This is the Sixth instalment of the Virtual Studio and as I mentioned last time I’m going to go into detail about a couple aspects that are specific to digital recording. How the digital domain changes our use of EQ and compression.
This is something I have heard very little discussion about and I welcome any input from other engineers and musicians in this area. As I have probably alluded to in the past, I first started recording in the days before the all digital gear was available and so I adapted my engineering practices to the new way of working. One of the reasons I want to talk about this is because most of the definitive works on recording technique come from a reference point that goes back to analogue tape and tube mics etc.
There are many “rules of thumb” around for how you process vocals for instance and for drums etc. but what I don’t hear talked about is how different the typical commercial studio of the “old days” is from a modern all digital system and what this means when we start twisting knobs. As wonderful in so many ways as the new technology is, there are things about the analogue tape machines and consoles found in the big studios that make them sound the way they do and subtle things that we take for granted and find pleasing.
You have to look at what is going on when a sound runs through the typical analogue signal path before anyone even starts twisting knobs. What is often overlooked is how much compression occurs naturally in this scenario. In a typical commercial studio with an analogue 24 track machine you have a signal path something like this; First a valve (tube) condensor mic picks up the singer’s voice and then the signal goes to a mic pre-amp which could be either tube or solid state amplified. From there it goes into a compressor and then possibly through a de-essor and maybe an EQ and then to the tape machine. Some of these components could be part of the mixing desk’s input module (namely the mic-pre and EQ). Once the signal goes into the tape machine it gets routed through a whole new array of circuits which print a signal to tape and then retrieve that signal via a playback head. So, along that maze of circuits and devices the signal is potentially compressed in 6 stages or more in the process of tracking and playing it back. This varies obviously with the type of gear and how it is used but basically the first one is the mic itself and this will remain the same in our digital world as well. Next is the mic pre and between the two there could potentially be a considerable amount of compression when you’re talking about a tube mic and a preamp run at a hot enough level.
If we used the same mic and preamp in the digital environment we would benefit from this compression effect as well, but when compared to a dynamic mic that is driving a mixer or mic-pre at low levels the difference is significant. Back in our commercial studio the signal is also compressed in the process of recording and playing back from analogue tape, not to mention all the gymnastics the tape machine circuitry has to do to accomplish this not-so-trivial feat. The bottom line is that the analogue gear may have been designed with the ultimate goal in mind of not colouring the signal any more than humanly possible but in the process, the things that happen to the signal along the way end up making it sound good even before any knobs are turned so-to-speak.
When compared to the digital world here’s where we stand; we may still have the nice effects of a good tube mic and preamp but once the signal is digitised there is no more compression of the dynamics unless we process it on purpose. So, the old “rule of thumb” about how much compression to add is not going to be valid when we take away the big analogue console, the analogue tape and the tube mic and preamp. This means that we will actually need to use more compression to compensate for what the analogue gear used to do naturally.
The problem with simply cranking the compression plugin a few more clicks is that it will not duplicate the subtle way in which analogue gear applies compression and will become noticeable in an unattractive way when too much is used. Keep in mind that the type of music you do will also be a factor. In a very sparse acoustic tune, for example, a lead vocal that’s way up front in the mix may not need as much compression as it would in the dense mix of a loud rock song. These things are also subject to personal taste but the bottom line is that a vocal should never jump out or dip down noticeably in a mix and it will sound wrong no matter what kind of music it is.
Another factor that is especially relevant in recording vocals is the perception of loudness. Sometimes a certain vowel sound or the harmonics in a singer’s voice will make one word sound much louder than another even though the actual VU level is the same. This is sometimes desirable when you want to emphasise a certain phrase or word but you still need to correct the levels so that it doesn’t jump out too much. This type of thing can really only be done by riding the levels in the mixing process or taking that word or phrase and adjusting the volume of the actual samples in the track.
Fortunately there are a couple cool plugins that help in controlling the dynamics without adding too many undesirable side effects. One of these is the tape simulation made by Cakewalk. It attempts to emulate the effect of recording and playing back on analogue tape and can be anything from a very subtle dynamic and EQ change to very distorted saturation effects. I find it very useful in applying just a touch of compression and brightness when used very lightly and the way it does this is much more subtle and pleasing then the same amount of standard compression. Just be careful because too much of the tape saturation will start to make the signal very aggressive and then distorted (albeit in a very controlled overdriven type sound as opposed to anything really nasty).
For basic compression there are many plugins that emulate a hardware device but in general I find them much more touchy and unforgiving. That’s why I always compress vocals as they are recorded using (at the present time) a DBX 164 and an Art Levelar. I find that the two are very different and compliment each other well. I end up with a very tame signal and no ugly “pumping” effect etc. I also like the Alesis 3630, especially for the price. If you have the big bucks there are many high-end tube compressors out now worth checking out. The reason the tube trend is popular is because people started to realise the really nice aspects of the way tubes do their thing. The big buzzword is “warmth” and what it really means is the right blend of harmonics that are added to the signal – yes we are deliberately distorting and colouring the sound but it’s these harmonics that the old analogue gear produced that we’re now trying to get back into the music.
I think we’ll continue to see software emulating analogue gear and I’m very curious to see just how close it will come to the real thing.
The other thing I wanted to discuss which ties in with the subject of analogue vs digital is EQ. The by-product of all the wire and circuitry and etc in the analogue studios in the area of EQ was that, in general, you lose some high end and gain various amounts of different harmonics. The tape itself had a limited bandwidth and highs usually suffered the most so it was common to boost highs on many types of sounds, usually at the mixing stage. What I find now is that in the digital world we simply don’t need to boost highs, almost never. If the signal is clean and sounds correct going in, it will come out like that as well. One of my first instincts was to boost the highs a bit on vocals and sometimes on drums and my first all-digital recordings were too bright as a result. So the good news is that while compression is a bit trickier in some ways, EQ has become easier in general owing to the increased signal-to-noise ratio and clarity of digital.
So that about does it for this time. There’s a lot more potential discussion about this issue but I just wanted to touch on the major points and start some thinking along these lines. Until next time . . .
Go to Part V of this article
Go to Part VII of this article
Digital recording column author Stewart Meredith has worked on the road and in the studio with the likes of Leon Russell and played on sessions in Nashville, Houston and Los Angeles. He was a content developer and beta tester for inovative software company Hotz Interactive in Los Angeles and has worked as assistant engineer in studios as well as session singer and keyboard player/programmer. He is available on a limited basis for freelance consulting in the London area.