Manifest: Patch Engineering

Today is Manifest, Columbia College’s end of the year celebration.  It features many artistic presentations of Student work from the year. Part of this is a Mainstage for musical performances where I worked patch engineer on stage. I was in charge of making sure that everything on stage reached the house and monitors console. To do this we had the snake box pictured below.

There were 5 stage boxes on stage that connected back to a main snake rack, one on each side of the stage and at the DJ table. Each of these boxes connect back to a large box on a panel. There is a break out connected to the end of each snake and can be connected to a split that goes to both the house and monitors console.

There are a number of ways to set this up. We connected each mic to the closest stage box and then patched that to input on the split that matched the input list. After load in, we needed to add a few extra mics for a choir and a sax. so we added them into the closest stage box and patched it on the split rack. This is a lot of work for the patch engineer, but keeps everything smooth for the console operators. It has the disadvantage that it is hard to tell what input on the stage rack goes to what input on the console.

Another way of dong it is to label each stage box with inputs for vocals and instruments. Patch them one to one on the split rack and then soft patch them to where each engineer wants them on the consoles. The advantage of this is each stage box input always goes to the same input on the console so it can easily be soft patched and anyone on stage can figure out which input on the consoles it is going to.

Octopress and Static Sites

When a reader loads a page in WordPress, the server pulls the data required for the page from a database, generates the page, and then sends it to the user.  I does do some cacheing, but it still needs to uses server resources to know whether or not to generate a new page.  This is a bit overkill for a site that rarely changes.  Instead, we can use a static site.

A static site engines like Jekyll and Octopress generate the site ahead of time and then always serve the same HTML files.  The server overhead required to power a site like this is much less because no processing is done on the server side.

I have built another blog using Octopress.  I can make the site on my computer by writing Liquid Markdown, which is then built into HTML by the engine.  The files it generates are then synced to the web server.

So far, I have liked it a lot more than WordPress.  Everything either works or gives you a useful error message.  After messing around with Liquid Markdown a bit, I was able to transition posts form this blog over.  I was able to easily embed HTML code in the blog to customize it where the Markdown fell short.  It’s handling of images is infinitely better than anything that WordPress has done for me. Adding a theme to the blog was as easy as running a git command to download it and a rake command to install it.  I was able to easily edit the theme to have better fonts and menus.


Octopress fits my workflow my better than WordPress.  Everything is right there to be edited and having useful error messages when something goes wrong is a huge plus.  Unlike WordPress where the most you get out of an error message is Internal Server Error or HTTP Error.


Last weekend I went to Axpona, a HiFI convention and Expo.  It focused on the best quality sound products.  In Live sound, you basically use any speaker and signal chain and use Equalization to make things sound good.  This won’t cut it in HiFi; It is all about matching a source, pre amp, power amp, and speakers to your listening environment to get the best sound possible.  I only had a few hours there before I had to work, but i was able to see a few rooms.


HSU Research

HSU research was demoing a lot of their product line.  The had 3 tops and 3 subs in their room.  The sound was very detailed, but there was a something not right in the low mids.  They played some classical and Jazz tracks and the clarity in the higher range strings and percussion was very good, but the bass sounded a bit muddy and the clarity in the low end was not so good.  These were very reasonable priced coming in at less than $1k for the subs and under $2k for some great sounding tops.

Logan Martin

I listened to the Logan Martin ESL15A’s.  These are electrostatic speakers as opposed to dynamic speakers.  Electrostatics usually have amazing detail and these were no exception.  THe ESL15A’s were the best sounding speakers i heard their.  They lacked low end impact, but they were so flat it was amazing.  I have never heard the Keys in XXX sound so even and balanced between each note.  I could hear so much beautiful detail and no one thing stood out as bad about them.  My biggest problem with them was the price tag, coming in at $24,995 for just the pair of speakers. 

Columbia’s LAS Program (Get out of jail free for week 4)


Here at Columbia College, I feel like the Liberal Arts and Sciences program is a bit lack luster. The program feels like it was made to make the university look better on paper, but in my opinion is detrimental to the college learning experience. The LAS core here has are far too many classes required, many the classes offered are terrible.

