New, Updated SonarTest

I have no idea what I'm doing, but I was unhappy to see my computer wimp out at a lower latency.. I usually work @ 2.9msec without any problems, but this test wouldn't have it.. I usually bounce tracks a lot so I've never even seen my cpu meter go over 30% until this test... maybe I need to do some more tweaking..


P4 2.53/1Gb of ram/M-Audio Firewire 410/Maxtor 160GB hd for storage


64 samples - 1.5 msec - dead
128 samples - 2.9 msec - dropout
256 samples - 5.8 msec - cpu 57%
384 samples - 8.7 msec - 50%
512 samples - 11.6 msec - 46%
768 samples - 17.4 msec - 42%
1025 samples - 23.2 msec - 38%

p.s.

cool test!

Toshiba Satellite A30 Notebook, 512Mb RAM, P4 Mobile 3.06G HT, Intel 82852/5 Graphics

Asio4all driver for AC97 :o)

MP off MP on
46.6---------28-------27
23.2---------30-------30
11.6---------39-------35
5.8----------47-------46
2.9----------68-------64
1.5----------84-------79


I was surprised to get this low with the AC97 - the asio4all driver is a real find. I typically use it at 5.8 and it doesn't sound too bad for mixing on the move.


Cliff Speight

**Superstition brings bad luck**

My system:

Dual AMD Opteron 246 2.0 GHz
Tyan Tiger K8W Motherboard (AmiBIOS 1.03)
AMD 8111/8151 Chipset
1 GB Registered DDR400 Memory
1 Maxtor 40 GB 7200 RPM ATA133 HD
1 Maxtor 160 GB 7200 RPM ATA133 HD
ATI Radeon 9000 Pro AGP 4X
Delta 1010 .27 drivers
UAD-1 (2) 3.5.1 software/drivers
Windows XP Pro
Sonar 3.1.1

MP Enabled:

46.4 17%
23.2 18%
11.6 20%
5.8 24%
2.9 33%
1.5 56%

MP Disabled:

46.4 28%
23.2 31%
11.6 35%
5.8 37%
2.9 49%
1.5 68%

Okay boys and girls, just updated my system from a standard Intel P4 3.2 ghz to the new P4 3.2 ghz Extreme Edition. Below is the data:

First, I ran the complete test again with my standard 3.2ghz P4. I used the ASIO driver in my Tascam FW-1884. I installed the EE, and ran the same test with the same driver:

P4 3.2 ghz---------P4 3.2ghz Extreme Edition
46.4: 25%----------------24%
23.2: 26%----------------26%
11.6: 32%----------------29%
5.8: 41%----------------36%
2.9: 61%----------------50%
1.5: Won't engage.------81%

There appears to be a significant improvement at the lower latencies. I ran this test immediately after installing the chip. I will run it awhile and report what happens..

The chart has been updated with P4EE and Opteron 246 results.

-S

I'd really love to see some dual xeon numbers.

ORIGINAL: deiseldave

I'd really love to see some dual xeon numbers.

Me too! Anyone??

Hi Scott,

Thank You for puting together a "real" test that will no doubt influnence my next purchase.

Using ASIO Drivers
46.4.....49
23.2.....54
11.6.....63
5.8......85
2.9......WNR
1.5......WNR


1.8 P4
2X 120gb Seagate Barracudas 7400 RPM
1 gb RAM
Delta 1010 V.42 driver


Dave

Just wanted to say thanks to Scott and everyone who posted for this thread. I'm just about to order a new system and this has helped me out a lot. If anyone cares to second-guess my hardware, I'd be much obliged. I posted the shopping list in the gear forum.

Click:
yep's new computer
If you care to set me straight before I cough up the dough. Either way, I will post my own results as soon as I get the new machine up and running...

Cheers.

Yep,

Definitely get the K8NS Pro instead of the K8N Pro. This one has the new nForce3 250 chipset. I've built a couple of systems with it so far, and it's a winner.

