As an artist first, and engineer second, it’s important to know that whenever I’m tracking, my performance is being captured at the highest fidelity (that I can afford!). It adds to the confidence of my performance, and it lets the engineer inside me know that when it comes time to mix, my job will be easier. So while I am definitely no “gear snob”, I do need a certain level of confidence in my signal path.

I started recording music digitally in the mid 90s on prosumer gear like the Digi 001 and Echo Audio cards. All seemed good enough at time, as I was using decent mics, but it wasn’t until much later that I had my first “wow moment” after buying my first “real” pre (Great River). Compared to the stock interface pres, or my cheap external low end pres, the quality was night and day. Depth, transparency, detail, etc…all the buzz terms that makes gear fiends drool. But as I graduated to better audio interfaces, the conversion aspect always remained a source of insecurity. So to finally put my concerns to rest, I decided to augment my Fireface 400 with a Mytek Stereo 96 ADC to see if the converter element was as relevant as what I’d read. To properly test this mysterious world of converters, I did some A/B analysis today. And since I know I’m not alone in this confusing world of converter quality, I decided to share my results with the Internets to help others.

The gear and process I used for my tests is as follows:

Vocals -> Bock 195 Mic -> API 512 Mic Pre
American Jazz Bass -> API 512 DI Input

All signals were routed out of the API to a Whirlwind IMP 1×2 mic splitter, then one side was sent to input 6 of my Fireface and one side sent to the left channel of the Mytek Stereo ADC. Levels were matched as closely as possible, but I normalized all audio to be fair. I hooked the Mytek ADC to my RME via SPDIF, and also clocked off that device. Screen grab of my RME Settings can be seen here. The rest of the gear used can be seen here

Monitoring was done out of the DAC on the Fireface into Focal CMS 50 Monitors, and a pair of AKG K240s headphones

You can download and hear the comparisons in three session formats (all 24 bit / 44.1kHz):
Logic Session
OMF / Pro Tools Compatible
Raw Audio (Doesn’t include all samples)

In my opinion, the Mytek unit provided more detail in the mids and top, and added slightly more deep bottom. The FireFace still held up, but did sound a little “cloudy” next to the Mytek. Depending on your budget, only you can decide if this sound difference is worth $895. I got my ADC used for about $625, so I’m happy with my purchase, and will most likely track all my mono performances through this unit.

Hopefully this helps anyone else out there suffering from converter envy.

Here’s a collection of some m4a metronome tracks your drummer can use during rehearsals, gigs or even recording. I know there’s metronome apps out there already, but the ones I tried did not keep a consistent click, were clumsy to adjust, and would sometimes even crash. By having each tempo as a track, you get rock solid playback, and you can even create a playlist to let you switch between specific tempos in the middle of a set. Each track is about 15 minute, and they range from 78bpm to 180bpm, each in 2bpm increments. I didn’t bother creating anything slower than 78bpm, because personally if a track is very slow, I’ll want my metronome in 8th notes anyway. Finally, I used the m4a / aac codec because it made the file says way smaller than even the lowest rez mp3s. Download for free here: http://bit.ly/pjm6Mj

tl;dr I’m learning the dark art of mastering, and will master your music for free so I can gain more experience.

For almost 15 years I’ve recorded and mixed my own music, but have always avoided doing my own mastering. Partly because I enjoyed a fresh, unbiased set of ears on my mixes. Partly because I never understood what mastering engineers did. But mainly, because people more talented than myself always told me “that’s just something you leave to a professional”. But about 3 months ago, I decided to finally give it a shot. And to be honest, it is tough! Even with today’s professional tools (Elephant, LL316, etc) getting your mix at radio-ready levels without smashing it beyond recognition is pretty hard. Even theoretically simple things like “adding a bit more top” are more complicated than they seem. Should I be using a linear phase EQ, or traditional EQ? Should I be using a “colored EQ” or something more transparent? Why did that 8k boost sound great on everything but the hi-hat? Without a doubt, making broad stroke changes to a stereo track is a very different beast than mixing, and also kinda terrifying knowing you could possibly undo all the hard work you spent on a mix. However, through a lot of painful but fun experimentation, I feel like I’m finally getting a good footing in the process. And now I’d like to take my skills to the next level by mastering some music that isn’t mine. So if you have a mix that you think is good, I’d love to take a crack at taking it to the next level.

