Recording Tips
Compression Q&A
Q) I record
with Cakewalk... I have experimented with many different compressor
and EQ settings and still can't get a professional sound. I don't think
I fully understand compressor and EQ settings. Can you help me with
settings? -Shannon
Compression and limiting is simple, and highly under-explained. All
it is... is an automatic volume knob that turns down the signal a certain
amount, based on the settings. Commonly the idea is that louder portions
of a given signal can be reduced so that the overall level is more even.
This is helpful to (1) lower the level surges and (2) allow the softer
parts to become more even with the louder parts -- giving better consistency
to the sound.
The first setting is the *threshold* which is the point at which the
device works. Kinda like the threshold of a doorway is the point at
which you step into the other room - it's a line you cross over. So
before the threshold is reached, nothing happens to the signal. *After*
the threshold (after you step over the line), the device starts turning
the volume down.
How much it's turned down is determined by the compression ratio. 4:1
compression ratio means that for every four dB of signal that goes into
the unit, it turns it down so that only 1 dB comes out. 8:1 means that
for every 8 dB that's put in, an increase of 1 dB is what goes out.
Often we see a meter reading called "Gain Reduction" which is how much
the device is turning the signal down (that's what I prefer to look
at on the meters).
The attack time is when (how soon or how fast) the gain reduction unit
starts altering the sound. The release time is when (or how long until)
the unit stops altering the sound. The slower the attack time, the more
the initial signal gets through the unit without being altered. Slower
attack times let through more peak (and the immediately trailing) material
because it waits longer to start working.
Limiting (which is really just more compression) starts up at 10:1 and
goes up from there. Since limiting is really turning down the signal
a lot at once, it's faster and mainly used on peaks which are occasional,
not constant. Used properly, limiting should be hardly detectable because
it's controlling shorter amounts of sound.
Since 2:1 - 6:1 is a more subtle amount of alteration to the sound,
it can be used for more containment, for instance on vocals, bass, etc.
Stereo mixes, however, don't sound good contained. In mastering, I'm
very careful about keeping the openness to the sound of a mix, so I
highly recommend not compressing the stereo output of your console.
If your meters are swinging too far "into the red" check on individual
voices and instruments that need the containment.
Frankly, I'd leave off the limiting in mixdown too. If you put too much
compression or limiting on your mix, there's nothing we can do to remove
it if there's too much of it. Most of the time, we use limiting and
compression in different ways on different songs, so we prefer to have
a wider range of options to pick from. (It's not a bad idea to
try some stereo buss limiting to make reference CDs for evaluation of
what really happens to your sound on commercial systems -- particularly
to reference your music with commercial CDs. Loud-and-limited
commercial CDs have been carefully crafted, but there's much to learn
by doing it yourself and listening critically. ... just lessen or remove
that limiter when you're making your final mix... or stick with Separations
and allow the dynamics to speak out musically.)
The other reason that I don't recommend compressing the stereo mix is
because it sounds too good. What I mean is that when everything
in a mix comes forward a bit, you hear more of everything that you've
recorded. This can be good except when you start to reduce the
parts of the sound that need to snap forward. Same thing with
a Finalizer in mixdown. Since it sounds so good, it tends to make you
not work as hard at getting the mix to be totally exciting and punchy
as you want it. Working hard on the mix (within reason) adds a level
of excitement to the song. It adds a "performance" element, if you will,
to the mix process. When you put more energy into the mix, it somehow
comes out more on the CD.
I had great fun mixing through a Finalizer several times, thinking how
good it sounded, until I got these mixes in to the mastering room. Right
before my ears, what I thought was exciting and vibrant was more soggy
*within* the mix. The highs and lows were good, but it wasn't up to
the song's potential. Once again, I was reminded that another gadget
often does *not* take the place of solid basic music-based engineering.
Even riding faders is a very correct form of gain control, and
with your ears as the guide, it's a very musical form too. If you are
unsure as to whether or not to compress your stereo buss, then make
2 mixes, one with and one without. These two mix versions can be referenced
to the other songs on your album to see what works best.
|
The most flexible format to insure that your CD sounds amazing
is Separations.
|
Q) What type of EQ & comp. for vocals, guitars, bass and drums?
