Mic
preamp buyers guide (Updated August 2010)
I've
heard in person a great variety of pre's on the
market, and compared them in good listening environments. Most judgment error
comes from failure to calibrate gain - when things are louder you will hear
them differently. Anyway, pre's are a huge topic so
I'm just going to cover it topically.
What
makes a good preamp? First, it’s the ability to provide transparent gain. If an amp can't
conform to the input wave form then that's distortion. Some distortion sounds
great, but I'll get to that later. First and foremost you want a thing that’s
transparent by virtue of simply making a small signal bigger. The adage is
"wire with gain". Getting this requires a well designed circuit and
solid components top to bottom. Simple is better. Step attenuators are better.
Every time you run through a component (esp. caps) you pervert the signal
slightly. Every time you run through a switch you create capacitance problems
or impedance mismatches as the wire diameter or conductivity changes. No one of
them is a problem but if you add up a lot of them you start to get a degraded
signal. Every time you run through an op
amp - of which there are many in a typical console channel - you compromise the
input. All commercial manufacturers are incented to
reduce cost so they buy as cheap as they can within the spec. Small builders
have so much labor in the equation that the cost of the parts hardly matters to
the net profit so they use better parts. Features complicate the signal - Roll offs, etc. Some
preamps use feedback loops which require hand matched components to get perfect
phase cancellation. Good makers sometimes
use relays so the signal does not pass through the switch. Good makers use step
gain control so the signal does not have to go through a pot. A great preamp is
a simple well made preamp. With a good
pre, you will hear the mic for what it is with
surprising ease. "Discrete" components, a big claim, are good because
it means that the designer got to select the quality of each capacitor and
transistor etc rather than relying on an integrated circuit. IC/Op amps can
sound great - but are not repairable when broken, and can only achieve a
certain level of headroom and conformance to the input. Many very high-end console preamps have op
amps in them, so "discrete" is more of a marketing spin than a
reliable indication of quality. However, if someone is going to take the time
to make a discrete circuit, you know they care about what they are doing.
Another
big brochure-word is "Jensen" transformers. These have been used in a
lot of audio gear, and are great transformers… but… they ARE transformers - all of which add some
distortion. Many big preamp makers brag about the sound of their transformers. Focusrite, for
example. Indeed, they sound great. But wait - its
because of their coloration, not simply because of the purity of their
amplification. Millennia makes a Mic pre in which you
can add a transformer wound to sound like a focusrite
transformer into the circuit, OR leave the path transformer-less. A preamp
circuit does not require a transformer. They are handy to the designer, and can
do cool stuff like enable balanced interfaces, but they are not required and
they do introduce distortion. Most
studios need balanced inputs simply to accommodate decent noise performance on
long wire runs, but in a home studio run unbalanced CAN be just as good. It all
depends on the extent to which ground potential differences exist among your
gear or ambient RFI runs around in the air in your environment. Many high-end
pres offer transformer-less paths out or in or both, and they sound more
transparent without the transformer. I've AB'd this
with several preamps and its true. The so called "focused" sound of
the transformers made by Neve and Focusrite
(and their many clones) sounds great - but can be approximated by adding eq - very carefully - The problem with adding up tracks
using a colored pre is that the sonic signature of that coloration takes over
the record. Same is true if you only use
one Mic for everything. To add a grain of salt here -
note that most great recordings were made with all identical preamps - in the
board - so adding-up preamp coloration is not necessarily bad - it might even
make things blend better - but….other factors have a lot more to do with
blending.
Tubes!? God bless 'em, are the marketers dream come true. Just when you
thought you had a great rig….
Once
again, this is about creating desirable distortion. I've read that tubes are
more linear than transistors and can therefore more accurately represent an
input signal. … But that all depends on what's around the tube circuit-wise. If
the signal is impaired even a little bit you won't notice this benefit in most
cases. Tubes will always have more noise - but very careful engineering can
limit this factor to a manageable level. One thing for sure, tubes are doing
you very little good unless they are in fact amplifying the signal. Many cheap
tube boxes just buffer the signal running the plates at low voltage. Make sure
any tube gear has high voltage rails. Basically if it can't kill you it
probably isn't any good. Like drugs. The bottom line
here is that expensive tube gear is killer, and cheap tube gear is just ok. I
personally don't buy the digital harshness argument any more. It wasn't bits
that were harsh, it was cheap A/D converters and 16
bit word size and newbie studio owners who made bad sounds on their new Roland
hard disk unit. Like, "Yes.. dude,
it does sound like shit". If you
have good converters and 24 bit samples it can sound as smooth as silk without
benefit of tubes.
I
know a guy who has almost every tube mic ever made,
an ssl, studer, pultec and lots of it, and perfectly maintained tube gear
of every description,
By
all means have some tube stuff, but buy quality (>$1000), and don't do
everything with tubes - the noise adds up.
Use them to take the edge off a harsh input sound, or fatten up a vocal
or something.
Tube
emulation does not sound like tubes. On the other hand, if you are simply
creating distortion why be picky about where it comes from? If it sounds good, do
it. Do not, however, rely on music store auditions. It never sounds the same in
the studio. Just buy the stuff and take it back when you are done. Retailers
get enough of my money - and since the average Mars sales person knows very
little about what they are selling from first hand experience - the testimony
of the sales people is tainted.
Some roles for channel
strips
Swing a dead cat….hit a channel strip. They're
everywhere!
