this space intentionally left blank

September 5, 2006

Filed under: music»tools»digital

Top of the line

Although it seems to be working well in my recording tests, I'm not really convinced that a laptop-based solution is going to be appropriate for playing live. For one thing, I think the laptop gives the wrong impression--that I'm triggering samples, or that it's setting up my loops for me. With stompboxes and their relatively crude interface of footswitches and knobs, the theatricality of the live performance is preserved, which was a large part of the motivation for my pretentious solo project in the first place. Besides: dork with the laptop.

In any case, I'm surprised that there seem to be very few devoted audio notebook builders, other than Rain Recording's LiveBook or NUSYSTEMS in the UK. Alienware also makes a desktop chip-based "pro audio" laptop, but I'm guessing that's a bit hotter and high-power-consumption than I would want to work around. Beyond that, you're left with the usual suspects: MacBook, Lenovo Thinkpad, Sony Vaio, or other general-purpose machines.

I've been trying to think about what I would want in a live-rig laptop. I feel like there are two basic priorities: latency and durability. Screen size isn't critical for my needs, battery life would be nice (although clean power is more important), and the keyboard obviously wouldn't see much use. A huge factor for latency would be the interface--and here's where I think it's interesting that laptop audio hasn't really advanced past the Soundblaster days. Video output has become increasingly sophisticated, with upgradeable 3D cards and different ports, but it would take a miracle to find a laptop that could handle high-Z inputs (which are compatible with the load resistance of guitar pickups). Still, that's what USB interfaces are for, even if they add bulk and complexity.

I'm sure that if you have roadies and money for backups, it's worth taking a laptop into a rock situation. But I've never played that kind of crowd--I've played coffeeshops and seedy college bars. I worry about pulling it off the stand by the instrument cable, or some drunk spilling a drink all over it (the Lenovo's Thinkpad keyboard actually incorporates drains for liquid, which might help). Especially for non-DJ musicians, where audience members sometimes wander onstage and mess with the instrumentation, there are a lot of unpredictable factors to consider. No, I think I'm better off with dedicated pedals for live work. I'm tempted to reconsider when I have some money sitting around--but until someone offers an integrated pro-level audio interface in a durable shell (to make the whole package as simple as my stompboxes), the cash is better spent on analog hardware or a new bass.

September 4, 2006

Filed under: music»tools»digital


This Axon demo recorded for is incredibly cool: it's a realtime pitch-to-MIDI converter that handles chords. Plus it can apparently detect where on the guitar you're playing, right or left hand, and change patches accordingly. So the top half of the fretboard can be a different instrument from the bottom, and each pickup can be a different instrument. Towards the end, the demo presenter plays a complete MIDI jazz trio with drums, bass, and Fender Rhodes organ, all triggered from what looks like a PRS guitar. They say it'll work almost as well on a bass as on a guitar.

Of course, it also runs $1,200, and it still requires installing a special hex pickup that listens to each string individually. It costs a lot to live this free, Wayne.

August 30, 2006

Filed under: music»tools»digital

Building a better distortion

Not a single effect has occupied my time as much as distortion in the three years that I've played bass. Chorus? Set it thick and shiny, then forget about it. Envelope? As long as it quacks, I'm not particular. Slap? Crank the compression and drop the mids. But getting a good fuzz bass has been like Pulling Teeth, pardon the pun.

Even for a traditional bassist, I think distortion is a challenge. Guitarists get all the best high-gain pedals, but those same effects tend to filter out the low frequencies, with the end result that the bass disappears in the mix. Dedicated bass pedals started adding a blend knob to combine dry signal with the overdrive. For some settings, this works--on others, the Blend knob doesn't really blend, and it sounds like two instruments playing simultaneously, or even worse, a horde of bees following a weak bass around.

For players who just want a little growl, the sound is easier to achieve. I hear a lot of good comments from Sansamp Bass DI users, which can create a growling tube amp sound. But I'm not a low-gain player. Even when I played more conventional bass roles, I wanted to be able to fill a huge space of sound. Nowadays, I need a distortion that can be a jack of all trades--good on single lines or chords, giving me a full-bodied low end as well as a sharp high fuzz. I could pretty much get that from the MXR M-80 live, but when recording it was hit or miss. And since I record my loop performances to a single track, I needed it to consistently sound right without a lot of extra EQ.

