Posting songs on the net

[RobinG]

This is not really a mando question at all, but hey..

We are a trio - guitar, mando, fiddle & three voices and we record our practices on a laptop with a cheap microphone. When I normalise the files and put on a bit of reverb it sounds great through a hifi, but awful on a computer: we want to put some of the songs on our Myspace site, but are worried that to most casual listeners we will sound clunky and unfunky, rather than like our real rocking selves..

I know that the pros test out their mixes on a variety of speakers, but does anyone know how to 'myspace optimize' a one track recording?
Thanks & Best
Robin

[tnpathfinder]

This is a great question! I'm looking forward to the replies.

[steve V. johnson]

Make your mixes, make mp3's of them and then listen to the mp3s on another computer, not the one with your recording and editing
software. #The smaller and worse the speakers on the second machine the better. # Listen on one machine, tweak on the other.

Start with high-pass and low-pass filters. # A high-pass filter lets frequencies above a designated "corner frequency" pass, while it attenuates (turns down) frequencies below that point. #How much those frequencies are turned down is often expressed as a number of db's per octave. #If you have recording/editing software that shows you a graph of your EQs, you can see how this works. # A 6db/octave slope is pretty gentle, while a 24db/octave slope comes closer to a 90º corner. #

Start with the lows. #Acoustic guitars in standard tuning and uncapo'd will make nice low tones down to around 150Hz. #Start with a high-
pass filter at around 150Hz and a 12db/octave setting. #12db/octave will probably let you hear how the filter works and let you move either way, more steeply or more gently. #If you want to hear 'radical' start right off with a 24db/octave setting. # Move the corner frequency up and down, listen and mess with the curves.

You can make mixes of the tune (or just part of a tune, for quick listening, but be sure to pick a spot that has some guitar down there) and listen to them on the 2nd machine, thru the junk spkrs. #

Low frequencies carry and use a lot of energy, and they can keep junk speakers very busy, cluttering up the ability to truly reproduce the
rest of the frequency spectrum. #If you can keep the unnecessary low energy out of your webmix, more's the better. #You can find a setting
that lets the guitar come thru and sound whole (if not "good").

Then repeat the process using a low-pass filter, for trimming the high end. #I'll suggest that you start at 5kHz, which is probably a good bit
too low, but by starting there, you'll get a real good idea of what the filter can do. #Move it up and down, mess with the steepness of the curve, make
some mixes, listen on both machines. #You have the routine now. # I know, it seems tedious, but once you've been thru it and are familiar with the
filters, you won't need to take nearly so much time at it, you'll know where to -start- at least.

Some other small EQ cuts are used by mastering engineers for this stuff. # Many rectangular drywall rooms (the most common rooms in America) have
a 'sound' that comes up somewhere in the range of 300Hz - 500Hz. # Try a parametric EQ (that is, one where you can set the frequency to cut or boost, the width, or "Q" of the cut/boost, and the amount to cut or boost), set it for ... oh... a Q of 1.4, a cut of 6db (which is kind of a lot, but again, this is so
you're certain to hear it working) and move it around between 300Hz and 500Hz. # My guess is that you'll hear some more clarity with the cut somewhere in that range. # Be careful, tho, because fiddles generate a lot of information around 500Hz. # Experiment with the width of the cut, and the amount. #You should be able to do ok with no more than a 4db cut. #I find I usually cut up to 3db...

Other places to look are in the range of 800 - 1.6kHz (even tho that's where the voices are, try it anyway...), around 2kHz and in the range of 3.5kHz on up. # Even tho your mandolin and fiddle, and the timbre of the pick on the guitar are all up there, you might be surprised at how some small trims up there can add focus and leave out frequencies that can really clutter up the sound.

I'd suggest that you not use any limiting or compression on your mixes between the main mix and the mp3 version. # This may be controversial, but MP3 is -data- compression and I feel that it does result in some of the same effect as (rudimentary) audio compression. #For this sort of music, I'd rather hear natural dynamic, to the extent that mp3s will allow that. #

Likewise, I don't normalize files for mp3s. #What I like to do, if I want to add gain or make several tunes consistent in level, is to gain them up to -3db below digital zero. #Ok... sometimes I'll go as high as -2db, especially for just one or two instruments or a soft ballad. # I confess. # I've found that this
lets things sound more natural, the dynamics of the playing come thru better. #And folks can always just -turn it up- if they wanna.

There's no easy, one-two thing that I know of for making mp3s, and the stuff I've suggested can be useful for regular mixing, too. #In that case, I do this
EQ stuff before any limiting or compression. #

The most important part is to use this stuff as suggested -starting points- and to experiment, and most importantly, to LISTEN and experiement. #The more fearless you are with the experiments, the easier it will get to do this stuff.

Have some fun!

stv

[RobinG]

wow - plenty of homework
Thanks for a commprehensive answer!
Robin