A Problem with The Mixer Control on My USB Audio Interface Made Me Think All Guitar Impulse Responses Sound Bad

For over a year I thought something was wrong with my guitar cab IRs when I played direct into my DAW. My setup is a bunch of pedals going into my audio interface like this:

IR Loader in DAW <-- USB Interface <-- High Gain Pedal <-- Boost Pedal <-- Tuner <-- Guitar

Yes, I play with no power amp, real or simulated, although I'm told a tube power amp is used to generate most guitar cab IRs, so maybe I get some tube sauce for free there.

Symptoms

• IRs sounded like they had too much room sound. As if they had a small ambient room reverb on them. Like how I'd imagine it would sound if you mic'd an amp from too far back. You do you, but for high gain I want my mic'd sound pretty dry.
• Sometimes my high gain tones sounded spitty and gross. I couldn't dial out the fizz and crackle without losing the whole top end. Sometimes it verged on unplayable.
  
• One week or one session things sounded great, the next week I thought I was crazy to like that sound. That's normal for me, where one day I think a tone is perfect and then next I dislike it, but this was nuts.
  
• I thought I was going crazy.
• I thought...maybe IRs are overhyped bullshit?
  
• I thought...maybe there is something wrong with my pedals?
• Then the other day I suddenly realized that I could not get the sound in my phones to sound like the sound that was recorded in the DAW. This should not be the case ever. I was supposed to be hearing 100% DAW playback sound from my interface, in both my phones and speakers - the same sound that is recorded in bits and bytes.
  

Cause

• The mixer knob on my audio interface had always let a tiny bit of direct sound into my phones/speakers, even when I had it turned all the way to "Playback". I knew this all along, but I didn't think about it, because I am a dumbass.
  
• The way I knew this was: with the guitar track record button turned off in the DAW, if I turned my phones up loud I'd hear a bit of high gain junk when I played. It sounded like an industrial guitar track. I figured, that's weird...but how much can it hurt?
  
• A lot. It can hurt a lot. It can make you think that IRs are stupid, when the entire rest of the world seems to like them a lot.
  
• This was especially apparent when I had my DAW session really quiet, which happens sometimes because I'm shit at gain staging, and I forget things. When that happens, I have to turn up my phones or speakers to compensate, and the proportion of direct sound to DAW playback sound would go up. It turns out that those were the really bad days.

Approaches

1. Live with it. Gain stage my projects so that the proportion of fizzy direct sound is as small as it can be compared to the project level. Probably a good idea anyway? You tell me, but probably.
   
2. Get a new audio interface and hope it didn't happen there. My current audio interface is very low end, and out of production, so I figured what are the chances this happens again? This probably would have been a good approach.
   
3. Hack up my current audio interface with a soldering iron.
   

Solution

I chose approach I nearly always take: hack shit up with a soldering iron. I don't recommend this unless you know what you're doing, and are ready to replace your hardware if you break it. I came so close to breaking mine.

1. I took the interface apart. Lots of screws holding in the mic/instrument combo jacks and MIDI connectors, nuts holding in the pots and 1/4" jacks, and a single riser board for the top set of pots. This part was easy and fun.
   
2. I located the stereo B10K pot, and discovered it is a standard voltage divider. If you don't know what one of those is, you can stop now. I sorta know what a voltage divider is. That gave me incorrect levels of confidence.
   
3. Turned it all the way up to "Playback" and found it was 1.6 ohms on one set of legs, and 10K on the other. Voltage divider? I decided it is. What else could it be, it's a mixer pot? Right? I'm probably wrong.
   
4. Got the pot off the board. What a nightmare. I am terrible at desoldering parts that have more than a couple of legs. My solder sucker is half broken, the tip is all melted and fuzzy, I can never get solder wick to work, my hands are shaky, and I'm just terrible at this. I ended up literally destroying the pot with a pair of nippers and pulling the chunks of plastic off the board, leaving only the legs. I was able to get the legs out by applying heat and pulling with the needle nose pliers. I mostly didn't scrape any board traces by accident. Mostly.
   