-S

ORIGINAL: yep

Just wanted to say thanks to Scott and everyone who posted for this thread. I'm just about to order a new system and this has helped me out a lot. If anyone cares to second-guess my hardware, I'd be much obliged. I posted the shopping list in the gear forum.

Click:
yep's new computer
If you care to set me straight before I cough up the dough. Either way, I will post my own results as soon as I get the new machine up and running...

Cheers.

Hey Scott,

Just out of curiousity, would Yep get any better performance with the Athlon 64 3400+ ? Also, is there anything new on the horizon that you know of?

Oh, I also appreciate this post...........it has helped me a bunch.

The 3400+ bumps the clockspeed up from 2.0GHz to 2.2GHz. It's definitely faster, although quite a bit more expensive, as it is the bleeding edge (other than the AthlonFX series).

Socket939 Athlon64s with dual-channel memory controllers are coming in about a month. We'll see what impact this has. Current AthlonFX and Opteron CPUs have dual channel memory controllers... but they all require registered memory, which takes a small latency hit. Socket939 CPUs with dual channel controllers should be the fastest choice yet when they arrive.

-S

ORIGINAL: EbonyFunk

Hey Scott,

Just out of curiousity, would Yep get any better performance with the Athlon 64 3400+ ? Also, is there anything new on the horizon that you know of?

Oh, I also appreciate this post...........it has helped me a bunch.

Hey, I expect Sonar to arrive Monday or Tuesday. My dual Xeon system is ready. http://www.cakewalk.com/forum/tm.asp?m=132372

The first thing I will do is run the test. But my question is; where is the actual bottleneck on the computer system as it pertains to a DAW? Is it the bus speed? Is it the transfer of hard drive data? Is it the Ram speed?

I know it isnt the processor speed! By all accounts, if I am running a dual 2.4G Xeon with a 533 FSB the entire system will be slower than a single 2.4 P4 system with a 800 FSB.

Thanks for any tips...

The first thing I will do is run the test. But my question is; where is the actual bottleneck on the computer system as it pertains to a DAW?

It mostly comes down to bus speed, bus architecture, and memory controller architecture. The Xeons have a slower bus than the high end P4s (533MHz vs 800MHz). They also share a single bus between multiple CPUs... amplifying the impact of having a slower bus. With Opteron, the bus speed is the same as Athlon64/FX, each CPU gets its own independent bus, and the memory controller is actually on the CPU... eliminating the middle man to the memory.

-S

Scott and all-
thanks for the advice!
Will definitely check out the Kn8S Mobo.
And, yeah, the Athlon64 3200 isn't bleeding edge, but seems to have a comfortable spot at the price/performance knee, which is what I'm looking for. My 3-year-old 1.5gHz P4 is definitely starting to sag under the weight of some of the newer plugins...
Cheers.

If other devices run slower than the FSB wouldnt they be the bottleneck? Why isnt the transfer of data from the hard drive to the sound card the biggest bottleneck?

ORIGINAL: yep

Scott and all-
thanks for the advice!
Will definitely check out the Kn8S Mobo.
And, yeah, the Athlon64 3200 isn't bleeding edge, but seems to have a comfortable spot at the price/performance knee, which is what I'm looking for.

Yeah... the 3200+ is a good value for money.

My 3-year-old 1.5gHz P4 is definitely starting to sag under the weight of some of the newer plugins...

If you are coming from a 1.5GHz P4... you are in for a real treat.

-S

ORIGINAL: Scott Reams


If you are coming from a 1.5GHz P4... you are in for a real treat.

-S

Yeah, I reckon so. I'm a pretty good recording guy, but a pretty mediocre computer guy, so once I get a system working, I am really loathe to alter it (I had Sonar 3 in a box for months before i actually installed it...).

But I have the impression that many of the newer systems are a lot less finicky then the old ones used to be, and thanks to your help, I expect a pretty smooth transition.
Best Regards,
yep

Anyone tried the Albatron K8X800 ProII
?

Scott and all,

Does it matter that Yep purchased a "matched pair" of Corsair memory sticks? I just purchased two 1Gb Corsair memory sticks -- individually; not a pair -- for a dual-proc K8W mobo. Tyan's website recommends them but, given the price, I'm not so sure they're optimal. Hope I didn't screw up.