There just isn't a formula, because every musical element is unique
and deserves the listening treatment. Eq'ing is a another article in
itself. For now, as far as compression, I like 4:1 for vocals and bass,
maybe 5:1 for guitars, and generally I don't compress drums unless I'm
going for an effect, or it's just needed in the mix. (I do like gating
toms slightly.) If the drummer isn't consistent with the volume of his/her
playing, I attempt to iron that out at the tracking session, not the
mixdown. Sampling and replacing drum notes is an option, but ultimately,
it's more of a win for everybody if the player (the source) is held
to a higher standard.
As far as getting a professional sound, there is just a learning curve,
and every engineer goes through it. I can remember hearing records thinking,
"The high end sounds so darn much better than what I'm doing!" Then
I'd work and work on the high end, and think, "That record has much
better bottom than mine does!" Then I'd work and work on the low end,
and then I'd think, "Why can't I get the mids to sound like that record?"
And the list goes on. Striving for warmth, clarity, smoothness, width,
apparent loudness, mono compatibility, etc... it's just a process. Experience
is the only teacher that is one-on-one with you at every moment of your
career, and the frustration that goes with it is a valuable component
of this process, because it motivates you to *ask* the questions and
seek the answers.
Guideline: Generally, one answer comes up again and again. "Less
is more." Tweak as much as you can at the source, then progressively
less as you go up the next layer up the chain, and so forth. Therefore,
doing nothing at the stereo mix buss is right in line with this idea.
And...remember Rule #1 - there are no rules. Be as unique as you want
to be - don't rivet yourself into any one idea if another one will work
better. Sometimes mistakes even lead to cooler ideas. We all experience
the mistakes and frustration, so don't be hard on yourself at these
times. Just acknowledge your feelings, let them go, and enjoy the ride
along the way.
Q) I'm compressing my digital stereo mix with
an ART PRO VLA tube compressor before we go into the PC for mastering.
The results have actually been pretty amazing. Glancing at your article,
this might not be the best way to go. How can I suggest [what you advise]
to my engineer when he's the one with the experience and I'm not? If
it sounds good to me, should I stick with it, or will it really cause
that much trouble later for the duplication house? -Jeremy
There are two things to address: the technical situation, and the communication
situation. As far as the technical part, there are no rules! If it sounds
good, then that's a step in the right direction. An engineer I know
puts a tube compressor on his mixes just because it adds a sound he
likes, even if the compressor isn't working more than a half dB. But....
in some cases Studio Monitor Madness
could be occurring - the monitors could be masking some loss in dynamics.
When in doubt, just compress the appropriate tracks within the mix (like
bass, vocals, etc.) and let the mastering engineer do the stereo work.
The duplication house (CD replication plant) will only have problems
if you give them a CD with uncorrectable errors or a defective master
- they're not concerned with sounds as much as a cassette duplication
company (but a good cassette company will be able to handle whatever
you give them... it's just easier for them if it's well mastered).
In the tricky area of communicating a preference with your engineer,
you have to keep balance in the chemistry that you've developed with
your technical personnel. I use the "donut" theory - say something positive,
then introduce the challenging part, then say something positive again.
Also one of the best ways to diffuse the energy is to come right out
front and say, "I've gotten some ideas from an outside source that may
be different from what we've done together for a while. I'm a little
uncomfortable about bringing these things up, because I appreciate your
work, and I'm not into stepping on anybody's toes, especially yours.
I don't even know if these new ideas are good ones or not, but I'd like
to know if you'd be open to trying some things. They may be cool, they
may not be cool. I'd like your opinion, and for the sake of stretching
what's possible, I'd like to experiment with some changes. What do you
say?"
The "how can I suggest otherwise" could also go like this: "Fred (a.k.a.
your engineer), I'm loving the results we're getting so far. And as
a curious sound person, I've read an article by a mastering engineer
that discourages compression, although he doesn't state this as a hard
and fast rule. Given my budget, do we have any room to experiment with
leaving the stereo compression off the next tune to begin with, and
then doing a separate version with it in?"
Hopefully Fred's response would be to accommodate your request. The
thing to be careful of is that once the compressor is in, there could
be some level increase that seems to be enhancement. *Buyer Beware!*
The key is to listen to both the compressed mix and the uncompressed
mix *at the same level*. You may have to lower the compressed version
to really go side-by side, but when the volume level (use your ears,
not just meters) matches, then see if the difference is really better.
Often, I don't compress, I only limit the peaks to get the CD hot.
Important: While the particular song you're working on now may
sound good compressed, in the context of the *whole CD*, it may not
need as much compression, or it actually may need more. When I master
a project, I like to hear the context of the whole project... the entire
personality of the CD. One song influences the others... even the order
can have an affect on how much processing different songs will eventually
have. But if you've compressed it while you're in the studio... wrapped
up in the moment... and you find that it's tooooo much later on - guess
what? Tough luck (or go back and remix).
