Taking Time

I recently recorded some music with an old friend of mine.  We hadn’t recorded together in many, many years.  It was striking how easy it was to work together again after spending a few minutes catching up.  As we closed in on some ideas for our second track, an idea hit me when we were listening to the vocals.  I had an idea to segment the vocal line into independent samples and control the rhythm of them using the sequencer.  We also spent some time transforming the rhythm of another part to accentuate some subtle noise in-between two phrases of laughter.

My friend also works with another artist occasionally who is signed to a major label, and he remarked that it had been a long time since they had taken time to get an idea exactly right based on their vision.  His other working partner preferred to move quickly and not spend too much time honing details.

I think there is a dilemma facing most artists, especially when using technology to realize ideas.  Sometimes the vision of the idea is outside what preset workflows are available in the tools at hand.  The choice presented is either to slow down and try out the idea by meticulously manipulating the raw materials or to simply let it go.

This choice is not always easy to make.  I think I need to understand how much work an idea might be before I take the time to realize it.  Interrupting the train of creativity can lead to losing the inspiration that propels me in the first place, so sometimes it’s a gamble.  However, I’ve found that the times I do take the time to at least try them out doesn’t set me back too far if the ideas fail.

Thus, to me, it’s usually worth taking the extra time to slow down and temporarily sacrifice inspiration to follow through on difficult tasks.  If the tools you have don’t make the realization process easy, then you might be on to something surprising, unique, or inaccessible to others.  Try it out.

Old Samplers & Modern Computers Redux

In my previous blog post about samplers and computers, I lamented the fact that Akai’s old Ak.Sys program would no longer run for me if I upgraded to Apple OS X 10.7 “Lion.”  Well, that’s mostly true.

So since Ak.Sys is a PowerPC app and Lion finally does away with PowerPC support, there’s a problem for those of us who like to use old samplers, still.  I liked being able to transfer samples to the S5000 over USB and also to adjust parameters on it from my computer.

This is going to be a short post, so here’s what you need to do if you want to continue using your S5000 over USB with modern computers.  First, download the Ak.Sys packages for Windows from Akai.  You’ll find them on their website.  There are four: you need the latest version of Ak.Sys, the USB drivers for the S5000, S5000 support for Ak.Sys, and the S5000 “themes” for Ak.Sys.

Second, if you’re using Windows 7 64 bit, you’ll need to install “XP mode” which allows you to run a 32 bit environment as a virtual machine.  If you’re using Windows 7 32 bit, then you don’t need to do that.  If you’re using a Mac, like me, use a Virtual PC program like VirtualBox, Fusion, or Parallels to create a Windows virtual machine.  Make sure you install a 32 bit version of Windows 7, not 64 bit.

Third, once you have your OS up and running install the Ak.Sys program.  Then, unzip the S5000 USB drivers to a known place.  Turn on the S5000 and connect it over USB to your computer.  If you’re using a Mac, tell your virtual PC running Windows to connect to the S5000 over USB.  Windows will then ask you for a path to the drivers.  Find the S5000 drivers you just unzipped and point the installation process to them.

Now, install the S5000 themes for Ak.Sys on the Windows machine.  Then, install the S5000 support files for Ak.Sys.  After these steps, you should be able to run Ak.Sys from Windows and see the sampler.

At this point you can configure your computer however fits your workflow best.  I like to map a drive from Windows to my samples folder on the Mac so I can find the files I want easily.  I use VMWare Fusion, and that supports a Unity mode.  That means I can treat Ak.Sys mostly like a regular Mac program so it’s not too far away from the good old days when it ran natively on Mac.

Revisiting and Rediscovering the V-Synth

It is well documented I am a music gear nerd.  Every once in a while, I go back and dig out the manual for a piece of gear I have had for a while to get reacquainted with it and to spur new ideas for sounds and music. Last night, I cracked open the manual for the Roland V-Synth XT, which I’ve had for about 6 years now but haven’t really used that heavily.

The Roland V-Synth XT. You can't see from this picture, but the rack case has a cool pivoting case. The eight knobs provide a lot of control, and you can switch the touchscreen into a fun control surface.

I’m glad I did. I always thought the V-Synth sounded amazing, and I remember playing the keyboard version in the store when it was released. I liked how one could warp loops and samples using the touchpad, and I thought the overall sound of the machine was great. Then, Roland released the rackmount XT version which also includes a vocal modeling program as well as a D-50 model. I splurged and spent some of my signing bonus at Microsoft on the XT.

Last night I read the XT’s manual cover to cover and switched on the V-Synth to play with some of the features I read about. The effects section on this machine is amazing, with three distinct processors. There’s a multi-fx processor, a chorus/flanger, and a reverberator with flexible routing across the modules. The multi module has just about every effect you can think of plus some stacks, like phaser and delay in one program. The chorus sounds amazing and even approaches the richness of the Juno-60’s chorus.  One of the oscillator types in the voice architecture is “external in,” so one can use the V-Synth as a fancy effects processor.  Fancy in this case means up to four filters with envelopes/amplifier running into up to three powerful effects algorithms.

With a synthesizer of this complexity, it is easy to get overwhelmed with depth. But somehow Roland managed to keep it intuitive. The analog-style waveforms sound pretty good, and the VariPhrase/PCM oscillators have a great deal of depth. Some surprises to me were that the V-Synth has a built in sampler that actually seems more full featured than the dedicated samplers I have. You trade polyphony (e. g. 64 voices in the e6400/S5000) for the ability to play back sounds and change speed and/or pitch independently. There’s also this cool mode in which you set up a sample with specified temporal points and each successive key press steps to the next one – great for sampled beats.

While I was trying out some ideas on the V-Synth after reading in the manual, I stumbled upon the first limitation that made me feel a little disappointed. I was using the square wave oscillator and the LFO to modulate pulse width. Then, I wanted to route an LFO to pitch as well. I assumed there would be independent LFOs for each, but that isn’t the case. I thought there was just one LFO in the V-Synth’s voice! I reconciled this by thinking the Juno-60 which I love so much only has one LFO… but then I found that it turns out each section on the V-Synth has its own LFO. Each section also has its own ADSR envelope, key tracking, modulation… so yeah, it was way deeper than I imagined. Score!

The Roland D-50 - one of the first synthesizers I fell in love with, circa 1987

The D-50 emulation in the V-Synth is also awesome. The D-50 was one of the first synths I lusted after in the late 1980s. While I don’t have room for another 61-key keyboard in my small room, the emulation in the V-Synth sounds just like the real thing, and it even is 100% programmable just like the D-50 using the PG-1000 programmer. I left the manual for this for another day, but it’s nice to know it’s sitting right there waiting to be rediscovered.

