Recording, Mixing or Mastering: Which Matters Most?
The Audiophiliac, aka Steve Guttenberg, interviewed mastering engineer Dave McNair and basically asked, “what’s the biggest factor in determining the overall sound of a recording?” over at cnet (you can read the article here). The answer to the question depends on a lot of factors. As a former (can I say recovering mastering engineer of 16 years) mastering engineer and currently a producer and engineer of new high-resolution audio recordings, I believe I have something to contribute to this topic. The original recording sessions are the most important stage in the recording process in establishing high fidelity…if that matters.
And it matters to me because I’ve been trying for 15 years to produce and release recordings that avoid the traditional recording pitfalls (as I see them…over processing, heavy use of compression etc). The same considerations may not be important to other recording engineers. Their job is to produce a recording that matches the sound of the current commercial hits. The actual fidelity of the tracks is a secondary concern.
Music used to be produced by musicians playing acoustic and electric instruments. All a recording engineer had to do was put up some microphones or plug in a direct box and route them via a bunch of microphones preamps to a recording machine. Decisions were made regarding mic placement, the application of compressors, equalizers, and other signal modifiers but the sound the came from the performer was captured on analog tape or a digital recorder of some sort. The fidelity of the music being played and recorded was locked down during the original sessions. There is no going back later (with some exceptions for NoNoise or very old recordings) and restoring the fidelity of a poorly recorded track. Whatever the signal to noise ratio present during those original takes persists through the rest of the production.
This applies to the overdub sessions during which additional instrumental and vocal parts are added to the basic rhythm tracks. In fact, if the sessions are being done on analog tape, each pass of the tape over the record and playback heads reduces the fidelity of the signals. A very small amount of the oxide is scrapped away each time.
Mixing engineers (which many times are not the same people as the recording engineers) spend additional hours tweaking the quality of each sound in the overall blend. Mixing engineers are not in the business of maximizing fidelity for an individual track. Their job is to bring all of the parts into an artistic blend through the use of volume, dynamics processing, spatial distribution, and equalization. The use of reverb, delays, and other specialized processors are widely used and actually tend to blur the clarity and dull the fidelity of a track. The fidelity of a track doesn’t get better during the mixing stage of production. The tracks are made punchier and less dynamic.
So what’s left for the poor mastering engineers? It used to be that if you’d done your mixing right, the mastering engineer would have nothing to do but put the tracks in the right sequence, enter the ISRC codes, rundown the album, and output it to DDP or SONY 1630 tape for the plant. Not anymore. Mastering engineers have new digital tools that guarantee that any remaining fidelity present in the mixes is smashed into an even amplitude plateau. That’s what the record labels want because that’s what the radio stations, YouTube, and Spotify wants. It’s the SOP (standard operating procedure) and is not likely to change.
So why bother with release formats that brag about 24-bits or even higher? There’s no reason to go to 24-bits because the output from the mastering studio uses much less than 16-bits.
The opportunity to establish the fidelity of a recording happens at the start. It’s the job of the recording engineer to get it right when the music is being played.
Hi Mark,
I was reading that Cookie Marenco is going to be doing a live recording session to demonstrate hi-res audio. She will be using the new Sony PCMD100 portable audio recorder. It can record PCM and DSD. The built-in microphones on that unit are only specified from 20hz-20khz. If she uses the built-in microphones, then is she really demonstrating hi-res recording at all? What is the point of this unit recording 96khz/24 bit pcm or 64fs DSD if the microphones can’t even take advantage of it? Thanks in advance for your reply.
I’ll probably head to the NAMM show tomorrow and check out Cookie’s panel. The PCMD100 unit is a very good portable recorder. It’s meant for documentary style recording…the blend of the sound in the room is what you’ll record. It couldn’t (and doesn’t qualify) for the hi-res logo…and they don’t show that on the website product page. I don’t know if she’s going to be using the built in microphones.
Absolutely no question about it; the quality of the original recording will dominate all post-processing and cut through all playback formats.Being PC here, let’s just say you can’t get 6 cents out of a nickel.
The priority on recording has the advantage that, with the mixing and mastering tending to be botched as you described, there is always hope that well made recordings can be re-mixed and re-mastered later, especially if time proves the music to be an important contribution.
There is always that hope…if the tapes and notes can be found. It’s even worse with the current “master” existing as a set of files. I have trouble locating my own final mixes.
If it is difficult to find the files on which the final mixes (recording masters?) have been assembled, then surely you, as a mirocosm of the recording industry, are showing by your lack of attention to archiving that you care little about what future listeners, with future technology, could potentially extract from today’s recordings.
This is nothing new of course – it was already happening in the analogue era when an edited master tape would sometimes be destroyed after the LP pressing master had been cut. (Example – Tommy by The Who.)
I do a pretty good job of archiving the 100 or so discs that we have produced. I have the original R-1 Exabyte data tapes…which are a real pain to restore but I still have a functioning system that some other high profile artists have benefited from. Then I have multiple hard drives of the multitracks and both hard drive and cloud based storage of the final mixes. There are occasional difficulties with an item or two…but I’ve never been stumped yet.
The question remains:
How do we know, if/when the music, that we buy, is treated like Mark just described?
Analyzing the files each time can’t be the solution.
Do we have to trust our ears – can we?
If is safe to assume that you will getting the over-processed, heavily mastered stuff in virtually every commercial release. High-end labels like 2L, MA, Chesky, Linn and my own AIX Records deliver the real deal. I can’t figure out a way to bring the two camps together. It takes too much money and clout.
“If is safe to assume that you will getting the over-processed, heavily mastered stuff in virtually every commercial release.”
It almost always been this way, something I learned 20+ years ago.
I started out in the late 60s-early 70s buying ever better equipment and for the most part got increasingly better sound. Then a point came sometime in the 80s that it seemed the better equipment I got the worse my music sounded, I couldn’t seem to make any real improvements.
Then I bought a few of my first “audiophile” albums, stuff from David Manley (ViTaL), Chesky, Telarc, etc and my jaw hit the ground. I had no idea that my system could sound so good, it was an enlightening but sad time combined, and started the slow death of an audiophile. I realized that as my equipment was getting better I was better able to hear into the commercial recordings the junk they were committing to tape. I”m buying the best playback electronics available while the recordings are being done on 128 (or more) track mixers containing hundreds of dirt cheap 50 cent op amps and all other kinds of bright, edgy, distorted components. After that my spending on my system slowed to just about a stop since there were just no more real improvements to be made listening to the music of my choice, my MoFi CD of Dark Side of the Moon sounded as good as I’d ever be able to get.
I did have a short glimpse of hope over the whole HDA thing till I once again woke up (with your help Mark) that converting old garbage analog recordings into a large bucket digital files wouldn’t make any improvements either. You can’t make a silk purse from a sows ear.
As I faced retirement and a downsizing of my whole life I sold off my 30K+ worth of high end gear, moved to a small place in FL, and replaced my system with a nice midrange 5.1 surround system. I ripped all my CD’s to my computer and did high quality recordings of my LP’s using audicity then sold all the cd and lps.
I pretty much cured myself of Audiophilia Nervosa and no long question if Stereophile really is right and that a $3,000.00 power cable really would lift layers and layers of veils from my music. LOL
The best part of being 65 is that I know nothing new is around the corner that will make the Frank Sinatra – All The Way CD that I bought yesterday sound any better than it does today.
Oversampling does improve impulse response & flatten high frequencies, and is also an integral part of attenuating filter adventure.