5. I soldered the 1.6ohm set of through holes to short them out using clipped off leads from an old resistor, and left the 10K stereo set open.
6. Apparently I didn't cut any traces or chip any important surface mount resistors.
   

Results

• Now, when I have my DAW record button off, and my monitor level all the way up, and playback is stopped, and I hit the guitar strings, I hear literally nothing in the phones or speaker. You obviously want to watch your ears doing this, and make sure you turn in back down fast.
  
• My IR setup sounds the way I imagined it should - very dry, no fake "room sound" effect.
• My guitar tone sounds the way I imagine it should - no industrial guitar fizz, unless I put it there on purpose.
  
• I feel stupid, but I'm recovering slowly.
• I didn't get an excuse to buy a new audio interface, which is a bummer because I am am a monkey that likes shiny new things, but good because the fun wears off on those even quicker than most gear.

Built my own vocoder in Reaper DAW - For No Reason

This is a seriously backwards project. It's built in Reaper using only stock plugins. Keep in mind the important fact that I am not a genius. There are probably a hundred better ways to do this just in Reaper, and there are probably 10,000 better ways to do it in an audio programming language. People build these with analog electronics, too, but those are pretty high on the parts count. I wanted to spend a nice day off on this, not a terrifying rest of my life.

An interesting but incorrect fact: the Wikipedia article on the Vocoder says it was invented to compress and more easily encrypt voice transmissions, which is obviously wrong! It's clearly used for making robot sounds.

To download a zip of the full Reaper project, click here. Here is what the project looks like:

Here's the output from my Reaper vocoder: https://soundcloud.com/superbonuspak/reaper-daw-diy-vocoder-experiment-1

And here's how I think it works:

1. Mod Generator is the voice that will modulate the noise. I called it that because originally I used the JS: synthesis/tonegenerator to sweep my frequency bands, before I replaced it with my voice. The sends go prefader from here to each of the Bandpass X tracks - prefader so that you can turn it down (don't want to hear myself dry) but it still goes to the sends. This has a copy of me on it saying something.
2. Noise Generator is the sound to be modulated. I called it that because originally I used pink noise. That still works fine. You can replace the ReaSynth on that track with the JS: Liteon/pinknoisegen and you'll get something that sounds less musical, and slightly more voice-like. Ha ha not really. But less musical.
3. Bandpass X (where X is a number from -1 to 5). This simply receives the Mod Generator voice and bandpasses it way narrow using ReaFir. Each band (each X) is a different notch in the spectrum, from about 300hz to 1500hz.  Why that range? Because I got tired of making bands. Why is there a band -1 and 0? Because after I did band 1-5 I realized I wanted to go lower, and I didn't want to redo all the filters. I just couldn't face it. Like Mod Generator, these are also pre-fader so it can be turned down all the way but still do the sends.
4. Noisemod X uses the Bandpass X send as sidechain input to a JS: SStillwell/expander. The noise comes in from the Noise Generator track via prefader sends, it is filtered with an exact copy of the ReaFir bandpass from its associated Bandpass X, and then goes into the expander, where the output from Bandpass X is sidechained in on 3/4. The expander has a high ratio and carefully tweaked threshold. These are the only tracks that are allowed to make any noise.

This is obviously not a practical way to build a vocoder, but I did learn some of the basics. I think it says something good about Reaper, that it's got the routing mojo to handle this. And I may use Reaper to explore more basic synthesis ideas in the future - it's almost like a bunch of little modules that can be connected by wires, very inexpensively.

Raised Bed Garden with Integral Fence and Access Doors

This is my raised bed vegetable garden, designed to keep out nature's little Dr. Lecters. We have a herd of deer, a gang of woodchucks, a squirrel riot and a insurgent army of chipmunks that will fight a dog over a bone. I don't know if this will work or not, but it's a try. Check it out, the sides open on hinges:

Eye hooks with bungee cords hold the doors closed, necessary because this thing is built to such loose tolerances that I am ashamed. I guess the idea was (there was an idea) that it would be easily disassembled and loosely coupled so it would go pretty much anywhere and hang together. In that respect I guess I succeeded. I guess.