Here’s a track I recently mastered:

One Thousand Times by flightcrashcompanion

And here’s a snippet of the original:

Unmastered snippet of One Thousand Times by flightcrashcompanion

The goals I wanted to accomplish with this mastering job were: 1) get it to a radio-friendly volume 2) add some sparkle 3) add some deep, sub-harmonic bottom that shook my subs without sounding muddy 4) add some tasteful glue so the punch of the drums sounded more natural

You can hear more of my work here: http://flightcrashcompanion.com

So if you have a mix you’re happy with and would like to get it mastered, I’d love to see if I could help make that happen.

The following drums were recorded in 4 passes using only 2 mics. Read on to see how it was done.

One Thousand Times :: Section A Full Kit by flightcrashcompanion

The Dilemma
Outside of a proper studio, recording drums is always a pain in the ass. Lack of mics and pres, lack of interface inputs, poorly tuned rooms, noise concerns, bleed from other sources and sloppy drummers all make guerrilla drum recording very difficult.  I have a decent setup for tracking mono sources, but as soon it’s times to track drums, things get complicated. Historically, the options have been: pay for studio time, rent a bunch of gear, or try to fake it with drum software samplers.  If your song is really straight forward, sometimes you can get away with drum software like BFD, Superior Drummer or Addictive. However I find I end up spending too much time tweaking it to sound “real”.  Recording in a proper studio is always the best choice, but laying out upwards of $800 for an 8 hour day can get expensive quickly, not to mention the added pressure of The Clock. Gear rental is sometimes an option, but gear selection and availability are often limited,  and deposits costs can reach into the thousands.

One Limb at a Time
I had heard of 80s recording engineers and producers using the “shells only” technique, where they’d record kick + snare + tom on one pass, and then the hihats + cymbals on another. (I think “Girls Girls Girls” by Motley Crue is an example of this). This was an intriguing approach, but my main interface at the time [Apogee Duet]  could only record two tracks simultaneously which severely limited my options.  Not wanting to be one of those whiny punks who blame their poor recordings on gear, I decided to see if I could make a quality multi-track drum sound by tracking each drum element separately.

Mic & Kit Setup
The following setup took a bit of trial and error. My first attempt yielded great direct sounds, but didn’t capture enough room  / overhead sound, where much of the character lives. I found my AT4050 in the Figure 8 position was very good at doubling as an overhead and room mic. Mic’ing a bit farther and higher than usual worked for the natural sound I was looking for. I also found it’s important to have the full kit setup when recording all your passes. Little things like snares rattling on your kick drum track and the vibration of other shells all add to the realism.

The Performance
A good drummer and a good performance are the biggest drivers for a good drum sound, and recording individual pieces is no exception. Aside from the obvious attention to dynamics, consistency and rhythm, it’s important to be mindful that the individual elements will eventually be summed together. For example, make sure you “pull off” the hihat and ride when appropriate to create the illusion of a player that only has two arms. You wouldn’t be bashing away on the hihat while you’re doing a big complicated drum fill. Or less obvious,  don’t hit the crash and the hihat on the downbeat – it’s unlikely you’d do so in a traditional performance.  Maybe as a drummer I pay  too much attention to the drums, but being wary of these things can make or break the realism of the recording. Also, after you’ve recorded each pass, be sure to “sample” indivual elements at multiple volumes. These can come in handy for sound replacement, patching mistakes, or even creating a playable sampler version of your  kit.

Lining It All Up
This is where things get a little tricky. You basically have three options:  be a good drummer and play each part in time, manually line things up in your DAW, use audio quantization tools  like Logic’s Flex Time. I’ve found the latter works really well, and is dead easy once you get the hang of it. As long as your performance is in the ballpark, Flex Time can quantize your drums perfectly with very little artifacts.

The Mix
To get the most realistic mix, I create busses that simulated a real drum mix. All room / OH tracks gets sent to an “Overhead” buss, and all the directs get sent to another buss. Then, both those busses are sent to a third buss and hit with a compressor which acts a glue for the whole kit. When mixing drums, I almost always employ some kind of double bussing, but that’s more a matter of personal taste.

Here’s a snippet of the overhead buss I created ( EQ + heavy Comp). I prefer a roomier drum sound so I led the final drum mix with this buss:

Section A – Overheads by flightcrashcompanion

And now here’s the direct tracks (EQ + light comp + CLA Drum plugin on the kick):

Section A – Direct by flightcrashcompanion

Since each element is on it’s own track, sound sculpting with EQ is a breeze, as is making edits. I can alter every piece of the kit without effecting other elements, all while retaining an organic room sound.

So that’s basically it! None of this is really ground breaking, but I’m very happy with these results, and am looking forward to trying this technique again.  Please feel free to ask any questions, and if you’re so inclined, check out the final version of “One Thousand Times” when it’s released on May 10th.