Programmer for the Roland D-50. Lots of sliders! This programmer also works on the V-Synth's emulation of the D-50.

Also on the rediscovery pile is the manual for the Eventide H3000 D/SE. This effects processor has been a studio mainstay since the 1980s, and to me the sound even surpasses some of today’s modern processors. When I’m ready, I want to crack open MIDI control of this guy and see what ideas arise.

Removing Roadblocks

As time goes on, I find it more and more difficult to devote time to some endeavors that are important to me.  At work, there are always a thousand threads going at once.  And at home, it’s hard for me to carve out time to work on music.  I know I’m not alone in these pursuits.  Whenever I find myself on a roll, whether it’s me coding in an editor towards a goal I have or following my creative path, I loathe being interrupted.  Usually this happens because of something completely preventable, but in the moment I find that I lose all steam.  To wit, there is nothing quite like a software upgrade or bad cable to dampen my mood.  When I have an idea for a new melody or rhythm, the last thing on my mind is troubleshooting a patch bay or upgrading software.

Therefore, I spend a lot of my time planning and preparing for those ebullient moments when creativity strikes or when I find the time to concentrate on solving problems.  In the studio, everything is connected together already, and most of the signal processing routes I need are already pre-configured.  At work, I spent time gathering the reference materials, test data, and scripts I will need to execute fast and without interruption.

Inevitably, interruptions happen – software goes bad, I make a mistake, or something comes up that I just haven’t thought of before.  The best technique I know to minimize the impact of these externalities is to defer acting upon them until it is truly necessary.  If I find a problem with my recording template, for example, chances are I don’t need to remedy it immediately.  Instead, I keep a log of tasks that need to be done.  If I can work around the problem temporarily and defer solving the problem to another time, that keeps me moving forward and not distracted on the problem.  And of course the best time to come back to small tasks like that are those times in which I’m either not creative or just not in the zone.

It’s possible, though, to spend all one’s time making, tracking, and dealing with lists rather than getting anything done.  I’ve read before that the most effective people never complete their own “to-do” lists.  Instead, they know how to get the most important things done and how to defer lower priority tasks.

The fact that I’m even writing about this evinces that I’m a list-maker and that I like organizing thoughts and tasks.  I’m prone to wanting to complete a list totally, and it’s nice to be able to catch myself and know that the list isn’t what is important.  Get the important tasks done and defer the rest.

I find that if I spend my efforts clearing the road, I can make a lot of progress whenever lightning strikes.  Since those times are extremely scant and valuable, that time spent preparing makes all the difference.

Old Samplers, SCSI, and Modern Computers

Using computers with old music gear gets harder and harder over time.  I built a large sample collection over the years for my hardware samplers, both by sourcing sample CDs as well as creating my own sample collections.  I have three old hardware sampler machines that I still use today: An Akai CD3000XL, an Akai S5000, and an Emu e6400.

I tried to make the shift over to software samplers that are available now, like Kontakt or Gigasampler.  But for whatever reason, I still prefer the old way of using those hardware boxes.  Sometimes I think it’s the way they sound, and other times I think my fingers just know how to program those front panels faster than I could do the same thing on a computer with a mouse.  Perhaps I just like the tactile interface.

Whatever the reason, it’s hard for me to let go of those old samplers.  And of course, I want to efficiently get sounds into and out of them so the sample library isn’t static.  I definitely do not want to edit the waveforms themselves in the sampler; this is one place where a large screen and mouse blows away working on the hardware.

That’s where the friction begins.  Of course I’d love to be able to plug in any of these devices to my computer via USB (more on that in a minute), but the older devices like the Emu and the CD3000XL don’t have USB.  Instead, they used the SCSI bus to get sounds in and out of their memories.  And if any of you out there have used SCSI before, you know what a pain it can be.  Is the chain terminated?  What IDs are in use?  Do I have the right driver?  Can I hot-swap devices?

I spent a lot of time troubleshooting this part of my studio, how the samplers communicate with a modern computer.  I’d like to share what I learned, in case any of you out there are also gluttons for punishment: you love your old samplers, but you want to use modern computer hardware with them.  Of course, your mileage may vary, and I don’t claim to have all the answers.  However, the following worked for me.  Let’s tackle each piece on its own, shall we?  We have the rough structure of the setup, then each sampler to deal with.

THE SETUP

In my studio, my main writing and recording workspace is an Apple Macintosh computer.  I use either Cubase or Live for tracking, and these programs live on my main monitor.  On the second, smaller monitor, I usually keep a few windows open, like a spreadsheet that contains my routing cheat sheets and so forth.  When I’m using a sampler, I will also open the sampler’s host program on this second screen as well.  What this means in practice depends on the sampler I’m using.

The rack of samplers in my studio. From the top to bottom: Akai S5000, Glyph SCSI switch, E-mu e6400, CD-ROM, Jaz drive, Akai CD3000XL

Over in the racks, I have a Windows 7 computer with a SCSI card connected to an Iomega Jaz drive, a Plextor CD-ROM drive, and a Glyph 3-way SCSI switch.  The switch is, in turn, connected to each of the three samplers.  I really only need to use the computer with the CD3000XL and the e6400, and it’s nice to have a CD-ROM and Jaz drive for reading and writing programs, samples, and so forth connected to each machine.

For SCSI adapters, I recommend some variant of the Adaptec AHA-2940.  Windows 7 doesn’t officially support the 2940, but since Windows 7 and Windows Vista are close cousins as far as drivers are concerned, you can install the Windows Vista drivers, and it works.  I use the AHA-2940UW, the “ultra wide,” single-channel PCI card.

The next issue we need to address is that the samplers in question use what is known as the ASPI Layer for SCSI on top of the SCSI buss in order for host software to communicate with the samplers.  This too is no longer supported in Windows 7, but you can coax this into working as well.  After downloading, unzipping, and installing the ASPI layer, reboot and it should be up and running.

Now that the computer is up and running with a SCSI adapter, driver, and ASPI layer we are ready to get the samplers talking to host software.  Let’s begin with the Akai.

AKAI S2000 SERIES

The steps in this section pertain to the “S2000″ series of Akai samplers.  Akai made many devices that used the same logic generation, all based around the S2000 chip.  These are the S2000, S3000XL, S3200XL, CD3000XL.  These are great-sounding machines, and one can fit some option boards in them for adding effects and a second filter.  The filter card is exciting, because you get a second multimode filter for each voice!  Akai produced a piece of software for this series called MESA that allows you to transfer samplers to and from the machine and to program voices, multis, etc. from the computer.  It turns out that MESA II still runs on Windows 7, and this is what I did to get it running.

Turn on the sampler before the computer.  Make sure the sampler is running the latest version of the operating system, which is OS 2.14.  If you don’t have OS 2.14, you can find it online by searching for it.  Place this on a floppy using the appropriate tools, and put it in the sampler’s floppy drive before booting.

