As If Key-Based Notch Filtering Wasn't Crazy Enough, Here's Another Tip

First of all, to save everyone time and bandwidth writing posts about how this can't possibly work, this is a really stupid idea. There are plenty of other ways to accomplish the same result, and really, it all comes down to writing a better chorus anyway. No one in their right mind would sit down with a calculator and pretend to do things that have musical relevance. In fact, I feel kind of dirty and guilty just posting this, so I apologize in advance and accept full responsibility for thinking up something this dumb. So, there's nothing to see here; kindly move along. Thank you.
 
 
 
 
[Okay, now only the geeks and the terminally curious are left...welcome, have a seat...get out your calculators, you'll need them.]

So I was working on this song that had a shaker and tambourine part, panned left and right of center. The rest of the mix was fairly dense, and it seemed like I just couldn't get a good mix with the percussion. If I mixed it low, the two parts got lost and the song lost some drive. If I mixed them high, they stuck out and didn't integrate well with the rest of the track. I could not find a "split the difference" setting where they had enough level to drive the song but didn't sound overbearing.
 
I tried a bunch of tactics, most involving dynamics (more level while being less spiky - close) and EQ. Unlike the notch filtering, what I wanted was for the parts to be a little more prominent, not less. I thought what the heck, so I reversed the notch filtering to create slight peaks. That didn't work.
 
To make a long story short and not dwell on the other failed attempts, I added a Sonitus delay and set the delays to correspond to the song's key by calculating the period. The two delay lines were tuned an octave apart. For example if 3.0 ms produced the right pitch, the other delay was set to 6.0 ms. I didn't apply any feedback.
 
Mixing the delayed sound with the dry imparted a sense of pitch and with a very subtle mix, the percussion parts blended in with the melodic instruments better. It was easier to find a middle ground between too low and too high. I also found I needed to do this only with the shaker, but this also helped differentiate it a bit more from the tambourine so the two parts were more distinct in the mix.
 
Again, like the notch thing, this is not a panacea but another tool that at least in this case, solved a problem. I haven't tried it on anything else yet, but assume it's applicable mostly to unpitched sounds unless you're into some kind of special effect. It might work well with soft cymbal hits or something...but as with my other tips, I'm putting this out there before I get too far along to see if the Cakewalk Forum Brain Trust comes up with any novel applications. 

It's definitely audible, even without feedback - it's basically a comb filter. The delayed sound mixed with the straight sound produces the sense of pitch. 3 ms is the period for a 333.33 Hz wave - close enough to E (329.63 Hz) that it sounds in tune. So the pitch is basically the same as the E at the 12th fret of a guitar's high E string, and the octave lower delay is pitched like an open high E string. Making the pitch any lower didn't sound very good for what I was doing.
 
Feedback does have its uses if you want to make the effect more obvious. However, this isn't really an electronically oriented song, it's heavy on vocals and relatively clean guitars, so adding feedback sounded more out of place - like a flanger stuck at one spot on its LFO cycle.
 
I'll post an example of a song that uses this technique, I'm finishing the mix either tonight or tomorrow. I ended up applying it to another percussion part as well (not the tambourine, a synthesized one that sounds more like a small cajon). The resonator effect is subtle, but it makes the mix work better.

I have a bunch of shakers and a handful of tambourines and encourage people to pick instruments that seem to sound good while they are actually playing their parts.
 
I wouldn't bring it up, but seeing how you took the time to lay down the dirty dare-n-sneer in the intro I figure someone should stick their neck out and offer an antidote to mixer fixer work flow preoccupations.
 
all the best,
:-)
 
 
P.S. Just about anyone, if they don't want to waste time driving down to a store and auditioning what is for sale on the shelf, can make a cool collection of shakers with stuff around the house. Personally, I enjoy checking out shakers at Guitar Center and Sam Ash.  Each one seems different and unique.
 
For tambourines, it's good to have a few brass and a steel versions laying around as they all hit the mix differently and if you have choices you'll find it's easy to hear a stinker and easy to hear when one seems just right.
 
 
 
 
edit spelling

 
 

Paul P

Anderton
 
It's only a 1.5MB file. Two percussion parts are playing simultaneously. The first half is without the resonator, the second half with.

I haven't used headphones, but I can't hear a difference happening anywhere in the file.
Should there be a second file ?
 