It takes 120 Credits to graduate with a Bachelor of Arts. For audio, 55 credits are required in the department. Beyond this there are 40 credits of required liberal arts and science classes required. That is almost 3 solid semesters of classes not related to the program I came here to study. This is a lot of time and mental energy not put toward learning about audio or things I think are useful to my education. I would live to take a class on DSP programming, but I doubt I will be able to fit it into my schedule with such so many humanities, social sciences, and humanities Literature classes. What even is the difference between Humanities and Humanities Literature classes. In my humanities class all we did was read books, and talk or write about them. Wouldn’t one of each course suffice. It ends up being a waste of time.
Of the LAS courses, I have taken a few stood out as the worst classes I have ever taken. The humanities class I took was about the religion and ideals of a couple of countries. There was an international student from one of these countries in the class and when the professor asked her about half of the things we were learning about in the class she had never heard of them. It ruined the professor’s credibility in my mind. I know that one student is a tiny sample size, but still there is something to be said for that. In my oral expression class, I feel like the instructor picked topics he thought would be “fun.” He made up topics for some of the major speeches by drawing from a hat, leaving us with terrible topics for speeches. Meanwhile my roommate was in another section that allowed him to do speeches on topics in his major. His final speech was arguing some aspect of a camera lenses while mine was on cats vs dogs as pets. For the time and money wasted on these LAS classes, the college could at least make sure we are at least getting something worthwhile.




Logical Confusion: AND, XOR, and NOT


For my Audio Electronics class, I had to build a sequencer and wanted to add a bit more processing to it.  The sequencer is based on the 4017 Decade Counter.  This chip has 10 output pins and counts to 10 by bringing each pin high in order when it receives a clock pulse.  I am using it control 10 notes that will get sent to an oscillator.  I wanted to do some processing to the incoming clock before it reached the sequencer.  I want to have a button that stops the clock pulses from reaching the chip, and a button that manually advances the sequencer.  To do this, I used 3 additional logic gates, the 4081 Quad 2 input AND Gate, the 4039 Quad 2 input XOR Gate, and the 40106 Hex Inverting Schmitt Trigger.


The 4081 Quad 2 input AND Gate has a long title, but it tells you exactly what the chip does.  Quad meaning it contains 4 separate AND gates, 2 Inputs meaning it compares 2 signals, AND meaning the output is high if and only if both inputs are high.


The 4030 is very similar except it’s an XOR so its output is high if and only if one input is high.

The 40106 is a bit different, instead of comparing 2 signals, it uses a special trigger that switches from high to low and low to high at different voltages removing confusion when the voltage is right around that middle threshold.  It is also an Inverter so when the input is high, the output is low and visa versa.

In my circuit, I use a button to hold the clock.  This gets Inverted so when the button is not pressed, the output is high. First the button is connected to a circuit that removes any bounces from the button and inverts the signal.  It is then connected to the 40106 where any intermediate states are removed and the signal is inverted again.  Since the signal was inverted twice which is a double negative and thus not inverted, I need to send it through the 40106 once more to invert it.

This signal then gets ANDed with the Incoming Clock using the 4081.  So If the button is not pressed, the AND gate just outputs the clock, but when it is pressed, the AND gate stays low.

There is another button to manually advance the clock.  This is Denounced by the same RC filter and the 40106, but it does not need to be inverted so we only need to use the 40106 once for this button.  This signal is then sent to the 4030 where it is XORed with the output of the 4081 and sent to the clock.


H is the Hold button, S is the step through button and C is the clock pulse.

Reflection On Tiny Portfolio

For My tiny portfolio, I made a page about the props I have made. For this page I tried to pictures of the props as well as a description of how they were made, what processes i used, and a description of how the prop functioned in the piece.  I tried to incorporate some of the concepts we talked about in class such as Keeping the text short an broken up, photo composition and keeping the photos simple, and having a clean Distraction free page.

Keeping the text short was a challenge with the word count of 300 words.  I wanted to use bulleted lists to talk about the basic process and build, but I could not get the required word count using bulleted lists and captions. I ended up using more of a prose approach to the text, but still broke it up with headers.

For the photos I used, i tried to keep them distraction free and follow some of the composition aspects we discussed.  I tried to keep the photos focused on the subject matter and free of background distractions.  This helped keep the photos clean and helped them show off the props.  The only exception to this was with the stills from the films.  I did not want to crop these because they show off the props in context.

I also tried to keep the pages clean and free of anything unnecessary.  The pages were made so it would be just the basic UI of the blog and only the images I need.  I picked the most captivating image as the header image for each prop and layed out the other images in a concise and easy to look at way.  Clicking the images brings up a light box with a larger version of the photo.

House and Monitors on One Console

On many smaller shows, you are the only engineer there. Just mixing house can be overwhelming, at times. Start adding in monitors, lighting, and stage management, and it can get real hairy, real fast. Mixing Monitors and Front of House from the same board can be challenging. If you run monitors from the same channel as house, you are in for a challenging night. Any EQ change you make in house will be reflected in the monitors. Moving the high pass filter around 100Hz might change the phase of something higher destabilizing your monitor. Even worse, the harsh sound a monitor needs to cut through the stage noise will sound terrible in the house. It can be a lot to manage, but there is an easier way if you have enough channels.