When the computer starts, ensure that you see the sampler listed during the SCSI card’s post screen.  For my computer, it looks like this:

Notice the Akai entry when the SCSI adapter scans the bus

If you don’t see the sampler in the list, that means that there’s no way the host software will see it, either.  Check for SCSI ID collisions, and make sure your SCSI chain is properly terminated.  Cables go bad, too, so replacing the cable is another option to try if yours isn’t working.

Once Windows boots, we need to make sure that Windows can see the sampler.  This is a layer on top of the SCSI adapter.  Open Computer Management and then open the device manager.  You should see an entry for the Akai when you expand the “Other devices” tree node.

Akai CD3000XL listed in the device manager means the CD3000XL is on the SCSI bus and the ASPI layer is working.

If you don’t see the Akai listed there but you did see it in the SCSI card screen, that means the ASPI layer isn’t installed correctly in Windows.  Try it again!

Akai's S2000-series editor called MESA II

Assuming the Akai entry is there, you’re ready to begin using whatever software you want with the sampler.  MESA is great for programming patches and transferring samples.  You might also want to use Recycle, Sound Forge, or Wavelab to transfer samples to the machine.

EMU EIV SERIES

In the 1990s, Akai and E-mu Systems had similar offerings in the sampler market.  Akai had its S2000 series, and E-mu had its EIV series, descended from the Emulator line of products that brought sampling to the masses more cheaply.  Okay, maybe not “the masses” (Ensoniq’s Mirage keyboard was probably the best example of that), but E-mu certainly broke the market open with the Emulator.

The E-mu sampler I have is the e6400, which is based on the EIV engine.  Before E-mu was bought by Creative Labs and ceased making hardware samplers, there was one more revision of the Emulator line: the Ultra series.  These had the option of adding USB communication, like Akai’s S5000 and S6000.  But since the version I have doesn’t have USB, I’m relegated to the SCSI bus again.

Getting the e6400 to work with the computer is similar to the above section for the Akai.  Make sure the device is present in the adapter card’s BIOS screen:

Emulator IV on the SCSI bus means the computer should see the e6400

Windows will attempt to install a driver for the E-mu 8 times – each time, just tell Windows not to install a driver and to leave you alone.  You should then see 8 Emulator IVs in the device manager under “Other devices.”

8 entries for the E-mu in the device manager

At this point, the computer should be able to communicate with the sampler from whatever software you want.  Of course, the software needs to support the Emulator IV series for this to work.  Recycle, Sound Forge, and Wavelab all should work.

EOS-Link, a program from E-mu that brings the front panel of an EIV sampler to a computer host

One fun piece of software I don’t use that much but surprisingly still works on Windows is called EOS Link.  This software was created by E-mu, and it provides a way to see what the device’s display reads and to program the device from your computer.  It is a literal translation of the front panel of the device, though, so it’s not as nice as Akai’s MESA.

AKAI S5000

The last sampler I bought new was the S5000, and to me this was a huge step forward.  It had a USB option card on the way, a nice big screen, many filters, and more good options for expansion.  It also uses the Microsoft FAT file system as its standard, which means that instead of messing around with custom disk formats, one can just drop files onto a regular drive and read it on the Akai.  My solution for the S5000 right now is to use Akai’s Ak.Sys program from my Macintosh over USB.  I’m running Snow Leopard, and generally this works just fine.  However, Ak.Sys is an application compiled for the now-defunkt PPC instruction set.

With the new Lion version of Mac OS X, Apple dropped support for Rosetta, which is a machine code translation layer that lets Intel Macs run PPC-compiled software.  This means it’s the end-of-line for Ak.Sys on the Mac.  However, it still runs on Windows 7, so if I ever want to upgrade the operating system on my Macintosh I’ll move the USB connection over to the Windows machine.

Ak.Sys, Akai's software for the S5000 series and newer MPCs. It supports drag and drop file transfer over USB.

The S5000 series dropped support for programmability over SCSI, so the MESA option is gone.  But that’s not so bad, after all.  Ak.Sys is better, and the front panel is actually not that bad on the S5000.  Transferring samples isn’t that bad either, since it reads and writes a commonly-spoken file format.

JUGGLING FILES

All right, so now we have all the samplers talking to computers.  How do we deal with the samplers’ files?  I use Remote Desktop from my Macintosh to connect to the PC over wireless.  When I do this, I can also map parts of my Mac’s file system to appear on the PC.  So as long as I keep the files together on my Mac, I can transfer them over to the sampler of choice pretty easily.  If I’m using big files, I might put them on a CF card from the Mac and read them on the PC.  This doesn’t happen very often, though.

NOTES ON SOFTWARE

I mentioned that you can use Wavelab, Sound Forge, and Recycle over SCSI to these samplers.  That’s true, and there are more pieces of software you can use than just those.  But unfortunately many of those pieces of software have dropped SCSI sampler support along the way.  The last version of Recycle to support SCSI transfers is version 2.0.  I don’t recall the last versions of Sound Forge and Wavelab that support SCSI transfers, so you might have to do some digging if you want to use those.

Boutique Musical Electronics

The old, beautiful ARP 2600. These are selling for many thousands of dollars currently, much more than they were around the time they were being replaced by Yamaha DX-7s

It has never been a better time to be an electronic musician, at least as far as options for composition, sound design, and collaboration are concerned.  After synthesizer manufacturers moved to digital instruments, for a short period of time the older, analog instruments plummeted in price.  Enthusiasts like to brag about finding the $50 Minimoog or a $100 ARP 2600.  I’m sure that happened for some people, but once musicians grew tired of Spartan, single-slider data entry (a la DX-7) the demand for used analog instruments brought their prices back up.

By the early 1990s, very few manufacturers were still making analog synthesizers.  So the used market for the old analogs really heated up: $1000 and above for the old Roland TR-808s and TR-909s, $1500 for the TB-303, and $2000 for Prophet-5s were not uncommon sale prices.  The problem with this equation was simple: no one was making new devices to stem the demand that had risen for the old sound.

That said, what was happening at the time was the rapid growth of digital systems.  Synthesizers and samplers based entirely on digital circuits got better and better.  So did the computer systems, too.  In the late 1990s, affordable digital systems that could not only sequence and track music but also create the musical lines proliferated.  In 1997, The Propellerheads released their classic program called ReBirth (now available for iPad, it turns out), which packaged the most coveted analog classics of the time into a single, self-contained program you could buy for a fraction of the cost of just one of the old machines.

Rebirth simulated two TB-303s, a TR-808, and a TR-909 in a computer program with faithful accuracy to the originals.