Moving to the VST effects rig was the perfect chance to be able to create my own distortion from a blend of the many free plugins available, and I spent three nights tweaking it until I could get a good result in the mix of several songs. I'll post a cover of something in a sketchpad as an illustration as soon as possible. Here's how I cooked up an overdrive that doesn't make me cringe.

First of all, I knew that the best bass distortions usually involve at least two different amps blended together. Tim Commerford, who may have the best single-line bass sound in rock, uses three amps with increasing amounts of dirt in each. John Entwhistle, whose bass sound was groundbreaking, split the sound by into separate high and low frequencies, then ran them through a mix of guitar and bass amplifiers. So my first goal was to achieve the same thing through two different plugins, EQ'd to fit together into one big sound. In Phrazor, you do this by using the two busses for each effects track--one receives dry signal, and the other is wired in series, with separate input volumes for each.

For the bass half, I used Audio Damage's Fuzz Plus 2. The Fuzz Plus actually doesn't do a great fuzz in my opinion--it's more of a very heavy drive. That's perfect for a growling low end that'll support but won't stick out too cleanly. I dimed the fuzz control, but cut the tone all the way down (Phrazor tells me it's set at 400Hz). On top of that sound, I ran a second signal chain through the series bus with a Tube Screamer into a Marshall JCM 900 amp emulation, both from the Simulanalog Guitar Suite. The Tube Screamer gives the amp a bit more oomph, just like its real-world counterpart, and the Marshall preamp has its lows removed and its mids severely cut. That gives me a wide high-end fuzz to sharpen and define the lower frequencies.

Balancing those two signals together gave me a full, uniform fuzz, but it was still a bit too digital-sounding. That's happened to me before, like when I tried to crank the overdrive for Sketchpad 6 ("Strange Chemistry") and ended up with harshness and clipping. To tame the harsh high frequencies without blunting the amp's sound, I added a low-pass filter after the amp simulation. A low-pass does exactly that: it lets frequencies below a certain point through, but anything higher is removed. The filter is set at around 9KHz, which is very restricted (most people can hear past 16KHz). Normally, a filter set that far down would make the amp sound muffled as all the treble was removed, but I also raised the resonance, so that frequencies near the cutoff were actually boosted. The harshness was still far enough past the cutoff to be eliminated, and the resonance accentuated particular thicker parts of the fuzz. The end result was a classically scooped high-gain distortion, but one that has lower gain at the bass end for a meatier tone. I could probably still spend weeks or years tweaking it, but it's a solid and reliable foundation for recording.

...I hope.

August 22, 2006

Filed under: music»tools»looping

Stripping Mobius

After two days of work, the process of virtualizing my pedalboard has taken a big step forward with the emulation of my Line 6 DL4 in Mobius. You can get the scripts here, and Mobius itself is a free download. The process of moving my effects rig to a laptop may be a really bad idea, but it wasn't even a possibility until I got this running.

The essential functionality I missed from the DL4's looper is the Play Once button, which I think was inspired by the Boomerang loop pedal. If Play Once is pressed while the loop is active, the mode is armed and the loop will stop playing after the current cycle. If the loop isn't active, the loop will play once (huh!), then stop. Pressing the button while Play Once mode is armed will stutter the loop, restarting it from the beginning of the cycle, kind of like a skipping CD.

I hadn't realized, until I began looking for software alternatives, how much I've used these functions in my compositions and covers. On "My Foundation," for example, I use it to trigger the chords in the chorus. In a lot of songs, Play Once sets the loop to end, while I simultaneously switch my other effects off or on--something I otherwise wouldn't have enough feet to do!

Now, it's not clear if I'm an oddity or if I just haven't transitioned fully to the dominant looping paradigm, but most other people with this kind of equipment don't seem interested in these functions--and it shows in the tools they write. Loopy Llama is a fine enough plugin, but it's clearly been coded by someone who has either more equipment than I do or a very different musical approach. Mobius itself is itself an emulation of the Gibson Echoplex Digital Pro unit, and so it doesn't even technically have a "stop" function, just different modes of "mute."