Any advice is appreciated.

Anyone know if the Asus SK8N is compatable with the UAD1?

Thanks,

berry

ORIGINAL: berry thomas

Anyone know if the Asus SK8N is compatable with the UAD1?

Thanks,

berry

I believe the nForce3 chipset is compatible in general. It's probably a safe bet.

-S

Scott Reams,

Check this thread if you haven't already. Mr Tankersly is giving you some deserved props.

Scott

ORIGINAL: ScottS

Scott Reams,

Check this thread if you haven't already. Mr Tankersly is giving you some deserved props.

Scott

Thanks. I just posted to that thread... interesting stuff.

-S

46.4ms........1225%
23.2ms........1227%
11.6ms........1230-1231%
5.8ms..........1335-1337%
2.9ms..........1543-1546%
1.5ms..........1862-1867%

I assume everyone missed the real genius behind this....Glennbo has actually overclocked a Pentium to produce 1762% more CPU Cycle effciency than is actually installed in his computer. It also means he has sekretly overcome the buss issue prevalent on all motherboards today. Amazing...I'm surprised everyone missed that.

hi i've had my daw for 2 months now. the computer end of things is the most confusing for me. here's my system as i understand it. amd, athlon 3000+, 2.16 512 ram, 160 gig . sonar 3 pe, delta 1010 lt. i performed the test but my latency slider (even when i move it) stays/reverts back to 11.6. even when i move the slider the number - 11.6 does not change. it seems i can only test at 11.6. the result was spot on scott's graph - amd green line 44% - 49%. can someone explain what this is all about. what is "overclocking the cpu"? i have optimized xp home ed. for audio according to specs i read in eq mags. etc. things really cleared up when i set hardware acceleration to "none". is there anything i can do to this machine besides get more ram? edit: asio drivers.

Sammy,
In ASIO mode you need to adjust the latency by changing the DMA buffer size in the Delta control panel.

As for overclocking, that's forcing a processor to run at a faster clock rate than it was spec'ed at by the manufacturer. Fun to try, but easy to lose stability. I wouldn't mess with it on something as important as your DAW.

yes wogg, greenkid had the same question. here are my results 46.4 = 35%, 23.2 = 38%, 11.6 = 46%, 5.8 = 60%, 2.9 = 91%, 1.5 = audio stopped. based on my 2.16, 512 ram which setting should i go with? default is 512 samples = 11.6. as i read on in this forum, i notice people talking about bouncing tracks. isn't this why i bought a computer and sonar so i wouldn't have to bounce tracks?

ORIGINAL: sammyp
...as i read on in this forum, i notice people talking about bouncing tracks. isn't this why i bought a computer and sonar so i wouldn't have to bounce tracks?

If you use that computer as a straight audio recorder like a tape machine, with no effects, softsynths, etc, then I am sure that your computer will handle well over 100 simultaneous tracks at very low latency. So if that's why you bought Sonar, fear not, you shall never have to bounce audio again. BUT if you want to use your computer to do stuff that a tape recorder can't do, like actually generate sounds and effects... that stuff can be pretty demanding, especially if you want to run more complicated (read better-sounding) reverbs, synthesizers, virtual instruments, etc.
This test is, if I'm not mistaken, deliberately designed to strain exactly those processing and synthesis capabilities pretty aggressively, not to test how many tracks you could record.

FWIW, anyway. Cheers.

yeah being new to the whole thing i've developed a slight paranoia over the computer thing - one of my good friends (a non recordist, non musician) computer geek, gave me alot of teasing for being inexperienced and buying an "off the shelf" H.P. AMD 3000. but i'm a 24 - 40 track kinda thinker - i only use reason "redrum" and the rest is mostly audio in sonar. i just took a 20 track work in progress and jammed the tracks full of reverb - eq - comp plugs etc ,reduced latency to 5.8 just hammer the cpu. the computer did fine for the ridiculousness of the plugs. i think my machine will do the trick for a while barring unforseen circ. i don't know if anybody else experiences the fact that READING THE SONAR FORUM TOO MUCH LEADS TO SERIOUS COMPUTER/GEAR ENVY SYNDROME.