Of course, the Internet too was ascending at the same time digital music instruments were becoming more mainstream, and with the rise of the Internet so too rose the communities of musicians, collectors, collaborators, technicians, and engineers working on musical electronics.  Those communities brought people interested in what were probably fringes of musical technology together for local meetups, gear exchanges, discussions, and so forth.  This ushered in today’s golden age for electronic music.

Now, there are literally hundreds (perhaps thousands) of small manufacturers across the globe developing musical electronics and software.  Moog’s famous modular synthesizer of the 1960s has been recreated several times over today.  Standardized rack formats, power supplies, and interfaces allow anyone with an idea and a soldering iron to plug in new sound generators and modulators.

Taking orders over the Internet also makes short runs of esoteric, specialized hardware such as the Monome and the x0xb0x actually feasible.  Many of these new generation of music hackers release their source code and schematics online for free.

Software is not left behind in this revolution.  There are countless developers out there cranking out code to implement new digital signal processors, synthesizers, and sequencers that plug in to each other using common interfaces, like Apple’s Audio Units API or Steinberg’s VST.  One can build a complete studio with multitrack recording, sampling, synthesis, dynamics processing, equalization, mastering, and more using nothing more than a computer and free software.  Using Open Sound Control (OSC) and tools like OSCulator or Max/MSP, one can even connect a Wii remote or an iPhone to the studio for musical expression.

The Monome 40h is a great example of a product that would not have been feasible before the Internet. These are specialized, niche interfaces but have wide visual appeal.

Most recently, we have seen new developments for musical technology in mobile devices, like the ubiquitous iPhone or now the iPad.  Gorillaz frontman Damon Albarn recorded his most recent album using only an iPad while on tour.

The number of musical instrument applications for the iOS platform alone is staggering, ranging from generative music inspired by Brian Eno to serious commercial music instruments from Moog Music and IK Multimedia.

Here at SXSWi 2011, Ge Wang of Smule demonstrated some of the invigorating applications his team developed for the iPhone and iPad.  To be honest, I was skeptical of the iPhone as an expressive musical platform.  Seeing and hearing Smule’s innovations and drive for true expressiveness in their iOS products made me a believer.

Perhaps the only drawback to all the great musical options out there is deciding just where you want to go with them.  New poly analog keyboards?  Old classics?  Piecing together a modular synth using boutique, limited modules?  Writing an album on your phone?  All options are on the table.  To me, that is a very good thing indeed.

On Augmented Reality

Overlaying text on the real world. Image courtesy of HowStuffWorks.com.

This year I switched groups at Microsoft, from the Zune marketplace group to Bing Mobile’s Augmented Reality division.  At Zune, I worked on a lot of engaging technology, including recommender systems based on directed acyclic graph processing (similar to Hadoop) and a continuous playlist generator called Smart DJ.  While I valued and enjoyed the work I did at Zune, I also felt ready to tackle something more nascent in media technology: active processing of the environment on mobile devices.  Thus began my stint into augmented reality (AR).

So aside from the sci-fi term of art, what is AR?  My first exposure of what this technology is appeared in the late 1980s or early 1990s.  I remember reading about a program for the Commodore Amiga that enabled a person to trigger drum sounds by striking virtual planes in the air.  A video camera was used to capture the human figure, and the software defined targets that corresponded to drum sounds.  Whenever the computer detected one of the planes was crossed, it would trigger the drum sound.  This illustrates the canonical definition for AR: a modality that combines the real world and virtual world, is interactive and is a real-time experience, and within it the virtual world is mapped to the virtual world in 3-dimensional space.

Word Lens, an application for the iPhone, translates text automatically

Other examples of AR come from the movies.  Didn’t it seem that every scene that depicted robotic vision overlaid infographics onto a real-time video of what the robot saw?  What was fiction five or ten years ago is coming of age now, thanks to pervasive wireless Internet connectivity, powerful handheld computers (you know, those computers you can also use as a telephone), and advances in computer vision algorithms.  Now it’s possible to use a handheld computer to source and process video, register the video with location and orientation of the camera using GPS, compass, and gyroscope sensors, and communicate with vast computing resources in the cloud over high-speed networks, all in real-time.

This sounds cool, but does the technology have potential?  Some of us remember the heady euphoria of so-called “virtual reality” in the 1990s, with technologies like VRML, backpack computers, and video goggles promising to usher in a new age of human-computer interaction.  In the ’90s, computers were not fast enough or small enough for this concept to work in true real-time, so they ultimately failed.  Users of virtual reality goggles reported the processing lag making them dizzy.

Yelp's monocle feature overlays business information on video captured from the real world. The balloons appear near where the businesses are located.

Forrester Research believes there is potential for AR “to trigger disruption in the years to come and open up new opportunities.”  Some companies are already tinkering with AR layers to drive interest and commerce.  Take, for example, the recently-released application Word Lens for the iPhone.

Word Lens allows you to point your phone’s camera at a sign and translate the words into another language.  Is this AR?  Yes: it blends the virtual world with the physical, is registered in 3D, and is a real-time experience.  Is it useful?  I haven’t traveled yet with Word Lens, but my guess is it could be useful in certain situations.  I think the jury is still out on startup time, though.  On the iPhone, you have to unlock the phone, launch the app, choose languages, and then point the camera at the sign.  Those are a lot of steps to follow, though granted almost all AR apps require about the same routine before you get what you want.

One of the first AR applications for iPhone was Yelp’s application.  Yelp is a crowdsourced review engine, and its mobile app is great for finding businesses or restaurants nearby.  The monocle feature took this a step further by turning on the camera and overlaying balloons onto the physical world with Yelp’s review scores near where the businesses exist.  When you rotate your phone, the balloons rotate with it.  This is great to make sure you’re going in the right direction to that highly reviewed falafel joint.

At Bing Mobile, we introduced Augmented Reality features in the Bing for iPhone application.  Like with Goggles, you can use your phone’s camera to recognize parts of the real world and do things in the virtual world with them.  We introduced a new view in version 2.0 called Bing Vision, and the way it works follows.  You enter camera mode and point your phone’s camera at points of interest in the world.  We have a set of recognizers that process the video from your camera and try to extract information from the image.  For example, whenever we detect that there is text in the image, we flash an indicator telling you we can turn that text into a search query.  Or whenever we recognize a barcode, an indicator points to the barcode and automatically searches for the barcode in Bing’s index.  You can do this with a variety of objects.

Cover art is a great example.  Take a picture of a book, CD, poster, etc., and automatically conduct a web search:

The cover of this book is recognized by iBing

Search results for the cover art

Barcodes are another cool example.  Whenever iBing detects a barcode in the image it displays, an indicator points to the barcode and it automatically searches for the barcode in Bing’s index to find out what it is, how much it costs, where you can find it, and so forth:

The indicator points to the barcode iBing recognizes.