Most engineers will put a very small amount of
limiting or compression on an input track if the dynamic range of the
instrument is too big. Generally compressing to "tape" is to be
avoided. Most good players know where to put their dynamics to fit the part
they are playing - but a lot don't. Thus having a compressor/limiter on the
input channel is handy. However, with today's digital gear you can record that
whole dynamic range and compress later. Its not about signal to noise ratio the way it used to be
with tape. On the other hand, saturating a tape can sound ok, but distorting a
digital input sounds like vomit. You can save a track with a limiter. Thus:
Have a good limiter. Cheap limiters sound cheap. They dull the sound when they
take out signal. Long attack times help - but if you are trying not to distort
your inputs, letting the attack through doesn't achieve that goal. If you have a limiter you hope will never hit
its threshold, you might as well lower your input gain another notch.
Most channel strips have compressors/limiters and/or
eq. The shorter signal path enabled by having all
this stuff inside a box is good. Signal path issues plague studios. Simple
signal path is good - it’s a long and winding road through cables and patch
panels and they all make a difference. On the other hand most top end boards
have compression and good eq in the channel - so that's
all handled in a good studio nicely. They HAVE channel strips. Big audio facilities have 48 + channels of
this costing several hundred thousand dollars.
So channel strips are an economic way for a small studio to have
world-class sound transcription. Most
channel strips make their various parts available separately via jacks too.
This is nice to since you can use the channel strip as several
"boxes" in your mix.
Options
in channel strips include low cuts, phantom power, phase invert, pad, etc. all
nice. Given the option of doing a low cut in a mic or
in the pre, I'll do it where it sounds best - and there's always a difference. Its nice to have a choice.
Phase invert is important when micing a single
source with multiple mics. Sometimes the part you
want will phase cancel - so putting them out of phase is will stop the phase
canceling. (XY mic pairs are good - huge difference
in phase cancellation. A lot of people
get fancy with multiple mics on a source - especially
strings - but simple is better if you can get away with it.
In
the super geek world of audiophiles there are "subjectivists" and
"objectivists". The former want it sound good and the later want it
to be correct. Point is, anybody who has spent one day
in a studio knows there is no correct. Correct is good, but it's not the
endgame, and its not achievable. The musicians will
happily contort their own sound for creative reasons. So, what was correct? How
was it supposed to sound? Never mind the jungle of other significant
impairments between the music and the audience.
The objectivists have a tough row to hoe.
About eq. You know what
it is and what it does. The watchout with eq is phase distortion. All filter networks distort the
phase of something other than the range being eq'd. It’s a mathematical certainty. Good ones do
it less. Bad ones do it more. Cuts do it less, boosts do it more. So: cut,
don't boost, and eq only when required. The reason
you match mics to sources is basically so you don't
have to use eq. Eq also
introduces noise or decreases gain through the system. It’s a crutch. The best producers know how a
sound should be in the final mix and record it that way to begin with.
On the other hand, its not
a perfect world, and not everybody can have a visionary producer. Thus, the
next best solution is to have high quality eq.. Basically eq is used for two sorts of things. One is to shape the
basic tone of a sound. For this, you use wide "Q"'s.
Preamp makers talk about "musical" eq. This
basically means a wide Q - typically an octave. Wide Q sounds more musical
because its very natural to hear this phenomenon in
real life. For example a wall behind a band. The other
use is to kill off a bad resonance, or accentuate a good one to give an
instrument character that will make it stand out in the mix. Its
common to engineers to sweep an eq to find a
"sweet spot". I do it
constantly. For the "surgical eq" you need
to do these last tasks, you need a sweep, and a narrow q. You can do miracles
with quality eq.
Good eq is not cheap and it’s a bad place to
compromise. Another reason for a good channel strip.
No eq will do 40 db in a 5
hz Q without significant
distortion. Thus, its hard to make instruments sound
too much like other instruments by using eq - but if
you have great eq you can do some pretty amazing
things. I made a guy's guitar sound like a lot like a Gretsch
one time with just severe eq - following the spectrum
of a Gretsch. He loved it.
Its this cheesy "close
approximation" or "modeling" which is at the root of "amp
modeling" or "mic modeling". All that
stuff will give you the tone of the thing to be modeled, but will never sound
like it. I hate that entire class of gear. It just never sounds right. If you
are a guitar player in love with your Line 6 amp, forgive me. I've made some
good tracks with these amps (POD actually), and they sound a lot like the real
thing. But they are not the real thing.
I don't understand why anyone would want a so-so imitation of the real
thing when the real thing can be had, in most cases. Seriously, the fake sounds
don't respond anything like the real thing. A Pod "tweed fender" does
not feel at all like a real tweed to the player - and
isn't the relationship between player and sound the essence of music?
Compressors. Who needs 'em? Most music gets seriously compressed by the
time anyone hears it. I suppose someday all sound will be the same volume! If you hire an expensive producer or engineer
they will likely show up with a rack of cool compressors.
These are powerful boxes… and they do more evil than
good in the wrong hands. The best minds in the audio business have delved deep
into this area, so I'm going to stop here.
Just remember this simple rule: Keep your ratios lower than you think
they need to be. Its easy to compress a thing more but
hard to undo the damage once the sound is committed to medium.
In the end a preamp is just a device to get the gain
of a thing from point a to point b - and vendor's
attempts to differentiate themselves with features does more harm than good.
What the world needs is wire with gain, cheap….and we'll fix all the other
stuff with good mic choices, mic
placement, and good musicians. (or.. harmony, melody,
and rhythm).