ORIGINAL: sammypi
don't know if anybody else experiences the fact that READING THE SONAR FORUM TOO MUCH LEADS TO SERIOUS COMPUTER/GEAR ENVY SYNDROME.

Yeah, dude. If your system is working for you then leave it alone. If all you need is a couple dozen tracks of audio, some drum samples and a couple of reverbs, a PIII with 512 RAM and a couple fast hard drives might do you alright, provided that you're cool with relatively uncomplicated reverbs.
$300 worth of new mics, guitar strings, piano tuning, bass amp, voice lessons, whatever is more likely to help your sound than $1000 worth of new computer that can run 30 softsynths if you never use softsynths. Cheers.

THANKS MY FRIEND, LUCKY FOR ME THE MUSIC IS WAY AHEAD OF THE TECH. GOT THE VOICE, , MICS, AMPS, GUITARS, MUSIC DEGREE IN THE BAG FOR THE MOST PART. (ALWAYS LOTS OF ROOM FOR IMPROVEMENT). UNFORTUNATELY, ENGINEERING AND MIXING (LIKE THE TECH SIDE) ARE LIKE LEARNING TO PLAY A BRAND NEW INSTRUMENT, EH?

Also, will duel Opera-Trons make my hihat sound any better?

You could sound four times betterer.
http://www.tyan.com/products/html/thunderk8qs.html

Sager Laptop; NP8890 / P4 / 3.4ghz with ht / 1 gb 400ddr ram / 800 mhz fsb / 2 internal 7200 60 gb hds / Echo indigo IO (WDM):

Latency -- w/out HT -- w/HT (meaning w/ multiprocessor checked)
46.4----------25%-------23%
23.2----------28%-------25%
11.6----------33%-------30%
5.8 ----------42%-------40%
2.9 ----------63%-------58%
1.5 ----------XX%-------96% and then craps out

Needless to say, I'm a little disappointed. If anyone has any tuning tips that really help latency, I'm all ears.

I just switched to ASIO drivers, and there's a slight improvement at 2.9 (the indigo won't let you choose 1.5) - - sits around 56%.

ORIGINAL: sammyp
...UNFORTUNATELY, ENGINEERING AND MIXING (LIKE THE TECH SIDE) ARE LIKE LEARNING TO PLAY A BRAND NEW INSTRUMENT, EH?

yep.
but don't sweat it. If your music is good, just cut some clean, listenable demos, and one day you'll have a trained professional to record it all for you.

Cheers,
yep

ORIGINAL: LettuceCheese

Sager Laptop; NP8890 / P4 / 3.4ghz with ht / 1 gb 400ddr ram / 800 mhz fsb / 2 internal 7200 60 gb hds / Echo indigo IO (WDM):

Latency -- w/out HT -- w/HT (meaning w/ multiprocessor checked)
46.4----------25%-------23%
23.2----------28%-------25%
11.6----------33%-------30%
5.8 ----------42%-------40%
2.9 ----------63%-------58%
1.5 ----------XX%-------96% and then craps out

Needless to say, I'm a little disappointed. If anyone has any tuning tips that really help latency, I'm all ears.

Actually those numbers look right on target for your platform. Damn impressive for a laptop.

anyone and everyone, how much does added ram assist one's computer in the sonar test. if you test with 512ram, then upgrade to 1gig what kind of percentage points can you expect to save, if any?

Finally got to this on a:

P4 3.0g/800mhz bus/HT disabled in BIOS/XP Home
Intel D865PERL MOBO
512mb Kingston RAM 300mhz
160gb Samsung EIDE hd/7200rpm/8mb/1 partition
ATI Radeon 9200 SE on AGP/8x
MOTU 828MKii using ASIO Ver. 3.0 driver

Results:
46.4ms.......29%
34.8ms.......29%
23.2ms.......30%
17.4ms.......33%
11.6ms.......36%
8.7ms.........40%
5.8ms.........48%
4.4ms.........56%
2.9ms.........75%
2.2ms.........Nope

At first I could not get below 4.4ms UNTIL I implemented Black Viper's Services settings as Internet Gateway. I will try again as Extreme Tweak when I can and update this info. EDIT: This can't happen until I offload Internet duty to another PC and dedicate this one to DAW-Duty.

