Anderton
The idea is to subtract the fundamental and its even harmonics only, 

I have pointed out that the rhythm track in a song in the key of A Major is almost as likely, at any moment, to have a fundamental, or undertone, at 146.8Hz or 164.8Hz as it is to be rooted at a multiple of 110Hz,
 
If you think about it long enough, it explains why a low shelf cut seems to be a nice way to make some space for other instruments.
 
The technique described is something I tried back when I was a school boy, but I soon learned that arranging the song more effectively and working with guitarists (or pianists) that play with a sensitivity for the tune yields the best results. I also learned that thoughtful mic placement on acoustic guitars is very useful, and that tone knobs were placed on electric guitars for a reason.
 
 
 
 
 
An idea that is somewhat related, that I have been fascinated with, is the use of impulse responses that are constructed of chord tones and have been identified as being matched with song keys for use in convolution processors. There was a fellow over in Africa preparing samples of these a few years back and he offered some freebies that I thought provided an ability to glue a song together in a sweet harmony.
 
At some point I realized that the results I heard were more or less the same sound musicians have, for the past few centuries, created by including a soft pad in their arrangements. The experience of using the song key impulse responses reinforced my appreciation of very quiet pads and the way they can make everything played on top seem to glue together in harmony. 
 
At some other point I imagined that the technique of using song key impulse responses, as I have described, could be hybridized into a system of using distinct impulse responses, each prepared for a specific chord voicing. I imagined a system where the IRs would be swapped out as the song moved through the harmony. When I was done dreaming about that it occurred to me that this is, more or less, what a really good reverb already does when it reacts to the music injected into the system.
 
Good times!
 
best regards,
mike
 
 
edit grammar

I love this forum and all the cool (and wacky) ideas that are proposed, but any technique that involves me doing math falls into the "taking the fun out of music" category.  I don't doubt it could work, but just too much fiddling!  

mike_mccue
 
I have pointed out that the rhythm track in a song in the key of A Major is almost as likely, at any moment, to have a fundamental, or undertone, at 146.8Hz or 164.8Hz as it is to be rooted at a multiple of 110Hz,
 
If you think about it long enough, it explains why a low shelf cut seems to be a nice way to make some space for other instruments.

 
But although a low shelf solve problems in terms of lower-frequency mud, it doesn't solve the problem of midrange instruments interfering with each other. The genesis of this experiment was rhythm guitar parts taking up too much sonic space, and realizing that a Nashville tuning - which shift the overall harmonic structure higher - often solved the problem. One thing led to another...
 

The technique described is something I tried back when I was a school boy, but I soon learned that arranging the song more effectively and working with guitarists (or pianists) that play with a sensitivity for the tune yields the best results. I also learned that thoughtful mic placement on acoustic guitars is very useful, and that tone knobs were placed on electric guitars for a reason.

 
But maybe you want a full voicing on the guitar, and you want all those "key" notes. I'm differentiating here between a full arrangement and a full frequency spectrum. Mic placement, arranging, and tone knobs on guitars (which of course do only high cuts) are all useful and part of the toolkit, but so is this.
 
Think of this approach as simply a variation on the technique used by so many mastering engineers who do a broad, shallow cut around 300-400 Hz, except applied to an individual instrument or two instead of en masse.
 
In practice, this works. Even if you don't think of it in terms of frequency response, being able to have a higher average level without dynamics is also useful.
 

An idea that is somewhat related, that I have been fascinated with, is the use of impulse responses that are constructed of chord tones and have been identified as being matched with song keys for use in convolution processors. There was a fellow over in Africa preparing samples of these a few years back and he offered some freebies that I thought provided an ability to glue a song together in a sweet harmony.

Yes, this kind of thinking can be very cool. I often add resonances at specific pitches and put them way in the background on parts like shaker and tambourine. It gives a sense of pitch that can really help blend into a track, if that's what you want (which of course isn't a given - sometimes the percussion is there to stand out).

sven450
I love this forum and all the cool (and wacky) ideas that are proposed, but any technique that involves me doing math falls into the "taking the fun out of music" category.  I don't doubt it could work, but just too much fiddling!  

I understand, but there are charts for pitch vs. frequency on the web. So all you have to do is read numbers, and enter them. I didn't want to have to sit there with a calculator, either 

sven450
I love this forum and all the cool (and wacky) ideas that are proposed, but any technique that involves me doing math falls into the "taking the fun out of music" category.  I don't doubt it could work, but just too much fiddling!  

I think I'm with you on this Sven.  I do like the ability to program and layer.  Technology allows for some awesome adjustments.  However, too much automation can be too much of a good thing.  Alas, I agree:  Leave the fiddling to the fiddler.
 
Fresh is Best,
-Erin

This is very interesting because it is easy to get build-ups in certain areas (e.g., around 300, around 2K), but you can't cut too many things too much.
 
Since I do not know how to record a song that isn't driven at least in part by an acoustic guitar , this could be very valuable.
 
Thanks, Craig.

I'm a big fan of carving EQ to mitigate masking, but ...
 
If the song is in A, and you notch every frequency in that scale and thereby lower every note the guitar is likely playing, how is that different from just turning down the guitar?
 

bitflipper
I'm a big fan of carving EQ to mitigate masking, but ...
 
If the song is in A, and you notch every frequency in that scale and thereby lower every note the guitar is likely playing, how is that different from just turning down the guitar?
 

As mentioned in the first post, I started off with six notches using the sharpest Q possible at the following frequencies: 110, 220, 440, 880, 1760, and 3520. I've since changed that to do only the first four frequencies, so no more notch at 1760 and 1520. Because the notches are so sharp, and because I have control over the notch depth, this can have a subtle effect if that's needed. If I want to do something that's more blatant than this but not as blatant as turning down the guitar, then I just widen the Q a bit.
 
This is the kind of thing that's somewhat difficult to imagine mentally, but if you have a rhythm guitar, vocalist, and piano all playing at the same time and apply this to either the guitar or piano, you'll definitely hear the results. Of course whether those results are desirable or not depends on what you want out of the mix.

"This is the kind of thing that's somewhat difficult to imagine mentally, but if you have a rhythm guitar, vocalist, and piano all playing at the same time and apply this to either the guitar or piano, you'll definitely hear the results."
 
 
It doesn't seem hard to imagine, nor does it seem hard to imagine that you can hear a difference.
 
 
 
However, It just doesn't seem like a "musical" solution. Turning down the fundamental and under and overtones of one of the 7 commonly used chords, the tonic or root chord, in a song key is a seemingly arbitrary choice and it doesn't seem sensitive to the idea that the frequencies chosen only appear in the voicings of a few of the other chords in a song key. 
 
 
It seems to me that stringing a guitar up with Nashville tuning is a much more "musical" solution, and writing parts with chord voicing that compliments a song isn't a bad idea either.
 
best regards,
Mike
 
 

scook
MMultiAnalyzer is on sale for ~$40 for the next couple of days.

Thanks for that scook. I have been wanting that one for a while. And I had forgotten to keep checking up on the rotating sale.

I got it now!!