Another great advantage to professional mastering is that you can take
home the mastered CD with you and listen to it... compare it to other
CDS and on several systems. Then, if changes need to be made (like less
compression or more compression), then you can go back and in a relatively
short time, make the change you prefer... and have a better product
that you're happier with for years to come.
Q) I am interested in a compressor for mastering
my music but I do not know if it is worth it to spend more money on
an elaborate compressor like the Cranesong or the Summit just to get
only 10-15% better sound quality than if I buying a cheaper compressor
with similar or with slightly less specifications than from the "High
End" compressors.
Specifications don't always tell the tale. I tried a [highly rated]
compressor that many mastering engineers thought was great at that time.
I didn't care for [that] compressor, so I took it back and tried six
others. The one I picked wasn't the most expensive. In time, that [highly
rated] compressor wasn't used at other mastering studios either, and
the one I picked gained in popularity. Specifications are a good staring
place, particularly when it comes to being on a budget.
The music I compose is cinema/documentary, classical/ballet and is from
synthetic sounds.
Clean and quiet and musical will be the best, and the high end units
will sound better. More open and smooth. However there may only be a
10% difference. It depends on your customer and how much of a difference
it will make to them and to the buying public. If it's just too much
for what they will or won't hear, then I'd say save your money and look
to some better speakers (unless you're very very happy with them already).
I think great speakers, power amps and cables can make as much or more
difference.
I have some R & B songs that I want to mix down the music without vocal,
so I can sing them live. My uncompressed mixes sound low level and transient
using Digital Performer.
Hard disc recordings for the most part have a more "vacant" less solid
sound than analog. Look at the waveform and you'll see much spikier
looking peaks than the fat, more rounded analog version. If you cut
your drum tracks on analog and transfer to hard disc you'll get more
wallop on the bottom end.
What would be some good compression settings that I could apply to the
stereo mix here to go right out and sing live with it. When I've tried
to do this on my own, it takes some of the punch out of the music. -Jeffery
Punch comes from the excursion (distance of movement) of the speaker.
Too much compression usually results from the threshold being set too
low (acting too soon) and just causing a blend. What you really want
is more fatness in the peaks, and more apparent contrast in the dynamic
range (higher peaks, lower lows) but still keeping the track full-bodied.
A limiter might be the ticket, enabling you to bring the overall mix
up without curving down too much musical information. A 10:1 ratio or
more with a fast attack and release time would be a good staring place,
but you'll just need to experiment to match your exact musical program.
NOTE: In the May 2003 Mix Magazine issue (a must-have issue), there's
an article called "Stop Mousing Around" where they have about 6 paragraphs
stating "MONITOR SECTIONS: THE MISSING LINK." A high resolution monitor
controller can be a powerful tool to hear the distinctions necessary
when listening to compression.
Q) Do you have any suggestion for getting a
more transparent sound? -Quincy
Separations can add a significant
amount of transparency. Besides that, good electronics and clean mics
help for sure. I'm also a fan of high end power cords that clean up
the AC lines - this translates to better top end and tighter bottom.
Clean mixing helps too - don't have bad caps in the console, some transformers
can add too much coloration, etc. Don't bounce too many tracks digitally
unless you really need the processing power - there can be some dullness
and lack of zip that occurs from track bouncing.
Keep lots of headroom - don't press everything to the top. Don't have
your recording room be too dead - some diffusion and live surfaces are
good. Higher quality reverbs add a great deal of transparency that the
cheap ones just can't match.
Sometimes cutting with EQ at 400Hz 500Hz 600Hz or 700Hz (anywhere in
between) can help give a more pure sound. If stuff is a little boxy
or cloudy a peak EQ (dip in this case) cutting with a moderate Q can
help.
The arrangement of the instruments can help - be sure stuff isn't crowding
out other sounds, be sure parts are played tightly, and pan things so
they don't overlap too much. If your sound seems good in the studio
and fluffy in the car, your monitors could be too bright or you could
be adding too much from 32Hz and down in the bass or kick.
When you solo stuff, it should sound almost too clear to be correct,
but when it's all blended it, then it should sound correct. Avoid using
the same frequencies on too much stuff. Spread the timbre out evenly,
keep the bass clean and even, watch for ground loops or hum in your
system...
Other than that, just keep testing out different ideas - we all discover
much that way!