Luckily, you can script Mobius, and it has an almost obscene amount of capabilities. The difficult part is overloading one button, so that it will trigger different effects depending on where the user is set, at the proper time (hardest of all). So while the other two are nice (having Record and Overdub on the same button saves space, as does Play/Stop), it's the Play Once button that caused me the biggest headaches. But I'm proud to say that it should work properly now, assuming that you let it use the InsertMode variable to store its state, and you don't have some weird quantize settings. It wouldn't be hard to change the script to handle that, but this is a good starting point.

August 20, 2006

Filed under: music»tools»digital

Pedalboard Update

I now own a used MIDI pedal. I am the dork with the laptop. But unfortunately, Niall's Pedal Board seems to take issues with the window manager in Windows 2000--at least, that's the only reason I can think of that it would work on an XP machine, but none of my Win2K laptops at home. And since I'm skipping the upgrade cycle until Vista, that means it's not very useful.

As an alternative, I've been playing with Phrazor for a couple of days, and it seems to do the same kinds of things that Forte does, but for much cheaper. Right now, fully functional betas are free to download, and you can pre-order the final version for 29 Euro, or about $40 (much cheaper than it will probably be once released). It takes a little while to understand--signal chains are built as "tracks" and then you recall their states with MIDI Note On messages--but it seems much more intuitive than the patchbay style interfaces for Bidule or Psycle. In fact, it's very reminiscent of Live.

Filed under: music»tools»digital»sampling

The Amen Break

A documentary by Nate Harrison explains the "Amen break," which is the most famous (and overused) sampled breakbeat in modern music. I've seen this wander around the music forums, but figured other people might be interested.

August 16, 2006

Filed under: music»tools»digital

My pedalboard is hyperthreaded

You start with the small stuff: the bundled plug-ins for Krystal, or maybe trying to find a better Audacity reverb. Then you get your hands on a real recording interface, which comes with a basic commercial DAW--Ableton Live Lite, or Cubase LE. The rush goes right to your head, especially when you stumble onto KVR Audio. Thousands of VST plugins, for free! Reverbs! Distortion! Even MIDI-controlled soft-synths. They make 12-step programs for those.

Still, you tell yourself you're just doing a little recording. You can stop any time you want--you just don't want to.

But one day you realize that these plugins are running in real-time--that they don't have to be applied to pre-recorded audio. They could even be used like guitar pedals--but cheaper, and all contained in a laptop. You find yourself researching virtual pedalboard technology until two in the morning.

Trust me, I know. And I'm your enabler, kids.

Is it actually a good idea to switch from standalone pedals to VST effects for an instrument signal chain? It depends on your circumstances. Virtual Studio Technology has many of the advantages of multieffects boxes--it's extremely tweakable, flexible, and cheap. It even has a step up in modularity--unlike a Digitech RP-100 or similar box, if you don't like a VST distortion, you can just replace it. There are plenty to choose from. And there are lots of specialized VST plugins that you can't really get in a stompbox form, like bitcrushers and vinyl simulators. Imagine it as a build-your-own multi-fx kit.

On the other hand, you might not actually be saving any space or money. You need a laptop, an interface, and some kind of controller (unless you plan on pecking away at the keyboard while you play, which I don't recommend). At least one of those probably has to be on wall power. Reliability could be a concern: the OS probably won't crash, but the plugins or their host might. The tools are complex--a lot of them are designed almost like an analog synth. And of course, you get to be the dork with the laptop onstage.

If you decide to give it a shot, remember that you don't need top-of-the-line hardware to run VST effects. It's always nice to have more power, but I've run multiple plugins without noticeable latency on a 366 Celeron laptop. Having a good sound driver is much more important than your processor speed if you're just going to use four or five effects.