Bing found the Moleskine I scanned and is ready to help me buy another one.

Scanning text is interesting.  Our application finds text in the image and then parses it to allow you to conduct web searches.  In the following example, pretend I’m reading my DSP book and want to find more information on a topic therein, the theorem of convolution.  First I point the camera at the text.

The "Aa" circle you see indicates iBing recognizes text in the image.

I tap the camera button to tell iBing I am interested in searching based on some of this text, so it parses the text using optical character recognition.

iBing is processing the text on the image I selected.

When the text is processed, I can touch words to add them to my search query.

I tapped the words "theorem" and "convolution," and now they are part of my search query.

Hitting the search button looks for more information on Bing.  I like how this works, and it’s usually faster than just typing the words myself.  This is the first version of our text feature, so it is a little limited today.  But we have big plans for it, so stay tuned.

More information about my selected topic, the theorem of convolution.

I had a chance to use iBing this holiday season to actually do something useful rather than just show off my group’s work.  My wife and mother-in-law were shopping with me one day for some audio-visual equipment.  The store was out of the product we wanted, but we happened to see a box on the floor that had a similar model number to what we wanted.  We didn’t see this product in the store at all.  So I pointed iBing’s barcode reader to the box, and it told us everything we needed to know: the model had similar features to the one we originally wanted, it had a slightly more expensive price, and we saw that we could purchase the product at lots of nearby stores that day.  Win!

So does iBing 2.0 satisfy the definition of augmented reality?  It ties the physical world with the virtual world, and it is a real-time experience.  One might say the experience is registered in 3D, but that is a bit of a stretch.  It’s close, and as time goes on we want to get closer and closer to a real-world experience.  Personally, I’m glad to have made the switch into this new world professionally.  There are a lot of fun problems to solve, and I like being on the cutting edge of something new and exciting.

Studio Tour, Part II: Writing and Recording

The proverbial clean slate. Imagine filling this guy up with notes one at a time using a #2 pencil.

I started out writing music by scribbling on manuscript paper.  The staves were there, but no notes or clefs; those were scrawled in #2 pencil after I jammed out the outline of a song.  Of course, this was before I had any synthesizers or a computer.  I wrote all my music at the piano, note by note.  I always wanted a way to scribe and edit music faster, and that desire along with my love of electronic, synthesized sounds destined me for computer-based writing and recording.

Around the time I was using a Korg M-1 to write music, an unfortunate event happened.  I had all my hand-written music in a folder in the back seat of my friend’s car along with some CDs and clothes.  We left for a few hours to go to an event, and when we came back everything was gone.  All the music, and all those countless hours of work with paper and pencil simply disappeared.  They were my only copies.

Going Digital

This device was also not the best interface for creating music. However, it was really cheap and had a disk drive. So at least I could use it to store my songs.

After losing all my hard work in one fell swoop, I essentially never went back to writing music by hand.  I had moved on to writing everything on the Korg M-1, but I still didn’t have a computer.  So that meant whenever the Korg’s memory was full, I had to erase something already stored there to make room for something new.  Eventually I bought an extremely cheap Brother sequencer that had a 3.5″ disk drive.  This allowed me to save all my data from the M-1, erase, and start over without permanently losing anything.  While the Brother unit was technically a sequencer, after trying to write music on it I swiftly decided it was better as a disk librarian.  Somehow, I think even the old Roland MC-4s and digital control voltage sequencers would have been better than that crappy sequencer from the typewriter company.

Inevitable Progression to the Computer

Later in high school, my friend Wes had an Atari ST computer which had built-in MIDI.  I don’t remember what program we used to sequence on that thing.  But it was cheap, and it worked.  Some of our best compositions were written on that Atari, sequencing our Emu samplers and Roland synths.  We sampled everything in those days, including movies and our own other synths.  That was basically the only way we would get the rich, full sound of multiple sound sources since we only had a few synthesizers and drum machines to work with.

My first non-toy computer was the Apple //c, but unfortunately its cool, sleek design eliminated the expansion board buss.  So that meant that even though there were MIDI interfaces and sequencing programs out in the market, I had to wait.  The real breakthrough in my process happened once I had a PC clone running Windows 3.1.  I managed to scrape enough money together to buy a Mark of the Unicorn MIDI Express, which was a 1-unit rackmount device with 6 ports of MIDI input and output.  Using that and a cracked copy (yes, I  will admit it) of Steinberg Cubase, I was finally in full control of my compositions in the computer.  Cubase was MIDI-only at the time, so all my music was written on MIDI tracks in Cubase, which sequenced my entire studio.

The Atari ST series had built-in MIDI and also had rock-solid timing. My friends and I used it to record our first record.

In the early to mid 1990s, I had the computer and MIDI interface for writing and sequencing, the Emu Emulator II which did most of the work in my studio (even with only 8 voices!), my old Roland Juno-60, and a Roland TR-707 for drums.  The next pieces to arrive were the TR-808 and a Korg Wavestation SR.  I didn’t really have any effects processing then, and the old Mackie CR-1604 mixer mixed everything down to stereo.  This setup was modest, but I learned a lot about workflow and digital composition using this configuration.

My process remained essentially the same from those days in the 1990s through the early 2000s.  I began working closely with my friend Mike on a variety of projects, and we both recorded the same way.  We would use the computer as a composition tool, record our takes onto DAT in stereo, and at that point the song was done.  In all honesty, I think Mike was more of a jammer than I was, so perhaps the computer’s influence was more my style.  I loved how I could sequence pretty much anything and have full control over the song’s arrangement.

The Next Step: Digital Audio Workstations

As great as sequencing entire compositions using MIDI was, that process also had its drawbacks.  I began performing my music more, and a lot of the sounds I conjured in the studio were lost forever when the patches and memories of the synths forgot them.  Many of the best sounds were happy accidents, which meant that I couldn’t retrieve them even if I wanted.  Sometimes I would also want to go back and fix certain parts of the mix after the stereo recording was done, but I didn’t have the tracks anymore.

I started using multitrack audio in earnest around the time I was composing the music for my first album.  Cubase had the ability to record and play back multitrack audio probably in the late 1990s, but I never got around to using it deeply until 2000 or so.  ProTools had been around for a while by then, but all of a sudden it was possible for inexpensive computers to do the same work custom DSP chips did in Digidesign’s hardware and the prices really came down.

The top keyboard is the PPG Wave 2.0, and the bottom is its cousin that lives in the computer, the 2.V. Notice how similar the user interface is.