I just downloaded it to check, and the second half indeed sounds different from the first half...I'm listening over the 5" Rockits connected to my laptop. It's a subtle difference, but I'm sure you'll hear it over headphones. Of course turning up the feedback makes it more obvious, but the demo was to show that you don't need feedback to hear a difference. Maybe I should have left the wet/dry mix at 50% to make the effect more obvious, but the setting was what I used in the song. FWIW I can hear the difference in the mix if I bypass.
 
i also think it's the kind of thing that once you hear it, you ears are trained to hear it. Like with the bus Console Emulator...when I first tried it, I couldn't hear any difference whether it was bypassed or not. But after using it for a while, I "learned" what the sound was and can now tell the difference, regardless of whether I can see the bypass button status.

It seems to have added a little length, which seems the most noticeable change.
 
Thanks for the tip,
Max Arwood

The second version (with effect) does have more life.  Nice job.
 
In both versions, the sound on the right sounds odd to me, but, of course, I am hearing it completely out of context. 

Good tip.. I like my Egg Shakers. Sounds real cool with stereo mike set up and you move them in and out. I sometimes hold 2 in my hand. 

Mike - what that paper doesn't take into account is that in this case, the sense of pitch does not depend solely to the delay. It is due to the delay being mixed with the dry signal. The resulting comb filtering provides the sense of pitch. The delay is just an accomplice. In this application feedback changes only the timbre, not the pitch.

Anderton
...the geek factor! I don't feel so alone...

Don't worry about that. For some of us, the science associated with music is almost as fascinating as the music itself.

Kev999
 
Don't worry about that. For some of us, the science associated with music is almost as fascinating as the music itself.

Almost ?  I could spend the rest of my life learning about the science and it would offer me an excuse for the fact that I struggle to produce anything musical.
 
In the last few days I've exploited the (very unfair, in my mind) access I have to scientific journals, being a returned-after-many-decades university student.  MIT's Computer Music Journal and Cambridge's Organized Sound have been my most recent targets for massive downloading.
 
 
 

Good ears, Jeff! Although some of the timing variability is due to swing, the part on the right was about 9 ms late on average. I fixed that but left the swing intact, and it works fine now. Thanks!
 
BTW I tried different keys and ones that related to the song's key worked the best. If I did something like A# for a song in A it didn't sound as good. Fifths weren't bad.

Craig,
 
Interesting that you used the delay technique.  You may be aware that this technique was used frequently back in the day in the studio (some very famous records have this technique employed).  They did it a bit differently and it involved two VARI SPEED tape decks.  I know that this has a scientific name of some psycho acoustic (the name I could not tell you) effect but in order to pull this off, you would make a physical loop of tape (about 15 to 30 feet) and route it around the control room (sort of like a giant ECHO PLEX cartridge). The reason for the large loop was so you could just turn this on during mixing and not have to worry about the decks running out of tape from a reel.
 
Turn the tape decks (giant echoplex with great audio) on, and then send tracks into this delayed loop buss and use the varispeed to find the sweet spot (the speed of the tape determined the amount of delay between the record and repro heads).
 
Difference was that almost everything would do through these tape loops (each of them set to different delay times), the returns would go to a DELAY RETURN channel and then to the reverb buss pre fader (the faders for the returns would stay off) so that the only audio  from this loop would be sent to the reverb buss (in essence, the delayed signal would only be heard through the reverb buss).
 
This then created a magical sense of space and much against one's common sense, more clarity in the mix.  The difference was that the delays were set to the tempo of the song as to mask its effect under other tracks that were more prominent in the mix.
 
Jim
 
PS - we did this by ear and not science and I still use this technique in almost all projects I work on but instead of 2 reel to reels, I use 2 instances of echo boy on tape sim mode and use the sync to tempo for the delay time (usually 1/32 and 1/16 notes) with the 2 delays panned hard right and hard left to the stereo reverb buss.

If I use your metaphor, yes, except that one used their ears to determine the length of the delays.  And then the reverb buss fader would create more or less AIR and SPACE on the mix depending on taste.  Sometimes you can hear a bit of the delays when their were beat pauses on the main tracks.  I can DEMO this for you with a WEB BASED stem mixer I have on my site if interested (let's you play with the volumes).  The reverb stem has the results of this technique and you can play around with its volume and the volume of the main stems.  Let me know and I will give you a link to check it out.
 
Jim