Mind you, I can live with this performance. For now

Here's my system, and here are my results:

Intel P4 2.4 800FSB HT, no over clocking
512MB RAM
120G HD (7200 RPM
Asus P4P800 deluxe MOBO


46.4 - 35%
23.2 - 38%
11.6 - 42%
5.8 - 57%
2.9 - 83%
1.5 - Warning....


2 buffers in queue, BTW,

Tony

Did I forget souncard?

Echo Mia with 6.08 drivers. Both ASIO And WDM produced similar results.

DOH!

Tony

ORIGINAL: sammyp

anyone and everyone, how much does added ram assist one's computer in the sonar test. if you test with 512ram, then upgrade to 1gig what kind of percentage points can you expect to save, if any?

No speed will be gained. Task manager shows Sonar needing only 81MB RAM for that project.

RAM timing is where it's at. If you're using budget CL3 RAM than an upgrade to CL2 will give you a point or two.

Tonester,
I think that may be why your 2.9ms score is a bit low. Have you looked at your BIOS and tweaked your memory timings as fast as possible?

I'm not to fond of Tom's product reviews, but this seems to be a good article on tweaking RAM timings.

i tried to download the test file and run it.. it says theres no audio data for it.. it's just a cakewalk bundle file (141 k bytes), am I doing something wrong?

i tried to download the test file and run it.. it says theres no audio data for it.. it's just a cakewalk bundle file (141 k bytes), am I doing something wrong?

Just load the project, make sure the audio engine is engaged, and note the CPU reading in Sonar.

Sonar3Test input monitors several tracks through plugins. The idea is to isolate the CPU, system bus, and memory. There is no audio data coming from a hard drive.

-S

2.8 P4 w/HT w/800 FSB, 512 DDR RAM 2700, 120 GB WD 7200 RPM(int), 120 GB WD 7200 RPM firewire ext. drive.. M-Audio Omni Studio w/delta 66, Sonar 3.1.1

LATENCY:------CPU:--------------HT:on/off
1.5-------------warning/dropout--warning/dropout
2.9 -------------56-58%-----------warning/dropout
5.8--------------36-37%-----------39-40%
11.6-------------26%---------------28%
23.2-------------21%---------------23%
46.6-------------20%---------------21%

I'm a Newbie looking for a few tips. I just installed Sonar onto my new dual Xeon system, and I would like to perform this test before sealing it up. I have a Delta 44 soundcard. When I open Audio Options I can't get the latency down lower than 5.8. The lowest I can get the buffer is 2. Is this where I am supposed to change the latency settings for the test??

I find that I have to change the buffer value, click OK, then OK again on the Options window, then go back into Options - Audio to see the change.

ORIGINAL: wordeye

I'm a Newbie looking for a few tips. I just installed Sonar onto my new dual Xeon system, and I would like to perform this test before sealing it up. I have a Delta 44 soundcard. When I open Audio Options I can't get the latency down lower than 5.8. The lowest I can get the buffer is 2. Is this where I am supposed to change the latency settings for the test??

Wordeye,

To get the latency setting down to 1.5 ms, you'll need to decrease the DMA buffer size to 64 samples in the Delta panel (Hardware settings tab.) Are you using ASIO drivers for the test?

Eagerly awaiting your DP Xeon results...

I'm a Newbie looking for a few tips. I just installed Sonar onto my new dual Xeon system, and I would like to perform this test before sealing it up. I have a Delta 44 soundcard. When I open Audio Options I can't get the latency down lower than 5.8. The lowest I can get the buffer is 2. Is this where I am supposed to change the latency settings for the test??