Around the same time, virtual instruments began to gain momentum.  I think the first virtual instrument (or VSTi) I used was an emulation of the old PPG Wave 2.3 published by Waldorf, the Wave 2.v.  I had an actual PPG Wave 2.3 by then, so it was fun to compare and contrast the quality of the real thing against the virtual version.  The verdict?  The virtual version sounded pretty good: not quite as present as the hardware PPG, but close enough.  Plus, it cost about 10% of what a real PPG would.

Going Too Far

I spent a lot of time collecting VST instruments, samples, and software.  I did this so much that I found I was spending more time managing software than actually making music.  At this point I decided to actually buy the software I used for making music and use only what I had legitimate licenses to use.  Part of this decision was motivated by wanting to reduce mental clutter, but part of it was also motivated by the fact that I realized I made my own living writing and developing software.  It felt hypocritical for me not to buy the software I used, especially as plans for my first solo album solidified.

While reducing software clutter, I also decided that since I had a decent collection of hardware that I should use it instead of move too much to the computer.  Pointing and clicking on knobs was too far abstracted from the tactile feel of the hardware.  So as I gradually scaled back my use of virtual instruments, I finally landed on my main workflow for writing and recording music today.  I had gone too far into the computer, and now it was time to strike a balance.

Today’s Workflow

Each recording and projet is a little different, so there is not one simple formula I use again and again.  But the tools and techniques I use have congealed into a cohesive process for me.  Here’s how it works.

I typically start by playing around with synths or drums.  Occasionally I will have an idea in mind before I get started, but not always.  For example, when producing a remix of another track I will strip down everything except the vocals to start working with the main structure of the song.  But most of the time, I just start playing.

All the audio in the studio comes in to the digital mixer, and the digital mixer has a direct connection over Firewire to the computer.  And all the control routings in the studio either originate with the computer or can be recorded and sequenced by the computer.  For those synthesizers with MIDI, I use a big MIDI patchbay to manage their connections to other devices and to the computer.  For those pieces that are control voltage only, I use a Kenton Pro-4 MIDI to CV converter as well as a MIDI/CV sequencer from Doepfer called the MAQ 16/3.  Additional control voltage lines can be sequenced from the computer using the MOTU Volta instrument, which connects to the rest of the studio via a MOTU 828 Firewire 8×8 interface.  The last bit of kit that is neither control voltage nor MIDI is a pair of drum machines, the TR-606 and TR-808.  These are synchronized with the rest of the studio using an old Korg KMS-30 MIDI to Sync24 converter.

This is a screenshot of a track I recently finished in Ableton Live. Lots of tracks!

Whenever I get an idea going that I want to build upon, I fire up the computer to play it for me.  With synths, I’ll play in what I want to record to the computer and build from there.  Sometimes, when I want modulation on the filter of a part, I’ll come back later after the track is laid out and use a knob controller like the Akai MPD-8 to sequence control voltage lines that are patched to the synth I want.

Drums are a little different.  With the drum machines, I usually build layers of 2, 4, or 8-measure patterns together on the machines I’m using.  That usually includes the 808, but lately I have also been using the TR-707 again and the MPC1000 for extra punch.  I don’t have enough inputs on the mixer to patch all the drum channels in at once, so when I’m sketching I will submix these and bring them in on a stereo pair.

Honing the Track

The next step after building up the layers of the song and defining the rough structure is to capture the audio as individual tracks in the computer.  With synths, vocals, and effects this is pretty easy.  I just hit record enable on the inputs I want from the mixer, and let the virtual tape roll.  The computer records each enabled part as an audio track, so I can either switch off the synthesizer (some of them get pretty hot) or just let it sit while the computer plays back the audio.  The digital mixer I use has the concept of record versus playback channels, so whenever the audio is recorded I toggle this mode on the channel to switch the input from analog to the computer.  Recording this way is great, because tracking parts in can be done quickly, and I don’t have to worry about the mix being different if the computer outputs were to go to another channel on the mixer.

This is a screenshot of the same track as above in the mastering stage. Note the ample metering; it's good to have a variety of ways to see the music when finalizing the recording.

Tracking in drum parts is a little different than synths because they are usually mixed together.  For this, I’ll record each drum part on its own or sometimes in layers of two.  Once all this is done, I can then distribute the drums to their own channels, busses, and effects routings if I want or mix them down in the computer to just a few channels.  I like to record changes on the drum machines live too, like mixing in or out sounds using the individual part volumes on the Rolands or muting and soloing tracks on the MPC.  Bringing a variety of patterns in makes it easy for me to mark them and cut them in the computer if I make arrangement changes later.

Once all the parts are in, I typically tweak the arrangement and create the final mix.  At this point, all the audio is being played back by the computer to the mixer.  Sometimes I’ll use some processing on channels in the computer before they come to the mixer, like EQ or compression on vocals.  All the effect sends are controlled by the mixer at this point as well.  Sometimes if I want to modulate pan or a send during the mix I will automate this in the mixer, but I don’t really do that very much.  I roll the sequencer and record the output from the mixer directly back into the sequencer, and then that’s it!  I have the raw mix.

Denouement: Finalizing and Archiving

This keyboard is what I wrote a lot of my early songs on, after the original manuscript theft debacle. It was okay, and I got to know it well. But there's really no comparison between writing using this and writing using a modern sequencer workstation.

Finishing up is a matter of saving patches on the effects units and mixer so I can come back to the mix if I need to.  I’ll export the mix audio to a file on the disk in the best quality I can, usually at 48kHz and 24bits.  Then, I pull the track into Wavelab or Cubase for mastering.  I give my tracks a little extra pop by smoothing the EQ out, giving the bass a nudge and adding a little more sizzle in the highs.  Then, I run that through a mastering compressor and limiter to squeeze a little more loudness out of the track without going overboard.  The final step is to export this processed audio as WAV, then MP3, and that’s it!

Coda: Goodbye Cubase, Hello Ableton

After all these years using Cubase, I think I am finally ready to let go of it.  Around the time Ableton Live version 3 came out, I made an investment in it for my live sets.  Now that they’re at version 8.2, Live is now a very powerful, full-featured tool for composition, DJing, and performing Live.  My fingers still know Cubase, but after getting to know Live a lot better, I don’t really use Cubase anymore.  The king is dead; long live the king!

Studio Tour, Part I: Mixing and Routing

My first recordings were extremely lo-fi.  I started with two tape decks, some Casios, and a Radio Shack 4-channel mixer.  I loved working with that old setup at the time, and I’m also glad I started where I did.  The economy of just a few pieces of cheap equipment taught me how to be efficient and how to work well with what I had.  Over time, of course, I could afford better equipment.  With this series, I discuss the equipment, workflow, and practices I use in the studio and have honed over time.

