The Stack of Dependencies

I've been making music with computers for... well... a fairly long time. In that time, I've endured many different variations of document archival integrity.

This issue is more acute for musicians and media artists than it is for most people. The problem is a stack of dependencies. We open our document in a DAW application. Usually, there are a number of plug-ins and virtual instruments that we use in our documents. Some of these applications and plug ins may require additional copy protection software to operate. The application itself requires a certain version of operating system. That operating system only operates on a certain range of computers. The DAW software may use audio hardware. The hardware requires drivers that are dependent on the OS version and computer hardware, etc, etc...

Each of these dependancies have their own life span and release schedule and they're all interrelated. Update your computer or operating system and you may be surprised that some upper functionality get depreciated. In my lifetime, I've seen the transition from MFS to HFS. From the first Mac OS to OS7 to OS9 to OSX. From serial and ADB to USB. From floppy drives to nothing. From the Apple External Drive Port to SCSI to Firewire. From 68K to PPC to Intel. Each one of these transitions demanded changes to everything running on top and in a lot of cases, some software didn't survive the transition.

Some changes are self-drected. At one point, I was using Performer, but got really interested in Max and the people around me were using Vision, so I moved to a total Opcode system with a Studio 4 interface. I synced this to a couple of ADATs. Then I moved to Pro Tools hardware with Studio Vision. Needless to say, accessing those Studio Vision documents today is difficult.

It is enough work to get a complete production system running, but near impossible to maintain a system with 100% backwards compatibility.



But really, the context of this post came from thinking about virtual instruments versus hardware. If you have MIDI data, you can play it to a virtual instrument, or you can play it to hardware. Now, forward 15 years. You can take the hardware out of the closet, play the MIDI data to it and it works. If you're using a virtual instrument. Well. Good luck.

So, what to do?

First of all, you need to render everything you have as a broadcast wave file when you archive a project. This includes virtual instruments and tracks with effects. Sounds pretty simple, but in practice it is more complicated. And a supreme pain. If you're done with a project, you just grit your teeth and do it. But, what about projects in development and those that never get finished? At what point do you shift into archival mode?

Another strategy is to limit your dependence on third party plug-ins and virtual instruments. This, obviously, sucks.

I've had some luck with accessing data with virtualization. Obviously, this isn't going to work on hybrid hardware/softtware systems. That is, unless someone comes up with a virtual ProTools III emulator, for example.

I have friends that archive their computer hardware. Sort of like freezing it in a preserved state. This is a good option if you have the space the store the equipment, and if you're planning a major shift in process, like switching DAWs. Most people I know sell off their old gear when they're done with it, though.

What are your strategies? Do you have any horror stories?

Siftables

David Merrill from MIT provides a short demo of 'Siftaables', intelligent blocks that are aware of themselves and others. Watch the video and think about the musical applications of this invention. At the end, he demonstrates a sequencer. Can you imagine the crazy stuff you could do with this and Max?

The Physical Loop

After sourcing all the parts, reading a lot of information, and receiving an eye-popping estimate from a Rhodes technician, I've decided to undertake the Rhodes restoration myself. It'll be quite a project, but the end result will be awesome.

My current Volta immersion has got me thinking about all sorts of analog/software hybrid synthesis ideas. This is a very fertile area to explore and there are many interesting possibilities.

But, back to the Rhodes. Why am I even doing this? I mean, aren't there some decent software Rhodes emulations on the market? Why have I romanticized this heavy, temperamental thing?

I could have certainly used some great piano sample libraries on ATAOIB and Brood XIV. It would have been much easier. I could have recorded any time I liked. I'm fairly certain it wouldn't have turned out the same. The defining limitations of the project were attractive to me. Even if I restricted myself to recording audio, instead of MIDI, there are a number of quirky aspects to the recordings that never would have happened. The chirping birds, for example.

Most people would go back to the quality of the sound. They would argue that a detailed sample of the real thing isn't the same as playing the real thing. I'm not going to go there. My argument about using a real piano, and a real Rhodes is act of playing it.