Using WDM there's 2 places you need to look. First in the M-Audio Delta control panel set your DMA Buffer Size down to 64 samples (in ASIO mode that will automatically set Sonar to 1.5ms).

Then open Sonar and re-profile your card (I'm not sure if that's required but it doesn't hurt).

Then you can slide the latency slider down to 1.5ms and everywhere in-between.

ORIGINAL: wogg

I'm a Newbie looking for a few tips. I just installed Sonar onto my new dual Xeon system, and I would like to perform this test before sealing it up. I have a Delta 44 soundcard. When I open Audio Options I can't get the latency down lower than 5.8. The lowest I can get the buffer is 2. Is this where I am supposed to change the latency settings for the test??

Using WDM there's 2 places you need to look. First in the M-Audio Delta control panel set your DMA Buffer Size down to 64 samples (in ASIO mode that will automatically set Sonar to 1.5ms).

Then open Sonar and re-profile your card (I'm not sure if that's required but it doesn't hurt).

Then you can slide the latency slider down to 1.5ms and everywhere in-between.

With ASIO drivers, there's no need to re-profile the card. And you should use the ASIO drivers if you can. It's specified in the benchmark to keep things apples-to-apples.

Here is what I got:

Without MP selected
46.4......34
23.2......37
11.6......43
5.8....... 56
2.9........82
1.5........Out


With MP selected
46.4......20
23.2......22
11.6......27
5.8........35
2.9........53
1.5........90/Out

I am using Asus PC-DL mother, dual 2.4 Xeons. 1 gig ram, 2 74gig Raptor Sata drives configured in Raid0. I think I will reconfigure, and eliminate the Raid and compare the numbers. Either way, I assume that with benchmarks, this system will always score about the same as any SINGLE processor 2.4 Gig system with a 533 FSB.

Finally built my new PC:

ASUS A7N8X-E Deluxe
Athlon XP 3200+ Barton
1 GB Corsair PC3200 CAS 2.5 Ram (Dual-Channel)
Matrox G450 Video
Seagate 80GB, 8MB cache (OS and Apps)
Maxtor 120GB, 8MB cache (Audio)
Antec 430W True Power PS
Windows 2000 Pro SP4
Aardvark 24/96 v7.13 drivers ASIO
Sonar PE 3.1.1

46.4ms.................28%
23.2ms................ 31%
11.6ms.................36%
5.8ms...................46-47%
2.9ms...................64-67%


There was no noticeable improvement with Aardvark's ASIO "turbo mode" enabled (Started with Sonar I/O buffer at 128, upped to 256, then 512, with only a tiny difference of a point here or there).

Interestingly, I also ran the test with NVidia's onboard soundcard, using Wuschel's ASIO4ALL drivers and got very nearly the same results.

JD

FIC VC19E Mobo
2.6ghz 533FSB P4 Intel Chipset
1 gig 400 DDR RAM
Xp Pro /TerraTec EWS88D ASIO 24 bit
WD 120gig 7200rpm
Raptor 40gig 10000rpm SATA
IBM Deskstar 80 gig 7200

46.4ms 16%
32.9ms 17%
23.2ms 18%
7.6ms 28-32%
4.4ms 35-40%
2.9ms 46-54%
1.5ms Nope!

But I wonder if there was audio if the system would playback with pops/clicks?

ORIGINAL: WireRage

FIC VC19E Mobo
2.6ghz 533FSB P4 Intel Chipset
1 gig 400 DDR RAM
Xp Pro /TerraTec EWS88D ASIO 24 bit
WD 120gig 7200rpm
Raptor 40gig 10000rpm SATA
IBM Deskstar 80 gig 7200

46.4ms 16%
32.9ms 17%
23.2ms 18%
7.6ms 28-32%
4.4ms 35-40%
2.9ms 46-54%
1.5ms Nope!

But I wonder if there was audio if the system would playback with pops/clicks?

Holy crap! WireRage, are your system specs listed correctly?

These numbers are better than the P4 3.2, and on par with the 3.2 Extreme Edition!