It started as a box of 1x12 #3 cedar, 8 feet by 4 feet (or so). It was important to keep the long side exactly 8 feet because so much of the wood is on those sides, so the short sides are butted in between the long sides and cut a couple of inches shorter so that half-lengths would span it. The mesh is 1/2" galvanized 19 gauge "hardware cloth". It doesn't remind me of hardware and it's nothing like cloth but what can you do - that's what they call it. This is the floor that stops the woodchuck from digging up through the bottom.

I leveled it in the yard somewhat, but not completely. That's important, because I should have leveled it perfectly. Why did I not? Well, why do I do anything? For some reason on the day I was doing this I decided it was good enough, not really thinking through the concept of swinging doors that would want to be level, and how that would match an off-kilter base. I think it's because I was facing the task of dumping 1.2 yards of topsoil into this with a wheelbarrow and a car ramp, and I got everything all mixed together in my head.

There you go, almost ready for dirt.No, wait, don't do it! What if I filled it up now and then I want to drive my U-posts into the corners and they hit a big rock and I had to dig the whole corner out? Or, worse, move the whole frame because the rock was too big? First I need to put in the posts.

Posts and dirt. I bent the flanges on the posts because I want them to go right into the corners.

That's a good one. It looks great. The face of the post is right up against the box corner so that when I face the U-post with a 1x3 cedar plank the bottom of the plank will be flush with the wood of the box, and therefore the door will close with no gaps. Sounds great but the fact is that this post is not completely vertical, so when I goof around with it to get it more vertical it will open up a gap down there. But probably no bigger than the 1/2" of the fence I'll be using.

A door. 4 by 8, the reason I fought to make sure the box was exactly 8 feet long all the time. So I wouldn't have to cut those long boards at all. The verticals are just a 1x3 cut in half.

Fastened with stainless wood screws. The long end is going up to the left.

Then a strip of 1x2 cedar is used to make the short sides flush, plus secure the mesh as well and hide the bitter nasty cut ends from the door user.

In my imagination this board would be flush with the face of the box. Imagination is a wonderful thing, but actual thinking would be better.This is a 1x3 board bolted to a U-post in the corner. That's the reason for the U-posts instead of T-posts - they have 1/4" holes in them every 6 inches or so, making it easy to attach things to them.

That's the four 8 foot 1x3 cedar boards stuck to the corner U-posts. That's a 1x6 cedar board perfectly level at a distance above the box that will allow the 4' x 8' door fully cover the opening. I attached it with a longer bolt through the U-post and the vertical face board. However in two of the cases, the horizontal board actually fell between two pre-existing holes in the U-post, and I had to make new ones. It's only mild steel, but it gave me a hard time. Plus I had to drill in place, which is always fun.

Hinge detail. These went on easily, but that's mostly because this a high-tolerance situation. See how the 1x6 bolted across the face boards becomes a header for the T-hinge? So the door in theory is flat, thanks to those 1x2 strips, and has a flat place to get flush against. Except for reality of course.

Reality intrudes in the form of the gap at the foot of the face boards where they touch the box, as I said. It intrudes in the form of a bulge in the side of the box, presumably from the weight of the soil. Basically the door corners don't want to go flush, so we encourage them with some biggish eye hooks and some rubber bungee cords.

There's two tomato plants, some climbing cucumber (they climb?), broccoli, cauliflower, celery and hopefully some romaine. Who knows, it's going to be a veggie brawl in there and I've never done this before. I guess we can throw some deer netting over the top. We could put another course of mesh around the sides if we need more height for the tomatoes...

Who knows what's going to happen.

Buckyball Squid - Oh, Yeah!

This is my favorite original.  It's got physics mojo, relying on some kind of magnetic pole thing to keep the squid arms floating away from each other.  I hung the squid head on a length of monofilament threaded and knotted through a plastic washer I made from a drink cap.  I chose to hang the head and then attach the legs.  When attaching the legs, you move them in slowly to give them a chance to tell you if they're going to freak out and stick to the next leg over. Don't let them.  You know how it is - magnets.

 Squid hole view.  The head is a stack of 3-ball-side pentagons with a hole in the center.