When you press a key on a piano, a physical action takes place that you can feel. You can press the key slowly and feel the action of the hammer. This physical feedback informs your playing like a closed loop. Playing a MIDI controller does not provide this feedback, it is playing by remote control.

This got me thinking about the Korg Wavedrum. For those that don't know, the Wavedrum was a small electronic hand drum with a real head that used the output of an array of internal microphones to feed a DSP. The DSP then translated this information into different drum sounds. You could play it with sticks. You could play it with brushes. You could play it with your hands. You could even lean up to the head and shout at it. The DSP faithfully translated the incoming noise as if it were real.

Strangely, the Wavedrum died on the vine and now they're a coveted rare instrument. I'd like to see this approach used on a keyboard instrument. Start with something like a Rhodes with a real keybed, real hammers that strike a physical thing that vibrates. Feed the output of each key's vibration into a DSP. This would be the foundation of an extremely expressive, and flexible instrument. Think of it as a combination of a Rhodes Mark 7 and ten or so Roland VG guitar processors. These days, you need to offer consumers something in hardware they can't get in software. Doesn't get any better than this.

I think the market is tired of the same technology recycled in different packages. You can add a color touchscreen or more polyphony or V-Beams to so many products before it becomes noise to the consumer. People are thirsty to innovation.

Are companies out of ideas or are they just too conservative? What do you think? Am I making sense here? Is this a product idea you would like to see?

Rhodes to Nowhere

16 years ago (ugh), I had a lovely Rhodes Stage 73 mk I. It played beautifully, was complete with all its parts and was in very good condition. MIDI, samplers and the DX7 had pretty much killed the Rhodes market and no one wanted them. I loved mine though. Sadly, when I moved from Boston to Texas for graduate school, there simply was no way I could take it with me, so, I sold it.

I envy the person who bough my Rhodes. I envy the person who is currently using it. I truly, truly regret that decision, but there was nothing I could do about it at the time. It was the right decision too because I moved another five times after that. The Rhodes would have been a boat anchor.

I figured someday I'd be able to correct this issue. Since then, ebay happened and interest in vintage keyboard skyrocketed. Even something as heavy and unwieldy as a Rhodes appreciated in value. Today, with the economy in tatters, and the new Rhodes Mark 7 on the way, I figured I'd make a serious stab at trying to find an old Rhodes. My urge to own this instrument again outweighed whatever inevitable grief I would endure from my wife about 'more junk in the house'.



I found one last night, and I brought it back to Cambridge, but it is in pretty sorry condition. It is missing the lid, sustain pedal, push rod, one cross bar, and six hammer tips. All the tolex is gone, and the bare cabinet wood is painted red. This is pretty far from the ideal Rhodes I originally had, but it is a start. Besides the hammer tips, it isn't really in playable condition, so I'll have to have a tech go over it.

I fall into the camp of eschewing 'improvements' to vintage instruments. I'm not too keen on unnecessarily altering the appearance and prefer to keep matters as close to original specifications as possible. I know there are people who will disagree with me, but when I personally think, for example, custom white Matrix 12s are an abomination.

When I bought my Prophet 5, it was also in pretty bad shape, but after some time and money, this instrument is in excellent condition. Assuming there is some future fix for the unobtainium inside the Prophet, and this is admittedly uncertain, I like to think that the next person who uses it will receive it in better condition than I did. This will be certainly true for the Rhodes.

All this reminds me that we don't really own musical instruments, as it is quite possible they'll outlive us. We're merely temporary stewards. Yes, a majority of instruments end up broken and discarded, but that isn't really necessary. A well cared for instrument can be handed down to the next generation. So, take care.

Creativity at the Speed of Life

As far as skills go, photography is a fairly recent acquisition for me. I had dabbled here and there, but everything changed when I got a Canon 20D with a EF 24-105 f4.0 lens. At that point, I became obsessed and went on a learning spree.

In a span of about three months, I went from not knowing what an f-stop was to doing product photography. I was very aware that the process of learning was accelerated, and this was due almost entirely to the internet. I could go to flickr and see what lenses and techniques produced what results. I could inspect the metadata of digital photos to see their shutter speed and aperture settings. I had instant access to any information I wanted. I could talk to people online, ask questions. The technology of digital photography helped as well. Instead of waiting for my photos to be developed, I could see the results instantly. I wasn't burning money taking and developing bad photos, there was no financial penalty to dropping the shutter.

