This article originally appeared in issue #4 of Surround Professional in April 1999.

Surround sound brings a new level of complexity to the entire process of recording, mixing and mastering not normally found in stereo. With at least four extra channels, there's now that much more room for error. After all, if some people can still screw up stereo (switching left with right, one channel out of phase, etc.), imagine what can happen in surround. One thing's for sure though, master tape prep has become more critical and necessary than ever before. Here are some suggestions and guidelines to follow while mixing in order to avoid any unforeseen occurrences later down the line.

Channel Assignments

Sooner or later during one's first surround mix, the question of channel assignment on the master recorder (be it tape or hard disc) always arises. Just what is the correct track assignment? Actually there's always been several ways, with some good reasons behind each one. Let's start with the one that's quickly becoming the de facto standard, just to get things off on the right foot and cut through any confusion right away.

A dedicated stereo mix, or Lt/Rt, or encoded AC3 can be recorded onto Tracks 7 and 8.

This format is the SMPTE and ITU standard, as well as the assignment matrix suggested by Dolby. This method transfers easily to the corresponding four audio tracks (L, R, C, LFE) of the most widely used video formats today such as DigiBeta or D5. This is also recommended by Dolby as it is the common pairing of channels in Dolby Digital encoding (although the AC-3 encoder can actually be configured to any track configuration). The surround products of Panasonic, Mackie and Martinsound, to name just a few, now support this configuration as well.

The following two assignment methods are also used, but less and less as the above the SMPTE/ITU standard becomes more and more widespread.

This configuration method is what many film studios use, although it's seen in some music production as well. This seems to make sense in that it's a somewhat visual representation of the way the speakers are laid out, but falls short when it comes to logical track pairings.

The above is the channel assignments preferred by DTS. Again, the pairings are logical but the placement different from the SMPTE/ITU standard. Tracks 7 and 8 usually contain the stereo version of the mix, if one is needed.

SDDS (Sony Dynamic Digital Sound) is a special case in that it's a 7.1 format. SDDS uses a track assignment that differs from the norm, but again makes sense in that it gives you a visual representation of the way that the speakers are laid out.

There are obviously other assignment permutations that are occasionally used, but all seem to be falling quickly by the wayside as the SMPTE/ITU track assignment method takes hold.

Media Prep

It's imperative that as much information about your project as possible is documented. Many potential problems can be avoided as long as the master is prepped and the following items are noted. Being lackadaisical about these details is a sure invitation to trouble since they're all significant.

Slate The Master

More than ever before, it's important to not only properly document the master tape or disc, but to prep the master in order to be make sure that there's no questions as to the actual track assignments. We've all had instances where the documentation doesn't align with what's on tape and this can lead to disaster in surround. Plus, with surround you can never really be sure that what you're hearing is actually what's been documented. Even an engineer who's mixed the tracks sometimes has a hard time determining which is the center channel and which is the left surround, so it's quite necessary to take any potential guesswork out of the process.

The best way to avoid confusion is to go back to the admittedly low tech but foolproof method of using an audio slate on each channel indicating the channel assignment (i.e. "Channel One - Left Front", "Channel Six - Right Surround"). This only takes a minute and provides an ironclad certainty that the proper tracks find their way to the proper channels.

Print A Test Tone

Print at least 30 seconds of 1kHz tone at -20dBFS, which is the SMPTE standard reference level, across all tracks. A 1k tone is a pretty good way to discover if there's any clock discrepancies since the purity of the signal will suffer as a result of clicks and warbles which might not be heard during the actual program material.

Also keep in mind that any program on tape media should start no earlier than 2 minutes into the tape, since that's where most errors and dropouts usually occur.

Other Things That Should Be Documented

Proper documentation of a master tape means that there's never a question that can't be answered or a mistake made because of confusion or omission of data. Even though this has always been important, multi-channel masters have a new variety of possibilities to account for.

Q Is the LFE channel filtered and at what frequency?

This is important if for no other reason than it's easy to figure out which is the subwoofer channel if the assignment documentation is lost.

Q What is the reference level in SPL?

This means the reference level that you used while mixing, and helps the mastering engineer to better approximate what you were hearing during the mix if there's a problem down the line.

Q What is the sampling rate?

Soon there will be a number of sampling rates available to mix to (192, 176.2, 96, 88.2, 48, and 44.1kHz) so this helps to avoid any clock or sync issues that may arise during mastering or authoring.

Q What is the bit resolution?

Generally this will be 24, 20 or 16, but an indication is necessary in order to set dither correctly. Dither is a small noise signal of only a dB or so that's intentionally induced into the digital word in order to remove unused or unwanted bits at the end of the digital word (i.e. when a 24 bit program must be converted to a 20 bit program). To simply lop off the bits (truncate) at the end of the word sounds bad, so dither is used instead.

Q If time code is included, what is the format?

If the audio program is linked to picture, the necessary time code format is required in order to maintain sync between the audio and picture.

Q Are the surround channels calibrated equally to the front channels or -3dB?

In film style mixing, the surround channels are calibrated at -3dB. Music style mixing has the surrounds equal in level to the front speakers.

Q What is the media format and how many pieces are there?

The entire master may actually be on several pieces of media on several different formats. A media count here can eliminate the confusion of an incomplete mastering or authoring job later.

Q How long is the program?

This is necessary because it's one of the factors that determine if data compression must be used.

Q What is the intended final audio resolution (i.e. 96/24 or 48/20)?

Once again, a combination of sample rate and word length determines if data compression is used and how it is set.

Q Any glitches, distortion, dropouts or anything unusual?

Always a good idea to indicate on any master tape, this stops the mastering engineers from wasting time (and your money) checking his own equipment when a defect actual lies in the master.

Yes, multi-channel provides us with multi chances to screw things up. But if we just take the time to do some media prep and simple documentation, we can eliminate any problems before they happen.