When I went to deploy my site to my prod environment (EC2 Amazon Linux), the Topspin 3.0 plugin kept shitting the bed with all kinds of weird output. The culprit appeared to be around <?=$_server['request_uri'];?> not returning any info, but that’s all had to go with. Did a bit of research, and realized my default php install didn’t have short tags turned on. All I had to do was edit php.ini and change short_open_tag = On and restart apache and it worked ! See below for more detail.

; This directive determines whether or not PHP will recognize code between
; tags as PHP source which should be processed as such. It’s been
; recommended for several years that you not use the short tag “short cut” and
; instead to use the full tag combination. With the wide spread use
; of XML and use of these tags by other languages, the server can become easily
; confused and end up parsing the wrong code in the wrong context. But because
; this short cut has been a feature for such a long time, it’s currently still
; supported for backwards compatibility, but we recommend you don’t use them.
; Default Value: On
; Development Value: Off
; Production Value: Off
; http://www.php.net/manual/en/ini.core.php#ini.short-open-tag
short_open_tag = On

Now that I’ve (mostly) finished converting my site to wordpress, it’s time to build a home for it. For this project, a micro EC2 instance backed by EBS made the most sense.

The AMI I ended up using was ami-8c1fece5 . It’s a 32-bit box running Amazon Linux. I’d never used Amazon Linux, but I liked the idea of getting familiar with a *nix flavor that I’m confident Amazon will heavily support. For this install, all I needed was a standard LAMP stack, and install everything was a breeze:

Login as ec2-user (not root)

Install Apache, MySQL, PHP & PHP-MYSQL Adapter, PHP Extras
sudo yum install httpd
sudo yum install mysql-server mysql
sudo yum install php php-mysql
sudo yum install php libmcrypt libmcrypt-devel php-mcrypt php-mbstring

Turn on short tags in php.ini
short_open_tag = On

Add the following to httpd.conf :
<Directory “/var/www/html”>
Options Indexes FollowSymLinks MultiViews
AllowOverride All
Order allow,deny
allow from all
</Directory>

Configure apache+mysql to start automatically on reboot:
sudo /sbin/chkconfig httpd on
sudo /sbin/chkconfig mysqld on

Start my services:
sudo service httpd start
sudo service mysqld start

Setup my root mysql user:
sudo /usr/bin/mysqladmin -u root password ‘password’

Temporarily change owner of your doc root
sudo chown -R ec2-user /var/www/html

Upload / or clone my wordpress repo to /var/www/html

Change ownership of my wordpress folder to back to user ‘apache’ so WordPress can do things like self-install plugins, templates, etc, directly from the admin control panel
chown -R  apache /var/www/html

You could go in to httpd.conf have the httpd process run as a different user (like ec2-user) so you can upload/download via ftp AND make updates. However since you can’t login into my machine as ‘apache’, this adds a layer of security.

Now I must change all referces of “localhost:8888″ in my database otherwise wordpress will shit itself when you try to access the remote version
Export a copy of my db
Find and replace all references from “localhost:8888″  to my new host name: flightcrashcompanion.com
Exexcute .sql file on my prod server

You’re done !

Over the past 10 years I’ve owned in-ears & ear buds of almost every variety, logging many listening hours on prosumer offerings from Shure, LG, Ultimate Ears, Sony and  more. Some wear out, some get lost, and some just don’t sound right.  But recently, I’ve found a very inexpensive pair that I’m really enjoying – JVC Marshmallows.  They have very tight, focused bass with little hype, but can go very low when needed – like on the subsonic booms of LCD Soundsystem’s Disco Inflator. The highs and mids are sugary without being too sweet – a problem on many of the low-end ear buds. And unlike some of the high-end offerings from Shure that give you a very flat, “accurate” sound, the Marshmallows add just enough sizzle and sweet to make the music come alive. They also stay put it my ears and rarely need readjustment. Finally, on the iPhone 4 they have plenty of gain with the volume around 75%, and more than enough gain when I’m drunk and require unsafe listening levels.

Last Saturday I discovered some very interesting things about my recording setup.

* The input on my Macbook Pro is not only analog, but also supports digital connections (via TOSLink 3.5 mm connection)! Sending two channels of digital IO directly from my macbook should have less latency than 2 channels via firewire (or USB). And it opens up a ton more options for high quality AD/DA.

* Although Apogee’s documentation states otherwise, you can indeed create a fully stable aggregate device using your Duet plus your macbook’s two on-board pres. With two external pres, I was able to record 4 channels of simultaneous audio! And the converters were totally legit! Some have claimed aggregating devices with the Duet works for brief period, but then eventually falls apart. I used this setup for 10+ hours and didn’t have any problems