine actually has two separate PCI bus segments. I'm thinking that the PT hardware on one bus and the disk hardware on the other would be sweet as you would truly double the bus bandwidth AND allow both input and output operations to happen at the same time. Unfortunately, it's got a fan the size of a small airplane and the noise is terrible. (As shipped)

Thanks for your interest.

Mark

[This message has been edited by Mark_Knecht (edited April 23, 2001).]

[This message has been edited by Mark_Knecht (edited April 23, 2001).]
Reply With Quote
  #40  
Old 04-23-2001, 11:29 PM
davec davec is offline
Member
 
Join Date: Nov 2000
Location: Montreal, Quebec
Posts: 651
Default Re: PT Performance

Mike,

Although you're right in that one appliance to do all your work is "convenient", when I look at the Windows98 SE footprint of 80MB, and I compare say Linux even with X-Windows your talking about 20% the size of Win98SE, and Linux is a true 32-bit multi-tasking/ multi-threaded/multi-user operating system. So it makes me wonder how much "fat" in the code there may be. Linux has come up with shrinked versions to support say Firewalls, or Webservers that can run on minimal systems.

I think with something so specialized as audio production, or video production, tweaking the OS could reap enormous benefits. I think what you're talking about is Microsofts inability to maintain stable releases.

What I'm suggesting is much like gaming machines like the X-Box, or PS-2, etc. which have been custom built for the gaming industry, or even MP3 players. Or like race cars, you don't take your off-the-lot honda-civic and pace it against a custom built hot-rod. IMHO, I think OSes should be, to certain degree, application specific for the sole purpose of optimisation.

When Macs first came out, they were hailed as the desktop platform of choice for it's desktop publishing capabilities (graphics and such) and of course the easy to use user interface. Once again, application specific.

And lastly, to paraphrase Apple co-founder Steve Wozniac, "...we should be focusing on maing things smaller and faster, not bigger and thus requiring more power...".


Mark,

Sounds like you've had some interesting background work. Are you an engineer?

On the seperate pci bus issue, there definately would be enromous gains provided proper queuing unless the busses are hot wired to the CPU. I'm curious, what 64bit OS are you running on those servers? The processor would also have to be 64bit otherwise there would conversion issues reading and writing to and from the 64bit bus.

A quick discussion with Eric Day about making the "inserts" multi-threaded sparked some interest which means we'd be able to capitalise on the dual and quad processor boxes and be able to power up more plug-ins. Don't know how far down the line that could be, in theory, the actual code changes should not be too great espscially as they are already using OOP, which means, in essence, all they'd have to do is instantiate the plugin object within a thread. Piece of cake.


BTW - I love this thread we're on, this is great bouncing this off you guys. It's been awhile since I've gotten into philantropical discussions on computers.


------------------
--D

[This message has been edited by davec (edited April 23, 2001).]
__________________
Peace...

--davec
StudioFX Multimedia
SpellWell Software
Reply With Quote
Reply

Thread Tools Search this Thread
Search this Thread:

Advanced Search
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump

Similar Threads
Thread Thread Starter Forum Replies Last Post
Macbook benchmarks/Davec Tests..... ArtGarceau 003, Mbox 2, Digi 002, original Mbox, Digi 001 (Mac) 0 01-31-2007 09:34 PM
24/96 davec tests? Tyler Crawfprd 003, Mbox 2, Digi 002, original Mbox, Digi 001 (Win) 1 12-15-2006 03:06 PM
Davec I,Davec II Tests Both Downloadable Here davec 003, Mbox 2, Digi 002, original Mbox, Digi 001 (Win) 3 11-27-2005 08:45 PM
hey.... what about the MAC DaveC tests? JonnyGinese 003, Mbox 2, Digi 002, original Mbox, Digi 001 (Win) 1 11-16-2005 02:02 PM
Results from Davec tests SEBJOERG 003, Mbox 2, Digi 002, original Mbox, Digi 001 (Win) 10 01-29-2002 12:53 PM


All times are GMT -7. The time now is 02:01 PM.


Powered by: vBulletin, Copyright ©2000 - 2008, Jelsoft Enterprises Limited. Forum Hosted By: URLJet.com