If I were taking up photography in the early 80s, this process would have been far more gradual. I would have to attend enthusiast meetings to talk to other like-minded people. I would have to wait for the monthly photo periodical to be delivered with the latest news. I would have to go to the library to find books and do research. The process of learning would have evolved much more slowly.

As a post-internet parent, I'm finding I can strike when the iron is hot. When my child has a question, we can look up information, see pictures, and find video at that moment. Learning is accelerated greatly when materials are an instant away.

It wasn't that long ago when a recording artist could spend a year or more crafting a recording. There was a lot at stake, releasing a record was an event. You're only as good as your last record.



Today, we expect more. We expect to be entertained daily. Or minute to minute. With RSS feeds and twitter, entertainment acceleration has hit warp speed. Consider these recent memes and how quickly they evolved. I'm not going to embed these inline as it would be an absurd amount of video.

The Christian Bale freak out (Feb 02)
The Christian Bale freak out remix (Feb 02)
The Christian Bale freak out remix with new video (Feb 05)
David After Dentist (Jan 30)
Chad After Dentist (Feb 07)
The Christian Bale / David After Dentist mashup (Feb 05)

Our entertainment used to come from (or at least was backed by) large corporations. Now, we're entertaining each other. We are all artists, and we are all consumers. This is a major contributing factor to the slow death spiral of the record industry. The scale of a record production is something that can by accomplished by a small number of people, cheaply. A major studio motion picture is less at risk because that sort of spectacle requires the coordination and dedication of hundreds of people.

At the beginning of The Work of Director Michel Gondry, Michel Gondry states that he opted for quantity over quality because quality fades and quantity stays. I'm not sure how serious he was about this statement, but the point is our idea of quality changes - it is fluid, and quantity is concrete.

If you are an independent artist serious about getting noticed, I think you need to re-calibrate how you expose yourself. You need to provide a constant level of noise. Instead of a period of silence, followed by a release of a clump of songs, you need to involve your audience in the process. Talk about what you're doing. Post text mixes and sketches. Engage your audience daily. If you're off the radar for weeks - or a year - then release your opus in one clump, the thing that you've spent a year on is competing for the same level of attention we reserve for a twitter post. I think there is a lot to discuss here, so please tell me your thoughts.

A different way to think about creative genius

Elizabeth Gilbert talks about writing, success, anxiety and emotional risk. Her delivery is evocative of Laurie Anderson's voice to me.

Amazon Kindle 2... but why

Steve Jobs said the Kindle is flawed because people don't read. On the surface, he's right. But, he's also not Amazon. Maybe people don't read books, but if they weren't buying books, Amazon wouldn't have a reason to exist. Before Amazon, if you wanted to buy a book, you went to your book store, paid and walked out with it. There was a tangible benefit to book cover artwork. Amazon has largely killed the local book store, and most books are bought, sight-unseen, and shipped to the consumer.

If the act of physically handling the book doesn't factor into the purchase decision, and the primary utility of the book is the data it contains, why are we still shipping books? Amazon is watching millions of books shipped to them, then shipped out of their warehouse. Tons and tons of pulped trees. How much energy is used to deliver these things? Can you imagine a more inefficient and wasteful method of delivering such a small amount of data? I'm surprised I don't see more people touting the green aspect of the Kindle.



The other thing I'm surprised about is the lack of understanding about e-ink. The Kindle doesn't use an LCD screen. It uses e-ink. E ink does not require backlighting. An e-ink image will persist on the screen after power is removed. Therefore, the Kindle can operate up to two weeks before a recharge.

So, I'm behind the technological benefits of the Kindle concept. What I'm not behind is the use of DRM. Amazon's music store has DRM-free downloads, so why do books have DRM? The reason I'm concerned about DRM is I don't want my investment in media to potentially die with the technology. If I buy a book, I know for certain that 10 years later I can take that book off the shelf and start reading it. I'm not so confident about an electronic file with DRM.