Obviously, the question for me is not "can it be done?" but "can it be done cheap?" The answer is a reserved maybe. The effects may be free, but you need a VST host program to run them, and you need to be able to bypass them individually. That seems to be easier said than done. I've spent some time searching, and here's what I've found:

  • If you've bought a keyboard or a cheap recording interface in the last year, you probably own a copy of Ableton Live or Cubase LE. Cubase isn't worth it--my experience with the bundled version was that it introduced unnecessary latency into effected monitoring. Ableton Live Lite is a great program, but it only supports one VST on a track at a time, and it requires modern hardware to run properly. You can get around the former limitation by using either MultiFXVST (seems more flexible, free) or Chainer, (much more stable and supported, $60). But loading Ableton just to run a plugin or two is kind of overkill.
  • For the same performance-oriented philosophy, but geared less toward tracking and more toward hosting, I've heard really good things about Forte. It's not a recording/looping/whatever tool like Ableton, but was originally designed to be a complete multi-set software synth host. I know several people are using Guitar Rig or Amplitube with it. It's not free, but at $129 it's not outrageously expensive, either, and the MIDI control and routing options are formidable.
  • On the other hand, Niall's Pedal Board is completely free, and if you can get past its goofy-looking interface it allows for MIDI-controlled bypass of individual effects, even if they don't offer a bypass option themselves. This really is like using a pedalboard--there are no presets or fancy routing options--but the price is right and the functionality is hard to find anywhere else. It also includes joystick input, in case you're like me and you can't stop hooking dance pads up to your laptop.
  • Finally, a conditionally-free option is DSound's RT Player Express. I say conditionally, because it's a free download to anyone who owns an M-Audio sound card (including the low-end USB Jamlab and FastTrack units). A lot of people already own these, so it might be a good deal for them. It only supports 4 VST plugins at a time, but you can probably fool it the same way as Ableton, using another plugin to subhost multiple effects. If you have a qualifying sound card, this may be your best choice. Otherwise, I'd probably recommend Niall's Pedal Board, which should be able to handle most VSTs without crashes.
  • Once you've got a host, you need effects. A good place to start is the SimuAnalog Guitar Suite, which claims to actually model the physical signals for several vintage pedals and amps. It's free and includes a Tube Screamer, two Boss distortions (SD1 and DS1), a phaser, and a Marshall amp. Sadly, they don't have a bypass option for host control, so you'll need a host that can do this for them.
  • CamelCrusher does have its own MIDI learn function, so it can be used with hosts even if they don't handle automation. Comes with distortion, tone filter, and a compressor.
  • And of course, I have to include a looper. Loopy Llama has an awkward control setup, but the loop functionality is excellent, including undo, wet/dry output, Frippertronics-style fading, and tempo sync. It also has its own independent MIDI control capability, including the ability to trigger from note or CC messages. If you have enough switches on your controller to handle it, this is probably the best loop plugin available. Hard to believe it's free.
  • For more plugins in all kinds of flavors, be sure to search KVR Audio, a huge database of software. There are lots of options I found there that I didn't include here because I thought they were too awkward or they're inappropriate for this specific task, but you might find a use for them.

August 9, 2006

Filed under: music»tools»digital

Control Issues

Problem: I'm using several virtual instruments, both at home and at work, but I don't have a MIDI keyboard, and even if I did, I'm not a very good piano player. It would also be nice to have transport control inside the booth at work, and that means wireless.

Solution #1: Use a normal USB keypad (or mini keyboard), and route it to MIDI using the command line utilities. There are four rows on a keypad. There are four strings on a bass. Some muscle memory still applies.

Solution #2: Use a gamepad, remapping joystick inputs with Rejoice. Wireless gamepads are relatively cheap, especially if I hack another xBox pad. But how to lay out the notes? Perhaps something along the lines of Band Brothers for DS.

July 2, 2006

Filed under: music»tools»bass

A Circle of Learning

Every few months, I listen to a bunch of other people's music while twiddling knobs on the Bass Effects Rig of Doom, and I think to myself, "you know, I probably don't need to have the distortion pedal set at 'Chernobyl' all the time. In fact, I really like that smoother, bluesier sound I could get with a lower gain."

Then I listen to my recordings the next day (because of course I'm the kind of person who obsessively listens to their own recordings) and I wonder why the treble sounds so muted. And I realize that the reason I was originally using the Face-Melting Fuzz setting in the first place was because otherwise it requires me to dig in way too hard to get a good sharp attack out of the notes. So I crank the gain back up, and enjoy the softer touch.

Then a few months later, I'm listening to some other music...

June 28, 2006

Filed under: music»tools»bass

Don't Blame Me

An old thread on the Lowdown surfaced today via e-mail: we had been discussing the problems of usability for effects units, and it got me thinking about how we use technology. Perhaps an overview for non-musicians might help first.