Instead of just sending using a prefader send on your house channels, most digital consoles allow you to soft patch the same signal to multiple channels.  This means that if you have a mic plugged into input 1 on the console, you send it to both channel 1 and channel 2.  You can stabilize the mic in the monitor on channel 1 and use channel 2 for the house.  This way, any compression, EQ, or inserts you use on channel 2 will not effect channel 1 at all.  When I am mixing on an X32 and there are less then 32 channels, I almost always use the split at least my vocal channels for monitors.

THere are a few cavets to this method.  First is that you need a digital console; Analog consoles do not have the same soft patching capabilities. Second, if you are splitting the input on a digital console, then you share the Preamp between the two channels.  If you need more gain in the house, turning the knob will also increase the gain to the monitor. Both of these can be solved by using a simple Y cable to split the 2 channels before they hit the pre amp. This requires you to have a bunch of Y cables.  If you don’t have a Y cable, you have other options on Digital consoles.  Some digital consoles have a handy digital trim that allows you to apply gain to the digital signal coming into channel after it is split.  Not every console has this, the X32 for example does not.  On the X32, you can get your extra gain from the Compressor.  Most compressors have output gain or makeup gain.  This can get you that extra boost you need to get the channel loud enough.



Dickbutt: Adding light

When I made the first small scale Dickbutt, simply placing it on an LED Luminary lit it up very well. The light spread evenly through the model and it looked really good. Even though the bottom lights up brighter than the top, this gives a nice effect.


But this did not scale well when I made the full scale model. The full size model suffered from the same issue where only the bottom lights up, but because it is much larger the effect is a lot more noticeable.


The top does not light up at all while the bottom is super bright.  To fix this I needed a larger LED.  I thought about puting a high power LED in the base where the luminary is, but the bottom would still be too bright.  I also thought about suspending an LED in the model when it is cast, but with the 5mm LEDS I had, this probably wouldn’t look right.  That is where some really cool LEDs I came across would be perfect.  I had some LED Filiment bulbs lying around so i figured i would try to remove one of the filiments and suspend it in Dickbutt.  I smashed the light bulb and carefully cut out the filiments.  These filiments turned out to be really cool.


They are tiny strips of around 25 LEDs that cast light out in 2 directions.  Each strip is about 1 watt and runs at about 60-70 Votls.  This is incredibly high for LEDs and is a lot higher than I usually work with.  I don’t have a power supply that puts out that high of a voltage so I added some 9 volt batteries until I got the voltage i needed.  I then suspened the LEDs in the mold and cast another Dickbutt.


The result is quite spectacular.  It lights a lot more evenly and looks really good with the white light instead of the yellow orange candle light.  The director and I are really happy with how it turned out.

The Room is Too Damn Loud

Doing live sound often mean doing audio in rooms that are less than ideal acoustically.  I recently worked on a fundraiser auction in one of the worst rooms I have ever heard. The room was too reverberant. When the room filled with 500 people, the conversation in the room was pushing at least 95 dB.  Too overcome this, we couldn’t just turn up the volume on the main speakers next to the stage.  Because of all of the reflections in the room,  you wou;d’t have been able to understand anything the presenter was saying.  Instead we put up tons of delayed speakers that were time aligned to the mains.


Here we used 3 K12’s per side.  These speakers played the same content delayed so they arrive at a listener at the same time as the main speakers.  These speakers kept the presenters intelligible in throughout the room.

As people are apt to do when drinking, they kept talking through the auction, meaning we had to push the gain.  My co worker metered the SPL in the middle of the room and came up with 105 dBA.  This is almost as loud as a rock concert or a loud sporting event.  THis kind of volume is a little insane for a small charity auction, but it was the only way to keep sound intelligible in the room.

WHen I Turned down the delays after the auction when music was playing, the further you got from the mains, the less of the vocals you could hear. Other instruments like the bass and cymbals carried just fine to the back of the room.  It was just the high mid frequencies in the voice that were smeared by the reverb and obscured by everyone trying to talk over the music.

Network audio

Today, Audio engineers are rapidly becoming network engineers.  Running Audio over Ethernet is now becoming the norm for large and small shows.  Currently, there is no one standard that has dominated the industry which leads to a bit of a mess; every manufacture seems to have their own standard.


  • 32×32 Channels
  • Devices are Daisy Chained to share channels between devices
  • Cannot be used with network switches
  • Used by Klark-Technic, Behringer
  • Used mainly for entry level equipment



  • 21-364 channels
  • .5-2 ms latency
  • Consists protocols for time synchronization, configuration data, audio, and video
  • Can be used with network switches
  • Requires special AVB compatible Switches
  • Newer Apple laptops support AVB natively.
  • Used by Avid, Motu, and Presonus



  • 128 channels per single device
  • Works with standard network switches
  • Latency from microseconds to 10 ms
  • Supports interoperability with AES67
  • Used in Yamaha, Lab Gruben, and  Shure