UnOfficial Protools Performance Certification

HIGHLY IMPORTANT DISCLAIMER

Do not run these tests with your speakers or headphones turned on. This is a completely fictitious scenario and the automation will produce results that will clip, cause resonance and can potentially blow your speakers at high volumes. So please remember to close all audio devices (i.e. speakers & headphones). Consider yourselves WARNED and I will not be held responsible legally in any way shape or form if you blow your headphones or speakers as a result of running these tests. Please do not opt not to take this test because of this warning. I had to apply some heavy automation and this situation cannot be avoided. You can actually listen to the sound but just at low volume cause I go crazy with all the automation sliders. Please do not take this disclaimer as a drawback to the test. I've spent nearly 8 hours cutting it based on comments from the community and ShanJ of course. Please try it. :)

Background

This was done fast so please excuse any problems type-o's etc.. I started this test thread (back in 2000), after moving over to Protools (aka PT) 5 from Cubase 3 VST, to basically discover which PC/MAC performed the best with PT. As a result I discovered that my own PC DAW (rig) was performing quite poorly :). So I quested forth and have gone through 3 iterations (built myself to specifications) of DAW’s along the way to finally hit a rig that can at least give me good enough bang for the buck to produce, engineer, master both my own and my clients compositions. This test was designed with PT6cs9 and I will have an update for PT7 shortly. Users of PT7 and 002 may go ahead and run these tests nonetheless. I highly suspect not many people will make it to Test-7 at this point. I'd like to see an Allenstein DAW take this test :) This test does take a bit of your time ( actually it's Protools that takes time loading the sessions); but at least it's better than manually creating the test :). Here it is..

Specifications

Over the years and evolution of both PT and PC/MAC technologies the original davec test became obsolete and received it’s fair share of complaints and cudos along the way. Well this time around, the test will truly exercise both machine and software alike. The following outlines the specifications of this new test. The test is setup to run for 60 seconds of audio recording and playback. I set the PC/MAC compatibility mode on all the session files so everyone should be able to try these tests :-).

Protools Setup Davec I (PT 5.x) Diagram

Protools is setup with 24bit 44.1Khz sampling (maximum possible). The automation memory is set to 200k. The buffer size is set to 1024 (maximum possible). CPU utilization is set to 85% (maximum possible) and leave the DAE Playback Buffer to default (i.e. Level-2).

Running The Test

To start download session file below and unzip into a new directory on your audio drive.

1) Open the Davec I session file
2) With your mouse select the created track.
3) Press Alt-Shift-D (to duplicate the track)
4) Record enable the track.
5) Repeat step 3 until your CPU goes into the red.
6) If your CPU is in the RED delete the last track created. IF it's still in the red, delete another track until the CPU is back in the green. Click on the CPU meter after every deletion to reset the meter (see warning below). Going into the RED means your getting a -9128 message from PT.
7) Hit RECORD, and record for 60 sec. (it doesn't matter what you record as the file still get's created).

If your CPU goes into the RED then delete a track and repeat step 7 until you can record 60 secs. without peaking.

*** N.B. If you can achieve the full 24 tracks with record enabled and can record for 60 secs. start adding Aux tracks with the same plug-ins and keep going until you hit RED.

Protools Setup Davec II (6.x and above) Diagram

Protools is setup with 24bit 48Khz sampling (maximum possible). The automation memory is set to 20Mb (maximum possible). The buffer size is set to 1024 (maximum possible. CPU utilization is set to 85% (maximum possible) and leave the DAE Playback Buffer to default (i.e. Level-2). Thx to ShanJ for pointing this out.

Tone Generator

People have mentioned over the years that recording silence isn’t necessarily a good example of a typical studio session. To compensate for this little oversight I’ve added an AUX track (in Bypass mode by default -- un-bypass when ready to test) with a single tone generator which feeds all the Audio and AUX tracks alike. This way when you hit record, you’re actually recording something. The signal is set at 20000Hz and -1.7 db (just to remove clipping) which should generate a solid tone to record audio. Again each track in the session file is fed this signal through a bus.

4-Band EQ

Full automation of all controls on the EQ are constantly shifting from one extreme to the other at an approximate rate of 1 extreme to another per second of playback

Short Delay, Medium Delay, & Long Delay

The delay automations are set for Master Bypass, Mix, Delay, Depth, and Rate. These too are occurring from one extreme to the other at about once per second of playback.

Volume, Pan, & Mute

All three of these controls are set to move again from one extreme or on/off as the case may be at about once per second.

