Return to EQ index

All information in these pages is copyright (c) 1989-2003 by Roger Nichols. All rights reserved. Permission for personal reference only, and may not be reproduced by any method without written permission.


Good Choices, Bad Choices
by Roger Nichols


ProTools HD
Last night I got to see a demo of the new Pro Tools HD system. Gill from Digidesign had it running on a Titanium G4 laptop with a Magma 4 slot expansion chassis connected to a 192k interface and a sync I/O box.


Whether or not you think higher sample rates make any difference, there are some changes in the new Pro Tools HD system that make it a highly desired upgrade to a Pro Tools|24 or Pro Tools |24 Mix system.


1) For the same amount of horsepower, the system costs about half of what a previous system cost. Each PCI card that goes into the computer contains 9 DSP chips, instead of 6, and they run 25% faster than the previous DSPs. Previous boards contained 3 different types of DSP chips, so plug-in shuffling would happen if the plug-in need a specific type of DSP. Now all of the DSP chips are of the S-RAM type that will run any plug-in.


2) Each card will handle 32 channels of I/O instead of 16. The cable connecting the 192 I/O to the card can be up to 50 feet long for 192k operation, or 100 feet for 96k operation. Instead of using a “Y” cable connects two interface cables to 888 or other I/O boxes, one cable goes to the first I/O box and then a short jumper connects the second I/O box.


3) One I/O interface handles 16 channels of analog, AES, TDIF, or ADAT audio, at about the same price as an 888/24 I/O. Instead of six 888/24 I/O and three ADAT bridges, I now only need three 192 interfaces to accomplish the same task.
4) If you are still working with 48k or 44.1k audio you can have up to 128 tracks playing back at the same time, while at 96k the limit is 64 tracks and at 192k the limit is 24 tracks. Remember that 96k take twice as much storage while 192k eats up four times the amount of hard disk space for the same length of music as 48k requires.


Bang-for-the-buck honors have to get to Digidesign for a job well done on this new round of hardware. Good work!


Live Concert Recording
I just finished doing surround mixes for a live concert DVD. I thought I had nightmares before, but these attained a whole new level of terror. The video part of the six hour long show was professionally shot with multiple Betacams all receiving the same timecode and locked to house sync (from the video switcher.) The video editors had an easy job… but the audio…


The audio was recorded on four ADATs. A fifth ADAT was used to record the timecode, sort of. When the first set of ADAT tapes was recorded, the timecode was recorded on audio track one of the fifth ADAT. When the second set of ADAT tapes were put in, the timecode was recorded on audio track two of the fifth ADAT. When the ninth set of ADATs was recorded, the timecode was recorded on the first track of a new ADAT tape in the fifth machine. Got it so far?


Now. There was no sync reference between the video machines and the audiotapes. What should have happened was that the same video syncing all of the cameras should have been sent to a box that generated word clock or ADAT sync. This would have insured proper audio sync later.


A potential backup to solving the “no sync” problem would have been to resolve Pro Tools to the SMPTE time code on the fifth ADAT tape. The SMPTE was locked to the video and can be used as a clocking reference if needed (although not the best choice.) Here is where the nightmare gets worse.


All of the ADAT tapes had numerous dropouts. Audio tracks would disappear for long periods of time. This included the timecode tracks. Not one of the timecode tracks made it without four or five major 30 second or longer sections with missing timecode. Without continuous good timecode, you cannot use timecode as a clock reference.


If the ADATs had been recorded in sync to start with, then the digital audio would be in sync with the video, and any good portion of timecode could be used to trigger the start of the audio transfer from ADAT to Pro Tools. But not this time.


After the audio for the entire show was dumped into ProTools, about 15 edits for each song were needed to compensate for the drift of audio to picture. Luckily there was a ruff mix audio feed recorded on the switched video. This audio track was recorded into Pro Tools along with the video. This audio track was used as a guide to line up the multi-track audio. What a mess.


Not all of the six hours of show was recorded on the ADATs. One of the performances had the first two minutes missing. Some performances had no ADAT audio at all. All of the video edits were done based on the audio feed to the video truck. Nobody bothered to check to see what audio was missing before they video editing was done. In those cases, the only audio available was the audio feed to the video truck. A bad mix with feedback from the on-stage monitors. Try making a surround mix of that!


The moral of this story? Make sure when you are recording multi-track audio that you have two methods of assuring sync. 1) Word clock derived from video sync, such as you get from an Aardsync, and 2) timecode from the video truck that is continuous with no dropouts. If there is potential for dropouts, record the timecode on two tracks on different ADATs. The chances of both tracks dropping-out at the same time will be less likely. Nobody’s pants ever fell down if they were wearing a belt and suspenders. Sooner or later applying that principle to recording will definitely save the day.


Session Documentation
Once more I am going to harp on session documentation. Whether you are using Pro Tools, Logic Audio, Digital Performer, CuBase, or Deck, make sure you document what you are doing. Make sure you name the tracks before you start recording so that the audio files mean something. An audio file named Kick is easier to figure out than an audio file named Audio 01-01. If you are using external effects or processes like a reverb or tube compressor, document the type of device and the settings.


A good idea is to include a Read-Me file that describes the configuration of your Pro Tools or other DAW. If you are using Digital Performer, was the interface an 828, a 2496, or a 24i? Were the mixes done completely inside the DAW, or were sub-mixes done on an external console? If so, which console, what settings, and what effects were added in the external console.


The first time you try to work on a session sent to you by someone else, or try to recall a session that you did yourself three years ago on a previous version of your favorite software you will appreciate a comprehensive set of notes.


++++++++++++++++++++++++
Roger can usually be found on the corner in front of Jimmy Buffett’s Krispy Kreme donut shop selling used pencils to raise the funds for his new Pro Tools HD system.


Return to EQ index