A little background

The workhorse mixer of the 1990s and beyond, the Mackie CR-1604

From 1993 to 1999, I shared equipment and studios with my friends.  We used the workhorse mixer of the industry during those times, the Mackie CR-1604.  The Mackie was a huge leap forward in terms of quality and price-effectiveness in those times.  The mixer had 16 channels, extremely good build quality, and sounded great.  And the price wasn’t bad, either; I think it hit the streets for about $900 at the time.

In the late 1990s, my housemate and collaborator Peter bought an amazing 32-channel Allen and Heath board that really opened all our eyes to what a good mixing desk could do for us.  It had parametric EQs, fantastic sound, and enough inputs to normal just about everything in the studio to a channel.  We were no longer bound by only 16 channels and only the lone stereo buss.

My first digital mixer(s). They were great for the price but not well-suited for complex mixes.

I eventually moved out of the house we shared at Studio 6250, and unfortunately that meant I no longer had access to the awesome A&H board.  I had to find my own mixer and workflow.  Unfortunately, money was tight at the time.  So was space, so I simply couldn’t afford the nice, large-format mixer I was used to.  I wound up buying two Tascam TM-D1000 digital mixers and chaining them together for my studio solution.  It wasn’t great, but I could finally start multitracking to my computer and mix down up to 24 channels digitally with 8 additional channels of analog.  It was the poor man’s 32-channel Allen and Heath.

I recorded my first album and lots of other music using this configuration, but it had severe limits.  The Tascams had only 4 sends, and the 4 busses were actually shared with the aux sends.  There were only two aux returns per board, and I consumed one of them with the stereo mix of my second submixer.  To make matters worse, the user interface on the TM-D1000 was counter-intuitive and slow.

Modernizing the mixing and routing matrix

The dual cheap digital mixer configuration stuck with me for about 8 years, until 2008.  Over time, I had been collating a set of requirements for the studio mixer upgrade.  I wanted a digital mixer capable of total recall with at least 16 analog inputs.  I also wanted at least 8-buss architecture and lots of sends and returns.  Digital mixers frequently have onboard dynamics processing, and since my old Tascams had this capability, I thought it would be nice to include it.  Basically, I was looking for a board that had all the features I got used to on Peter’s old Allen and Heath as well as those I had on my cheap Tascams.

This is where I spend a lot of my studio time: in front of the mixer. I have a controller keyboard nearby too, along with the subwoofer and nearfields.

Tascam released a new digital mixer after the little TM-D1000s called the DM-24, which was close to what I wanted.  It was fully digital and had onboard dynamics on each channel.  I toyed with the idea of buying one of these used, but then they released the DM-3200, a substantial upgrade.  The biggest reason I decided to wait until I could afford the 3200 was that Tascam produced a Firewire card for it that gave the mixer 32 channels of inputs and outputs over Firewire.  This would be a huge step forward in terms of workflow and reducing complexity for me.  Instead of having to manage a separate audio interface and routing that to analog or digital I/O on the mixer, I could plug one cable into the mixer and have full workstation support without additional hardware.  Plus, 32 channels of simultaneous audio at 24bit depth and up to 96kHz sampling rate far exceeded what I could do before.

Today’s configuration and workflow

An overhead view of the 3200. Like other digital mixers, the faders are arranged in pages: channels 1-16, channels 17-32, channels 33-48, busses 1-16, and aux sends 1-8. There's a remote page as well for controlling the computer.

So, the 3200 came home with me one day, and it is now the control center of the studio.  One of the great things about the Tascam is that the routing matrix is completely programmable.  There are 16 analog input channels with trim, 8 additional analog inputs from an expansion board, four assignable analog returns, and 32 possible channels of inputs coming from the computer.  It also has ADAT lightpipe inputs, which I use as digital inputs from my sampler.  There are 16 busses and 8 effect sends, which can be configured as stereo pairs with pan.  And finally, it also has AES/EBU digital inputs and outputs, which I use for two effect processing chains.

Its EQs are great, with four bands that are completely configurable from low shelf, parametric, and high shelf types.  And the dynamics processing is extremely powerful, with each channel having its own gate/compressor with sidechaining.

This sounds like a lot of parameters, and that is no lie.  The board has memory, and I have set up some templates to keep things sane.  One of the great workflow features I take advantage of is being able to quickly switch each input channel from an analog source to its digital source from the computer.  For example, I compose with a few synthesizers and drum machines on the analog inputs.  When I’m ready to commit to these parts, I record them into the computer over firewire and then toggle the channel inputs for those channels.  At that point, the parts are played back from the computer instead, which frees the synths or drum machines for other work.  I can also bounce the channels down to a submix and reuse the analog inputs if I want as well.

This shows you the patchbays in the studio on top with the modular synthesizer on the bottom. It's nice having all the signal routing in one place, which makes it easy to route signals into and out of the modular. Also, my cheat sheet can be seen on the left.

Once the mixer was the digital nexus of the studio, I also set up normalized routes to it from the important pieces of gear in the studio.  I have more than enough effect sends now, and I keep all my processors fixed on the same routing pretty much.  Two processors for delay and room effects are on the first two stereo sends, and the rest of the processors are either on busses or their own sends.  Everything comes back to a channel on the mixer, and I typically use slots 33-48 for the effect returns.  Almost all the synthesizers come into the mixer on their own analog channels, and sometimes I wind up submixing drums initially until I am ready to multitrack them.

The mechanical nexus of my studio is a bank of patchbays, like most other studios utilize.  I also keep a modular synthesizer near the patchbays in case I want to run a part through more filtering or processing.  It would be great if the 3200 had enough inputs and outputs to route everything in the studio without compromise, but that just isn’t the case.  Plus, sometimes it’s easier to physically route signals using cables than it is to program a routing digitally.  Something I learned from studying human-computer interaction is that when there is too much information to keep in one’s head, embedding that information into the world or the interface one is using helps.  To that end, I keep a color-coded printout of the signals in the patchbays taped to the bays themselves.

I keep all the vital signs of the studio on a virtual desktop so I can switch and quickly see DSP usage, the meter bridge, and so on.

While I’m composing, recording, or mixing, I also usually have a virtual desktop open on the computer that shows me all the vital signs of the studio.  On this desktop I can see the overall performance of the DSP engine, my master spreadsheet of routings, and a view of the 3200’s meter bridge in one pane.  It’s great to see all this information at once and just switch over to another virtual desktop with my audio workstation when I’m composing or arranging.

A nice touch on the 3200 is its transport and automation section.  Starting with the million-dollar super analog SSL desks of the 1980s, high-end consoles began to have their own computers installed to control mix automation.  The 3200 has this as well.  Even though I don’t use the automation much, it’s great to have when I do need it.  The mixer responds to and sends MIDI over USB, which means that almost everything can be recorded and sequenced in the computer.  There are a few exceptions, though, and that’s where mix automation comes into play.  Usually I record alone, so that means I have only two hands to adjust parameters during mixdown.  Being able to record just about anything I would do on the mixer for recall later is perfect for the few times I need it.  Pans, EQ, sends, and of course fader movements can all be recorded by the onboard computer in the 3200.