Installing The Test

For your convenience (and to save you 5 hours of cutting) Click on the link below to download the test sessions in ZIP format. I enabled PC/MAC compatibility and a few MAC users have verfied running the sessions on their rigs. The size of the file is 8Mb compressed.

With the same file explorer unzip the test pack into the newly created directory (double click the ZIP’ed file).

Loading The Test

Turn your speakers/headphones way down or better yet off before loading the tests. Open Protools and select session Test-1 to start. Enable the Tone Generator plug in if it's not already. Also, I did not save the sessions with the system usage window open, so you'll have to do this yourselves until my next release :( Sorry for the inconvenience. Follow the steps (i.e. levels) in sequence, level-1 first, level-2 then level-3. You must complete all three levels per test session to qualify your results. These tests are intensive since we now have rigs with quad core quad CPU capabilty which will fly through these tests. Don't get discouraged if you don't succeed in getting through all 7 tests. I doubt anyone without either Hyper-Threading enabled or multiple CPU's will make it that far. DO NOT expect any of the numbers you saw in previous posts from other people. This is a completely different test designed to fully exploit all the PT features at an intensive level which is what a regression test should do. :-)

Level-1

Record enable all audio tracks and run for 60 sec's. If you peak then start deleting tracks 1 by one (from right to left) until you can record 60 sec's.

Level-2

This is to test the automation features enabled on all tracks. Once you've recorded 60 sec's. hit the play back and let run for 60 seconds. If you peak start deleting tracks again (from right to left).

Level-3

Once Level-2 is achieved, bounce your recording to 24 bit 48KHz WAV. If you can bounce then post your results if not, again start deleting tracks from right to left (Poeple who've achieved Level-2 encountered problems with Level-3 so please continue on to Level-3 in your tests -- after all what's the use recording 32 tracks if you can't bounce them at the end.). If you pass Test-1 without deleting tracks proceed to Test-2 and so fourth until you can go no further. Once all Levels for a given test are completed this is your final davec score to post in the DUC.

Don't forget to delete the created audio files in the Audio directory once you're done with the test. They can grow pretty big and eat up you're HD. Simply <Click> on the "Delete_Audio.bat" file in your test folder after every test. This will automagically delete the audio files. :)

Posting Your Results Performance Chart

These results are for Protools version 6.x and above as I do not have 7.0 and 002 yet I can't add the extra 32 voice channels (coming soon though)... although I suspect few people will make it to Test-7 for the moment. I sure would like to hear from you if you did!!!

Where To Post

Post your results here. Please do not start posting all over the place as this will make it difficult for newbies and old farts (like myself) alike to research and follow what’s going on with who and what :)

Seeing People's New Results

This test has only been launched since 2005-11-26 so give it time to start coming in. To see results people are posting for this new test please go here.

PT Rig: (PT version and gear model i.e..001 or 002 or MBox)

OS: (98 or XP or MAC or other like NT, 2000 etc.)

CPU: (How many/Class/Speed)

Laptop: (yes/no)

MOBO: (Make and Model)

Front Side Bus: (Speed)

Hyper Threading: (On or Off)

Memory: (Amount & Speed)

Drives: (how many, type & speed)

Graphics Card: (Make and Model)

Tracks: (audio+auxiliary)

E.G. - Tip: Copy the following table to your clipboard and paste into the results post. Don't forget to modify for your results! :)

PT Rig: PTv6cs9 and 001

OS: Windows XP Home

CPU: 1x P4 3 GHz

Laptop: No

MOBO: ASUS P4P800 SE

Front Side Bus: 800MHz

Hyper Threading: OFF

Memory: 1Gb 800Mhz

Drives: 1 IDE Boot 7200RPM, 1 Seagate SCSI-2 9600 RPM

Graphics Card: ATI Radeon 64Mb

Tracks 32+128

Links To Best PC/MAC/Laptop Configurations On The DUC

In Closing

I’d like to thank everyone for supporting these tests and inspiring me to continue on with this. I would especially like to thank ShanJ for his moral and technical support (he helped me catch-up on a few years of information :). Finally I would also like to thank Eric Day (Digidesign Software Engineer) who corroborated on the initial davec test. I hope everyone enjoys and above all Good Luck! We all want to hear from you… For those of you curious about my setup you can view it Here

Davec I & Davec II Tests

Protools 5.x (Davec I) & 6.x (Davec II) Regression Test Sessions

Help Support This Site

Running The Test Scenarios