Support
Looper's Delight
|
Support |
|
|
Jeff,
again, a stack of unsorted wisdom (and street knowledge):
> Now for processing I usually sample a bit of noise and the do a noise
>reduction run on it.
Now it's time for me to talk with tongue in cheek: if you remove the
noise later, why do you add it in the first place? Reconstructive
surgery is something which can be done by a mastering engineer, but is
not what is the primary goal of mastering - and more importantly, it's
something which will leave scars, to stick with the metaphor.
Try to optimize your signal chain in the recording process to get rid
of unwanted noise. One action word here is "gain structure". Another
one is EQ. Noise is most of the time an issue in the high frequency
band. So if you have, say, a reverb effect and you can reduce the
reverb's high frequency content without affecting what you're after,
then please do so.
> I might also at this time adjust levels in the recording as needed.
Which may or may not be a good idea. There's level adjustment before
and after the compressor, and there are places where either one is
appropriate. If you do it, try it (in case your software allows for
it) to do it non-destructively, i.e. by using some "volume envelope"
feature, not by editing the audio file.
> I may or may not do some eq on it then using whatever. I am getting weak
>here. I try out all the other stuff but it's just at random and mostly
>don't like the results. Maybe a little eq works out. Ignorance is not
>bliss here.
A few ideas I'm throwing in (at the risk that you know them already):
* At all costs, use an audio app (I don't remember which you're using
- was it Ardour? or Audacity?) that allows you to use the effects
plugin-style and several of them at once. They will interact with each
other, so to find your optimum sound, you need to tweak them in
interaction.
* Rules of thumb for EQ in mastering (there are exceptions to these
rules, which I don't want to go into right now):
* destructive EQ (e.g. highpass at 35Hz or removing an unwanted
resonance with a high-q parametric) is the first step in the
processing chain. It's also the only EQ (for that removing unwanted
resonance) with high Q settings.
* use parametric EQ, not graphic
* the maximum allowed "Q" setting on your EQ is in the 0.7 range
(tranlates to: the minimum bandwidth setting is 1 1/2 octaves).
* an EQ setting in excess of +/-3dB means something is wrong in the
mix or source material. Go fix it there.
* Rules of thub for compression in mastering:
* the typical number of compressors in the chain ("compressor"
being the general term including limiters and clippers) is 2-3.
* the limiter (a compressor with infinite ratio and minimum attack
time) is always the last thing in the chain (safe for
dither/noiseshape - see below).
* the gain reduction of a limiter should not exceed 3dB
* the gain reduction of a clipper should not exceed .5dB
* a single compressor can be used to a) tame microdynamics (this is
usuall done during mixing), b) tame macrodynamics, but not the same
two things at once in an efficient manner.
* if you adjust your source level pre-compressor (either while
playing or in your mastering process), soft and loud passages will
sound vastly different (especially with a fast compressor).
* rule of thumb: microdynamics - short attack/release.
macrodynamics - long attack/release
* mastering compression time settings usually happen in the 100ms-3s
range.
* Level practices: when comparing the sound of two different processed
versions (or comparing your stuff to a master from another source),
make sure they have roughly the same loudness. First-oder rule of
thumb: RMS value. Second-order rule of thumb: K-System metering.
And another thing:
> It's that missing thing that I am trying to improve.
So let me ask "what is that thing you're trying to improve"? Take some
recordings you think sound good on any system, and where you think
"damn, that mastering job would sound good on my material!". Compare
it to your stuff, putting them on the same level (as mentioned before,
and also in private email). Try to take notes. The guitar sounds
edgier? It's less boomy? Etc.
Then, use those meter things. Frequency Analyzer, Meter (Peak and
RMS/K-System), Phase Scope. Compare both your and their master (best,
with audio turned off). Again take notes. Is it flatter/pronounced
areas (how wide in octaves)? What is the relationship between peaks
and RMS in different parts (soft/loud)? What is the relationship
between RMS values in soft/loud portions? etc.
Ok, I should start and do a web tutorial. Have you already checked out
Bob Katz's online tutorial (I didn't so far, but absolutely have to!)
Best,
Rainer
ps: did you ever check your favourite albums for the mastering
credits? I find that e.g. Howie Weinberg is responsible for a lot of
the really good stuff, with Bob Ludwig coming in at a close second.
And there's another nice quote from Gateway Mastering (Ludwig's
company): "mastering will most probably the highest per-hour in your
entire project". 'nuff said...