Why a Kindle book is better than a physical book from Amazon
1) instant gratification - downloads in seconds

2) travel - you can carry 20 (or 1500) books in the space of one

3) you don't have to pay for shipping
4) kindle books are cheaper

5) you don't have to store the books when you're done

6) doesn't require pulping trees

Why a physical book from Amazon is better than a Kindle book
1) initial expense of the reader

2) doesn't replicate the experience of reading words off dead trees

3) can't loan a book to a friend

4) DRM

Why a dedicated book reader is better than a netbook, laptop or iPhone
1) no power tether (e-ink)
2) low battery consumption (e-ink)
3) easy to read in sunlight (e-ink)

Why a netbook, laptop or iPhone is better than a dedicated book reader
1) color screen
2) don't need another gadget to carry around
3) the expense of a dedicated piece of hardware

Will you be buying a Kindle?

More Expressive Control

The French Connection is a controller designed to emulate a couple of the signature aspects of the Ondes Martenot, an early electronic instrument designed in 1928 by Maurice Martenot. The Ondes Martenot was actually a system of components, including some specialized loudspeaker elements that shaped the electronic timbre from the instrument. The speakers are as important to the sound of the Ondes as the Leslie speaker is to the Hammond B3.

Note articulation is decoupled from pitch selection like a violin or guitar. Notes are articulated with a special spring loaded button on the left hand side. Pitches are selected on the keyboard, or from the ring controller which allows seamless glissandos. It is easier to see than describe in words.



The French Connection takes the note articulation and ring controller aspects from the Ondes Martenot and packages it as a CV controller. The French Connection does not emulate the fantastic manually-articulated keyboard-based vibrato as it simply uses a standard OEM keyboard. Even so, the French Connection does not supply a velocity or aftertouch CV output which would have been very welcome on a CV controller - especially at what a French Connection costs.

You can emulate the decoupled note articulation of the French Connection with a little ingenuity. For example, a Roland Juno-60 allows you to articulate the VCA with the left hand controller. You can do the same with any modular and some appropriate controller.

Wendy Carlos wanted huge glissandos with precise pitch control, so she made something called the 'Circon' The Circon is sort of like playing an oscillator knob by hand, but with the physical articulation of the knob itself widened a great deal, and mapped against a circular graphic of a keyboard so you know where you are. Wendy would record the Circon in a clever way. The output of the oscillator gets fed into a Fairlight Voicetracker pitch to MIDI convertor (a nice piece of kit in its own right). The MIDI output gets recorded into a sequencer, and sent back out to a Circon patch on a Kurzweil K2000. Ha!

Wouldn't it be great to take what you can do with the French Connection and play it polyphonically? How about adding another dimension of expression as well? That is the dream, and it is made real by Haken Audio. The Haken Continuum Fingerboard provides three dimensions of expression continuously across the keyboard, for each finger, over time. Whew. Draw a bunch of piano key size rectangles on a piece of paper, and imagine pitch as the x axis. You can select a pitch, then glide up to another pitch. Now, the pressure you exert on the key (remember, not just initially, but continuously) can control another parameter, and the y axis of the key can control another. That is a lot to keep track of just for a monophonic line, but polyphonically? Wow.



As you can imagine, the Continuum generates a lot of data. Yes, it supports MIDI, but the Continuum also supports a special protocol for communicating with Kyma. It also supports an external 16 output control voltage convertor box so you can get up to four notes of X Y and Z output. Obviously, this is a company that 'gets it'. However, I could find no mention of OpenSound Control support at the Continuum web site. Why make a controller capable of generating such a rich set of data and then cram it through MIDI? MIDI intro Reaktor? Really? I hope I'm mistaken about the OSC support issue, or this is something they're working on.

If you want to be blown away by what the Continuum can do, check out this example. Unbelievable. You must watch this video.

Expressive Control

With the focus on the hardware side of the synthesis market shifting to controllers, we're beginning to see some interesting developments.

I'm very excited about the VAX-77 keyboard. The most visible distinguishing feature and top selling point is the one I care the least about: the fact that it folds in half. The VAX-77 folds in half to qualify as a carry on for airlines. Apparently it is difficult for backline companies to supply MIDI controllers. The folding feature certainly qualifies as a nice touch, though. I definitely see a benefit for fitting the keyboard into small cars.



