Adventures in Hi-Res Recording
By Rich Tozzoli
When Is 96 kHz Appropriate?
“We started hi-res recording when the Digidesign 192 I/O first came out, so we’ve been doing it for several years now,” says engineer and mixer Mark Dobson. Working with producer Matt Serletic since 1996, Dobson has had a string of platinum successes with artists ranging from Santana and Aerosmith to Willie Nelson.
Dobson and Serletic weighed the pros and cons of 96 kHz versus 192 kHz. For recording acoustic instruments especially, 192 kHz seemed to deliver important benefits. “When you use 192 kHz on instruments like a solo piano you can hear the tail end really nice. At 96, the decay is a bit faster,” Dobson says.
While working at 192 kHz offered some sonic advantages, some of Dobson and Serletic’s system components had trouble keeping up. “When we tracked at 192, our FireWire drives didn’t work that well, and the system was on SCSI, so it was tough,” says Dobson. Having to split files between several drives also presented organizational challenges. “We did a Matchbox 20 record entirely at 192 kHz, and it was tough keeping track of files on multiple drives.”
Though these challenges would have been surmountable, Dobson and Serletic realized that for many types of projects they were working on, 96 kHz provided all the resolution they needed. “As we added instruments to the mix, the difference [between 192 kHz and 96 kHz] was less noticeable. From then on, we’ve just done everything at 96 kHz, and it’s been fabulous.”
Let’s say you’re committed at least to checking out high-resolution audio. What do you need to do to get ready? The challenges Dobson and Serletic faced working at 192 kHz are good lessons for anyone working in hi-res. Working at 96 kHz or higher sample rates requires some careful component planning.
The first challenge with 96 kHz recording concerns file size. A 48 kHz/24-bit recording uses 8 MB a minute per track. File sizes are twice as large at 96 kHz/24-bit, up to 16 MB a minute. If you’re tracking at 192 kHz/24-bit, it’s a whopping 32 MB a minute. To work at these higher sample rates, you’ll need larger hard drives — probably several of them — to get the job done.
You’ll also need faster drives. Drives need to read and write the additional information contained in those hi-res audio files while still keeping up with the session. For working at 96 kHz, Dobson discovered that he needs to use 7,200 rpm hard drives to maintain a decent track count per drive. “We just use a song per drive,” he says, “and we’re able to get a very large voice count on each drive.”
Of course, the more edits you have, the harder each drive has to work to access the audio. “If you have lot of edits,” warns Dobson, “you need to consolidate them or you won’t get the same track count. So we track on one drive, and when it comes time to mix, I just consolidate all the 96 kHz files, and they play back fine.”
The speed at which the disk spins is just one part of the puzzle. The speed at which the disk drive communicates with the host computer is another.
Today’s external 7,200 rpm drives can come in several flavors, each with its own transfer rate. USB 2.0 can transfer data at up to 480 Mb (Megabits)/second; FireWire 400 transfers at up to 400 Mb/sec; and FireWire 800 transfers at up to 800 Mb/sec. Of course, not all computers support each type of technology. What’s more, the actual transfer rate that a host computer delivers can vary too according to a number of factors — but let’s not even go there.
So, what type of external hard drive is right for hi-res work? Once again, the rule of thumb is that speed is a good thing.
I recently tracked a live 14-piece band at 96 kHz with engineer Paul Antonell at his Clubhouse Studios in Rhinebeck, NY. When it came time to mix, we had some issues with some of the drives. However, when we set up the host computer with a 7,200 rpm FireWire 800 drive, and consolidated the tracks, it was flawless. If your computer can handle it, I recommend a FireWire 800 drive for its better transfer rate — although I’ve had no problems with FireWire 400 either.
The Right Chips
96 kHz Plug-ins
Once you’ve got audio recorded at 96 kHz, you face another potential challenge: plug-in support. Even if you’ve got a collection of plug-ins that can handle the extended frequency range, you’ll need the power (either in your host computer or in a Pro Tools|HD system) to do the processing required for 96 kHz audio.
I talked to Dan Gillespie, DSP Engineer at Eventide, about using plug-ins at the higher sample rate. “You need two times the processing power, and memory, to work at 96 kHz,” says Gillespie. “With every addition or multiplication, you have to do twice as much math.”
Do the plug-ins actually sound better at 96 kHz versus the lower 44.1 and 48 kHz sample rates? “Certain filters and EQ designs will run more cleanly at 96 kHz versus 48 kHz,” Gillespie says. “And things like distortion and aggressive compression will sound better. With other things, such as low-pass filters and some EQs, it doesn’t matter as much. Overall, each plug-in has to be optimized to work at 96 kHz, and the code has to run more efficiently.”
Backing Up Files
When you’re done mixing your 96 kHz session with those 96 kHz–optimized plug-ins, you’ll want to back up your work. Because the size of individual audio files is larger with hi-res audio, the overall size of each session is larger. As a result, backing up sessions can be more complicated and time consuming. “For backups, we take each one-song 120 GB drive and use it as the safety for another song,” Dobson notes. “So each drive has two songs: the master of one song, and the safety of another. If we have 16 songs, we have 16 drives. It works out well. Then we get a terabyte drive and back everything up to that, so we have every song in three places. The typical song is usually around 40 GB, so it takes time.”
Is It Worth It?
Given the potential challenges in buying the right computer equipment, recording and mixing at 96 kHz, and then backing up your sessions: Is it all worth it?
Most of the professionals I spoke with answer the question of whether 96 kHz is worth it with a resounding “yes.” And I agree. Working with a high-resolution source expands your options. Sure, a large percentage of projects will end up as Red Book 16-bit/44.1 kHz CDs, or even as (gasp) MP3s, but many of us have also released projects on DVD-A, hi-res DTS and Dolby Digital DVDs, and SACDs. With even higher-resolution Blu-Ray and HD-DVD formats on the horizon, it’s good to be prepared.
I’ve had several projects mastered by Dave Glasser at Airshow Mastering, where he has sent high-resolution 96 kHz PCM files through DSD converters. The multi-channel DSD files are then used to create SACDs, which sound simply amazing on playback. It’s a pleasure (and a rarity) to hear at home what I heard coming out of the speakers at the mix studio. Even the 16-bit/44.1 kHz CD mixes have warm, beautiful highs and lows — certainly due in part to the original 96 kHz audio captures.
Don’t forget that consumers are buying “Home Theater In a Box” systems by the millions, and hi-res content (especially high-quality surround sound) is becoming more common. That’s not to say all the consumer confusion issues over “what plays what” has been settled, but we’re heading in the right direction. You can play your part by providing good 96 kHz content.
In the end, of course, the deciding factor is your own ears. If it sounds better to you at 96 kHz, well, then it is better!