A view of the 3200's transport and mix automation section

Looking forward

In the next installment of this series, I’ll explain the digital audio workstation I use and how that has evolved.  The mixer and computer are tightly integrated now, and since I almost always worked with a computer for composition, this combination works really well for me.

Music Explosion

I’m finding it harder than ever before to manage all the music I have accumulated over the years.  I suppose I began buying music around age 8, though I couldn’t pay for it myself yet.  My mom or grandmother would take me and my brother to the mall, and sometimes we would get an allowance with which to entertain ourselves.  With about $10 a pop, I could usually afford a cassette or 12″ record at the local record store.  I would come home and practically wear out the music I bought on my little boom box or mom’s turntable.  Some of my first purchases were Van Halen’s 1984, Falco 3, and Kraftwerk’s Trans-Europe Express.

This 16GB Zune HD is probably my favorite music player. It's very small, has a built in HD Radio tuner, and it sounds great. But it won't play my iTunes-purchased music.

In 1987 I got my first CD player, and now the tape and 12″ album collection I had built began to diversify into this new medium.  I also started buying more music at this time, and both Depeche Mode and New Order began to dominate my collection.  Later, when some of my friends and I could drive (and had jobs), we would drive from the suburbs into the city and check out all the myriad record shops for better selections than we could find at the mall.  I also could afford to spend more on music, and that’s when I began to notice that it was getting harder to manage all the pieces of music I could play.

Of course, I didn’t stop there.  In college I began to DJ, both at the university’s radio station (WRCT at Carnegie-Mellon, shout out!) and at local parties and clubs.  At this time, DJing was still all about vinyl.  Sure, the radio station had some CD decks, but they did not have pitch control.  They definitely didn’t have any kind of tactile interface for cueing.  So while my personal listening collection of CDs grew, so did the collection of vinyl I used in my DJ toolkit.

The Diamond Rio PMP-300, my first digital music player. It had 32MB of solid-state storage for music.

The next prong of media diversification in my personal music library happened in about 1996.  The MP3 compression format from Fraunhofer enabled the storage of literally tens of thousands of songs on a server at my employer that we could play throughout the day.  I was excited by how good the compressed songs sounded and impressed at their tiny file size.  It sounds so quaint today, but this was truly a giant leap forward.  Without advances in audio compression technology, we would never have digital music players with such vast libraries that we have today.  Forget about selling digital songs over the internet unless you’re ready to suck down linear-PCM encoded files clocking in at about 10MB per minute.

So as my nascent digital library grew, I thus began to buy digital music players so I could listen to them anywhere.  Sure, like every other kid in the 1980s, I had a Sony Walkman, but these new digital players were different.  The first one I bought was the Diamond Rio PMP300 (1998), which had 32MB of storage on solid-state memory.  Solid state!  That meant I could jog around outside with the player and not have to listen to tape sag each time my foot hit the pavement.  Plus, it was very small. The drawback?  32MB of storage just didn’t seem like enough, even with the new affordances of MP3 compression.

The Special Edition of the Rio had 64MB of storage for music. But it was still not enough!

The next step up from the original Rio was the same thing, only with more memory: the Diamond Rio PMP300 Special Edition (1999).  This one had double the memory, and I also bought a 16MB card for it, bringing its storage capacity to a whopping 80MB.  I ultimately decided that it just wasn’t enough fun to have to decide which 14 songs I want to put on my player.  Sure, I could compress with a lossier bit rate, but that sacrificed the quality of the songs.  If this were truly progress, shouldn’t I be able to listen to good quality and have more than one album on my player?

Somehow I skipped the first iPods, and the first one I owned was the iPod mini (2004), which had 4GB of space.  This was finally enough, I thought.  It wasn’t solid state, but rather it used a new tiny hard drive called a microdrive.  But the microdrive was resilient enough to withstand walking or running, and the battery on the iPod lasted a pretty long time.  The only trouble was, by this time my music library had grown past the 4GB point.  I think I had about 7GB of music at the time, and it was beginning to get tiresome deciding what to put on the player and what to leave off.

My first iPod was one of the 4GB minis. Finally, 4GB seemed like enough space, but somehow the music library was growing faster than music players' capacity.

Since then, my digital music library, listening habits, and music players have all transformed significantly.  I bought a 30GB iPod Video to hold the entire music library on one device.  I began to buy music digitally from Apple’s iTunes digital music store.  And I went to work for Zune to develop music recommendation algorithms and social information processing code — of course, when I arrived at work I got a Zune there, too.  Now, I’ve owned two cellular telephones that also are music players, and I have a 1GB iPod shuffle for working out.  I still have my Zune Pass “all you can eat” music store subscription, which means I have an unlimited supply of music at my disposal.  My iTunes library clocks in at over 8400 items, representing 32 days of music at 82GB.  Oh yeah, and I also have a giant physical music collection of both vinyl and CDs.  I’m feeling overloaded.

Where do I go from here?  I know I’m not alone in this predicament.  I don’t think new music players with more capacity or smaller footprints are going to solve this problem for me.  Tagging, searching, and sorting my iTunes library helps, but the user interface is still more accounting in Excel than it is flipping through records.  To make matters worse, I now feel that my physical music library is just sitting there decaying.  I play records and CDs sometimes, but usually I am just dialing up a song on iTunes or Zune.  I read a lot of blogs, which just blast new music at me 24/7/365.  Smart DJ, Genius, and Pandora are all there to help me find music I want to listen to, and they do a decent job.  But I think what I really need is less, not more: fewer devices, programs, and sources of music — to savor the experiences music provides rather than to be such an avid consumer.  It’s hard to turn off all the voices pushing new music in my face, but I think that’s the only way to stop being a collector and to start being a listener again.

At Zune, one of the most striking pieces of analysis of our customers’ listening habits demonstrated that in general people listened to new music almost exclusively.  Building the histogram of play events against release date was telling.  I think music has been commercialized into a consumption culture for quite a long time, from the content producers to the distributors and also to the electronics vendors that create new formats, devices, and technologies.  But alas, I think now I am going to strive for simplicity.  I might just enjoy all that music sitting around here rather than find new ways to gobble up new bytes.

Even though I have some newish players with smaller capacities, overall the amount of music they hold is growing. Interestingly, though, I think people are starting to find they just don't need 100GB of music in a portable device. I'd guess that the days of the Microdrive devices with >100GB of storage capacity are numbered.