No, I'm excited about the VAX-77 because it is the first keyboard (that I'm aware of) since the venerable Kurzweil MIDIboard to provide BOTH release velocity and polyphonic aftertouch. Even the Korg Oasys doesn't do this.

The disappearance of polyphonic aftertouch is curious. It was difficult to implement, and consumers didn't seem to demand it, so it went away. Pretty much everyone started to source their OEM keyboards from the same supplier, and that was that. Infinite Response is doing something different - building their own mechanism, which must be very expensive. Real analog with aftertouch is brilliant. It sounds like instant Vangelis.

The disappearance of release velocity seems even more odd, because I don't believe it to be much of a technical challenge for keyboards that already support initial velocity. Regardless, I'm glad to see it return. The first time I ever played release velocity was on a Matrix 12 and it was amazing. It took a couple minutes to get used to, but the results were fantastic.

Keyboards with release velocity
Novation Remote (and possibly other novation products)
Peavy DPM C8
Oberheim Matrix 12
Kurzweil MIDIboard

Keyboards with polyphonic aftertouch
Roland A-50/A-80
Yamaha CS-80
Kurzweil MIDIboard
Ensoniq EPS/SQ-80/TS-10/12/VFX and others of that vintage

I like a nice piano action controller when I'm playing piano sounds, in which case, I'd prefer to play an actual piano. For everything else, I prefer an extended (76-key) weighted synth-action controller. Playing clavinet or drums or synths feels exceptionally weird on a piano action keyboard. So far, the best controller I've found for myself is the Roland A-50. I like the way the keyboard feels, it has polyphonic aftertouch and both a paddle and wheels. I also augment this with a Kurzweil Expressionmate ribbon controller.

I hope the VAX-77 is successful, because I think there is room for Infinite Response to expand their product line. We'll know more after they start shipping. If you know of other keyboards that send polyphonic aftertouch and/or release velocity, let me know in the comments.

Twitters You Should Be Following

Here are some music-tech people I enjoy following on twitter. If I forgot you, I'm sorry. Please, if you have some additional suggestions for followers (including yourself), please do so in the comments.

MODULAR FRIENDS
Surachai and Justin also run the Trash Audio blog
Brandon runs the Bleepology blog and tries daily to curb himself of gear lust.
Noisesource works in the film industry. For fun, he goes out to injure himself with his longboard. Also, check out his blog.
Dave Jones breaks his modular when he's not working on the FREAKING LINUX KERNEL.
James Cigler is exceptionally knowledgeable and is well-known for producing the web's SECOND-BEST modular videos.
Wetterberg is a music hacker's hacker, mucking with modulars, max and anything else filled with awesome.
Parasitk actually plays out and stuff.

MUSIC TECH BLOGS
James Lewin of Synthtopia always posts interesting, relevant items. He also lives in my home town.
Peter Kirn - extremely nice and knowledgeable. He is so knowledgeable, in fact, he needs two twitter accounts.
Music Thing may be dead, but the twitter account lives on.
Ladies and Gentlemen, Chris Randall and his orchestra.

MONOME FRIENDS
primusluta
corporation
modernscience
jonbro
firebrandboy
TRACERfire
HelgeG
madebyrobot
aldaraia
mapmap
toiletooth
rwbingham

TWITTER TOOLS
Me
Follow teh funny with favrd
Search Twitter
Follow subjects with hashtags
TwitterSheep produces Twitter tag clouds
I use Twhirl as a desktop twitter client
I use Twitterrific on my iPhone

Not What Ships Are Built For

Edison has released a free, downloadable album. Make your ears happy.

Why Use A Modular?

As Chris Randall, a recovering modular owner points out, modulars are terribly inconvenient. If you're looking for maximizing musical productivity, a modular is possibly the last route you want to go. Why would you, as an artist, intentionally handicap yourself?