JD

SuperMicro X5DAL-G
Dual 2.4Ghz Xeon
2 gig 266 DDR RAM (Crucial)
Radeon 9000
Echo Audio Mona
XP Pro SP1
Seagate 7200rpm (ST340014A)
Seagate Ultra160 SCSI RAID
IBM DPSS-33695N SCSI Wide
WD 200gig 8mb cache (WD2000JB)
Adpatec ASC-39320 RAID Controller


ASIO - w/o MP
46.4 32
23.2 34-35
11.6 36-37
5.8 43-46
2.9 55-57
1.5 79-83

ASIO - with MP
46.4 20
23.2 21-22
11.6 23-24
5.8 28-30
2.9 38-40
1.5 58-61

Hi JonD & WireRage !

JonD is right. WireRage - there is sonething definetly *not quite right* with your numbers.

To explain, I run a 2.6gig 800mhz P4 which is O/Clocked to 3.315gig stable - it runs at 13 x 255fsb - memory is DDR 500 running at 1:1 - I use a Delta 66 with the latest .42 drivers and I get as follows:-

1.5ms - 90% - stable & no dropouts
2.9ms - 56% - stable & no dropouts
5.8ms - 39% - stable & no dropouts
11.6ms - 30% - stable & no dropouts

I would suggest that the Terratec Drivers WireRage's EMS88D are reporting *funny* numbers *PLUS* it appears that WireRage has done his testing at 24 bit.

Timboe

I would suggest that the Terratec Drivers WireRage's EMS88D are reporting *funny* numbers *PLUS* it appears that WireRage has done his testing at 24 bit.

The test file is 24bit. It should be run with all of Sonar's audio settings configured for that.

As for the odd numbers... it looks like he may not have S3 Producer edition... and thus may be missing some plugins.

-S

*dang * .... the test file *is* 24 bit - dont know why I missed that.

As you say Scott, its probably a case of not all plugins / synths being installed and running in the test file.

Timboe

Okay, thanks to everyone's help here, I just built a machine whose price and performance so far I'm very happy with.

Athlon64 3200+ ($279)

Chaintech VNF3-250 mobo w/ the new nForce3 250 Chipset ($77) (I would have got the Gigabyte K8NS Pro, but it was not in stock anywhere, and the chaintech one has been getting pretty kickass reviews on the nerd forums for a little more than half the price, but a little less functionality).

1 Gb Corsair ValueSelect RAM, PC3200 DDR400, ($180) (matched kit of two 512 sticks)

Win XP home ($90)

Floppy drive from Staples ($12)

Antec case w/ 350 temperature-controlled power supply and rubber-gasketed drive bays for quiet ($66, including shipping)

Already had a dual-monitor video card and monitors.

Already had two fast ATA hard drives and a DVD R/W.

Already had cables, mouse, keyboard.

Call it yep's 80/20 rule DAW.

So all told, for about $700, I got a system that can handle practically as many plugins as I can think of using (so far) at the lowest possible latency without coming close to the redline on the CPU. Pretty good stuff, and I tip my hat to you fine folk for the help. For another $200-$300, I could've bought disk space, mouse/keyboard, cables, and a dual-monitor video card (CompUSA currently has a pretty good one for $50, after rebates), and come away with a pretty killer DAW for less than a grand.

Here's the funny thing- I tried the sonar test, and I cannot get my CPU meter to go below 18%, nor above 44%, no matter how I adjust the latency/buffer settings. Like, the 18% is with about TEN SECONDS of latency- doesn't matter. In almost the entire useable range I'm averaging about 35%.

This is all very preliminary- I'm not hooked up to my regular soundcard(s)-- so far, this is all with a Tascam US-428 that I have in my room near the interweb, and the tascam is USB 1.0, so the latency slider bottoms out at 5.8 ms. this is not an "audio engine won't engage" thing, it's just that 5.8ms is what the setting is when slid all the way down when this soundcard is hooked up. I will post more when I can check it out with the Delta or aardvark.

Cheers, and thanks again.

Which K*W are you using? 2885 or 2875?

Thanks,