You can split bass and guitar effects a lot of ways--by function, or by architecture, or by quality--but a big philosophical issue for musicians is single- versus multi-function boxes. The traditional FX pedal is a single function. It does one thing only, and hopefully does it well. Individual effects boxes can be digital or analog inside, but they tend to have manually-set knobs--they're not programmable, but they are very tweakable onstage. Musicians, being tactile (and sometimes not very bright) people, often like these standalone effects because they're simple and direct.

The other side is multi-fx pedals, which are capable of doing many things, often at once. For example, I have a Digitech RP-50 guitar pedal for experiments, which can add EQ, distortion, delay, and one other kind of effect (chorus, envelope, pitch-shift, etc.) all at once. Multi-function effects are almost always digital, because putting so many kinds of analog circuitry in one space would take up too much room and be incredibly expensive. Cheap sound processing chips can do credible imitations of the analog original, and it can be nice to have so many options available. So a lot of people start out with a multi-function pedal until they figure out where they want to spend a lot of money.

But if you're going to build what amounts to a sound-pedal-emulator, it's nice to have a save/recall function for settings, because now you're not just tweaking one pedal at a time, you're modeling several. Lo and behold, that's exactly the way it works: you set up everything the way you want it, and then the pedal typically saves all those settings into a "patch." This is cool, because it allows musicians to jump from one sound to another very quicly. You might have one patch that's lots of stereo delays and light distortion to sound like U2, and then another with very heavy distortion and a sub-octave to sound like Cheap Trick.

Naturally, there's a tradeoff for that much power and flexibility. Switching between patches tends to take time as the pedals--which are running on as little power and memory as they can possible get away with--unload one chunk of code from RAM and load the new patch to start processing. It's not a lot of time, maybe 20 milliseconds or so, but for that 1/5 of a second you are muted--effectively dropped out of the song. That's one reason why I quit using multi-FX and went back to individual analog pedals. I just couldn't stand having to pre-trigger my sounds to keep from missing a beat.

Zoom recently released the B2.1U, a multiFX pedal with only 8ms of switching time. Obviously, a few of us were interested in anything that would give us more power without the delay. However, one board member, well known for his vocal advocacy of a particularly flexible (but equally expensive and elaborate) effects unit, went the other direction. It doesn't matter how fast the switching time is, he said, because you shouldn't be using more than one patch anyway. Just set that up so you can change it a little, and you should be good to go. And he dropped this flame into the thread:

If one needs more than a single preset during a song performance there is something wronmg with either the device design, the user, or their conception of the song materials.

In other words: if you can't make it work with one patch, it's your fault. Even though the pedal organizes itself in presets and that is the primary mode of user interaction, it is not a design flaw that it can't do so quickly. It is your fault that you didn't plan around the huge gaping silence when you wanted to make a different noise.

I hate it when people make that kind of argument. It is a huge problem with technology. Blame the victim, not the crappy machinery. Sound familiar?

I am so tired of being told that it's my fault when things are hard to use. You used to hear this a lot from the Slashdot variety of Linux afficianado: the computer isn't hard, you just have to learn how to use it! Well, yeah. Maybe if it's that hard to learn, there's a flaw somewhere in the design, not in the people. If a game controller is intimidating to people, it's not that they're unintelligent or uninterested, it's that they're not willing to geek out for years in order to learn the muscle memory required. There's the infamous "blinking twelves" on an average VCR. All kinds of machines have this problem.

It's not that technology should all be simpler, because there is a time and a place for power and depth. But the tech community has been really good at shifting the blame from the designers, where it belongs, to the users, where it probably doesn't. What I've noticed is an inferiority complex by anyone who's not a self-defined geek. People are afraid of technology. They're convinced that they're not smart enough to use it. So they won't. And if the communications revolution of networking has shown us anything, it's that we don't benefit from people who just opt out of the web.

If you're reading this, you're likely to be a go-to tech person in your office or family. You can do something about the blame. Sure, you can't redesign someone's effects, or re-build their operating system. But (and I'm trying this myself, it's not easy) you can be more patient and friendly when explaining technology to someone. Make it clear that it's not their fault--and they're more likely to learn more, until they won't have to ask for help so much.

See? From effects pedal to broad societal doom-and-gloom in one post! Now that's value-added.

Future - Present - Past