It feels honest
Today's synthesizers and software tools are extremely powerful and amazing sounding. In fact, in a lot of cases, you don't really have to play them at all. You can call up a preset, use the manual (or this) to hold down some keys and they'll play themselves. If I were a composer for film and television and needed to deliver product on a tight schedule, I'd be all over these tools. They make you sound brilliant, in no time flat. But, I don't personally find this artistically satisfying. I don't see it as much different from pasting together a composition using a loop library. Remember the recent blogostorm over the Calling Berlin sample shipped with Fruity Loops?

Why throw the baby out with the bath water? Why not force yourself to only use self-created sounds, or initialize all presets? I'll try to answer this question below. The honesty aspect is only a portion of the complete issue.

More fundamentally, why must electronic musicians shoulder the additional burden of being synthesists? How many other musicians are required to build their own instruments for legitimacy? I'm not suggesting everyone who makes music with synthesizers needs to be a synthesist and make their own sounds. I'm speaking only from my own sense of artistic satisfaction.



The latest technology dates itself
What I've found over the years is the cutting edge, complicated sounds of modern instruments don't age well. They tend to betray the time period in which they were created. You can hear this a lot in film scores. There was a trend from the late 70s to early 90's where productions on a budget went a bit heavy on the synthesizers. After all, it was a lot cheaper than hiring a bunch of union scale musicians. These scores stand like monuments to the synthesizer technology used at the time. I think the shift back to traditional orchestral scores (or, in some cases, orchestral sample libraries sweetened with smaller ensembles of real musicians) reflects the fact that synthesizer technology, like hair and clothing, is a form of fashion. The conspicuous synthesizer that sounds modern and fresh today, is tomorrow's mullet and leg warmers.

Simple sounds are more expressive sounds
The whole point of a new software synthesizer is new sounds. New sounds tend to get more elaborate and complicated in order to impress people enough to buy the product. I find all this sound-design pageantry gets in the way of musical expression. How easy it it to convey your unique, personal vision when you have 16 arpeggiators squirting along while 100 wave sequenced oscillators are modulating burning pianos? The sounds they produce are so intricate and complicated that they leave little room for... a performance.

In imitative synthesis, it is usually easier to do an ensemble than solo instruments. Have you ever tried to synthesize a solo violin? Wait, back up. Have you ever tried to play a violin? Go ahead, drag a bow across a string. Pretty disappointing. If you're lucky, you'll get a basic waveform out of the instrument. That is pretty much all there is to the 'sound' of a vioin. For a violin to sound like a violin, you have to know how to play it. Same with a synthesizer.

I believe there is an inverse relationship between how complicated a sound is and how expressive you can be with it. A simple sound, with many vectors of expression sounds more like a musical instrument to me than a complicated sound bed triggered by a key press. The test is really simple, when you hear it, do you hear the instrument or the musician? Can you tell the difference between two different musicians playing it? This is why I love the French Connection. The basic sound is nothing anybody would want to listen to by itself. Take the same sound and add human input, and suddenly we have something expressive, personal, musical and emotive.

Self-directed upgrades
Adding a module makes the entire instrument new again. You're not simply adding the capabilities of the new module to your system, you're adding the capabilities of how that new module interacts with everything else - it makes everything new again. When you buy a monolithic synthesizer, it never changes - you're stuck with what it is forever.

It is true that some systems these days get new features via software upgrades. The difference is upgrading a modular is self-directed. You don't have to wait for a software upgrade - you're not at the mercy of developers. If you don't like a module, there is always someone waiting to buy it. These incremental changes are relatively inexpensive compared to the cost of a new synthesizer. As a single entity, my modular has held my interest longer than any single instrument.

Failure
A modular never gets completely borked. If a module fails, you send it off for repair. The rest of your system continues to function.

The Shiva factor
I think the first 11 years of my life were spent surrounded by a pile of Legos. Solid. I remember Lego play occasionally punctuated by school or sleep. Modulars are synthesizer legos. You can build cool things, but if you want to build something else, you have to tear your existing cool thing apart. I find the destruction/creation cycle inspirational.

Fun
Finally, my modular is fun. I've had more fun with my modular than any other synthesizer I've owned in nearly 25 years of doing this. Is a modular for everyone? Hell no. but I'm having a blast.

What do you think? Do you use a modular? Are you interested in modulars?