Buckball Penta-Cannoli Torus

Guy at work got a set of these as a present.  They are ephemeral, occasionally frustrating, odd, and sometimes rewarding.  Here's a 5 sided torus made from little cannoli shapes.

 First make 5 of the pentagons.  We left the hole in the middle.  The shapes have 4-ball sides.  The hole may help with allowing the cannoli tubes to flex when you place the last tube.

 Here are the five.  If they don't want to line up square, you may have to unzip your cannoli and flip it the other way.  You know how it is.  Magnets.  The last one goes in tight; you have to simultaneously pry open the nearly-complete torus and make sure you get the match-up correct, too.  Took me two tries.

Building An Awesome Fishing Joke

I came up with this one at work today.  It arrived suddenly, like a cleansing bolt of inspiration from the realm of the divine.  You know what I'm talking about.

Q:  What did the alien trout say to the fly fisherman?
A:  Take me from your leader.

Genius?  I leave it to you, dear reader...but I think you do well to consider the alleged source.

Inexpensive Meade 4.5" 114mm F=1000mm F/8.8 DS-2000 Reflector - Disassembly and Focuser

This is a telescope someone gave me as a gift many years ago, purchased from a chain photography store.  It's part of the DS-2000 starter kit, which comes in both refractor and reflector form, and has a computerized drive system.  But I never got it collimated - or rather, I screwed up the collimation so badly that it saw poorly.

I have decided to clean it up and get it working, if possible.  To do that, I'm going to have to take it apart.

There it all is, taken apart.  Just do it.  Three sheet metal screws hold the plastic end rings to the ends of the tube.  Three machine screws hold the focuser to its port in the side of the tube.  Three thumbscrews hold the three tines of the diagonal to their slots in the tube.

I read that the focuser is "lubricated" with something that is practically tar, and that this gunk will eventually cause the stripping of the cheap plastic gears.  And I don't doubt it - that stuff was like...like...I can't explain it.  Tar is a good word.  First take the screws out of the square lid covering the gears.

Then you'll see a thin spring metal device that I believe is intended to push the shaft gear into the linear gear of the focuser tube.

Then under that, the gears.

This is about as complicated as the interface between two dirt particles, but there's one thing to watch out for, which is this side-play tensioner or what-have-you.  It's a screw pushing a rubber nub into the smooth side of the focuser tube directly opposite the gear drive.  You can see the nub just sticking out in the inside.  If I lost mine, I'd make one out of a couple layers of bicycle inner tube glued together or something.

Here's the mirror!  It's kinda ratty looking, but it makes your face look really big and is therefore fun to play with for a few moments.

I wanted to put a center mark on it.  I'd free-styled a red sharpie dot on it when I first wrestled with it, years ago, but I did it when it was still in the tube and - whatever.  It was ridiculous.  Took me ten minutes to get rid of it with isopropyl alcohol and a series of cotton swabs.   So I tried various techniques of marking the center, including this one.  Thales theorem rocks, but actually I wanted a ring around the center so a laser collimator would still work later.  Some people use those binder reinforcement rings, but WTF - I can do the same thing forty times harder, without spending $1.29.  Right?  Am I right?  So I wanted to mark a circlish thing around the center and hit upon a genius method.

It's metal ruler with a stop, in the form of some vice grip pliers, attached so as to place the end of the ruler, when the stop is at the edge of the mirror, a fixed distance from the center.

In the center of the mirror you can see the faint lines of Pigma Micron 005 I used to delineate a square with chopped corners - an octagon.  Then I stuck tape up to the lines and cleaned up the mirror.

Not very clean yet.  It's amazing, after getting a nice camera and cleaning some old lenses, how difficult it is to clean optical glass.  Some of the flecks are little grease spots that smear around but never want to leave - you almost have to nudge them off the edge of the mirror.  Some of them are pits from either bad manufacturing or awful storage conditions (my bad).  Some of them are scratches I put there.  Some of them either change every time I look at them or are permanent.  After about ten minutes of struggling, I got it as clean as I could and reassembled the scope.

I'll try collimating it next.