Tuesday, February 15, 2011

New Song: Kraku - Krapularipulihumppa (Hangover Diarrhea Polka)

Alternative content

That's the cheeky song I made this month. It was composed while I had a hangover and I finished it while I was drunk...

Here's the download link to it:
Kraku - Krapularipulihumppa

The name is in english "Hangover Diarrhea Polka" or something close to that...


Sunday, February 6, 2011

Technological Idea: DCO that emulates a VCO

In my previous post you can hear what the digital oscillators sound like when ran through analogue filters. There's still surprising amount of digital preciseness to the sounds you get. If you somehow get rid of that preciseness, you should probably end up with much more analogue sound.

This preciseness doesn't necessarily come with digital oscillators alone. Some DCOs are very capable of it too. One great example of this is Dave Smith Instrument's Prophet '08 analogue synth. Its DCOs are super precise and cutting. They're probably even more precise sounding than most VAs, which might be the results of really high speed control over the DCO cycle times (much higher than 44.1 kHz). To be honest, I've no idea in what frequency the Prophet '08's DCOs are controlled. All I know is that they sound super precise and they're a perfect example of a precise DCO sound. This got me thinking one technological idea I had a while ago.

What if DCO could emulate very convincingly the unstable VCO sound?

The basic concept is fairly simple:
Trigger the DCO using an analogue modeled simulation of VCO's cycle triggering circuitry. For example take Korg Legacy Collection MS-20 software synthesizer and throw away everything else except the algorithm which triggers the oscillator cycles. Now put this very same algorithm inside the chip/cpu which triggers the DCO cycles. What you should end up with is a stable DCO which still sounds like your unstable vintage VCO.

Here's hoping for that one of the big three synth manufacturers would release a new analogue synth with such oscillators... Not gonna happen :(

Technique test: Sylenth1 used as oscillators for a modular synth

A short while ago I tried out what it would sound like if I used LennarDigital Sylenth1 synth plugin as the oscillators for my Eurorack modular synthesizer. I've had this idea for quite some time already, but it was only recently when I actually tried it out. I send the unfiltered oscillator sound to various Eurorack format filters and triggered the envelopes using MIDI and then recorded the results back into my DAW. Here are the results:

Alternative content

The filters can be heard in the following order:
0:00 - 0:21 = A-124 (Wasp filter)
0:21 - 0:45 = SY02 (MS-20 filter)
0:45 - 1:12 = Boogie Filter in 24dB mode
1:12 - 1:42 = Boogie FIlter in 6dB mode
1:42 - 2:15 = A-106-6 in 4L mode (XPander Filter)
2:15 - 2:44 = A-103 (303 style filter)

Note that I didn't try to aim for any cool sounds with this experiment, but tried to draw out the different characters of the filters. I also mixed the bass very loud so you'd hear their character and still hear the timing of the sounds (hence the drums in the background).

What's surprising is that there's still quite a lot of that digital preciseness in these sounds. This was mentioned by several people who heard the results of this test. Yes, the analogue filter gives a very different flavor to the sound, but there's still some of that liveliness missing from these examples. This probably means that a lot of the analogue character many people love comes from the unstable oscillators. This is actually quite logical when you think of it. The filter does just what it says on the lid: filters out different frequencies. Sure it also adds distortion etc. but it doesn't modify the harmonic content the same way as for example pulse width modulation or unstable oscillator cycle time does, which is the oscillators territory.

It would be interesting to try this experiment the other way around; analogue oscillators filtered using digital filters. How analogue/lively would the sound be then?

But anyway, it's an interesting technique and I might start using it in some of my tracks in the future.

One final thing:
Naturally this technique works with any software and hardware VA-synth/sampler/ROMpler/FM-synth/etc. you can lay your hands on. Just take your favorite sampler/FM-synth and use it as the "oscillator" for your modular synth.

The Quest for Dynamics: part 2

Let's think about the audible frequency range as two separate (and partly overlapping) ranges: high and low frequencies. Everything below 250Hz or so could be thrown into the "lower frequencies" category (no hard and fast rules here). Rest of them are higher frequencies for the purpose of this example. No "middle frequencies category" on my list here today.

Here's how the listener perceives these frequencies:
  1. Higher frequencies = loudness & presence
  2. Lower frequencies = power
If you feel that your track isn't kicking your chest, you need to concentrate on the dynamics of the lower frequencies. If you feel that your mix sounds stuffed with sounds and/or doesn't have depth to it, you need to concentrate on the higher frequencies. In this post I'm going to concentrate on the higher ones.

There is no substitute to keeping the frequency spectrum as clear as possible by designing the sounds in your song so that they don't overlap much in the first place. However, the mixes still tend to swallow certain instruments, like drums in this example. Let's hear the two example audio clips.

A cheeky tune called March of the Underpant Gnomes, with a proper drum mix:
Alternative content

...and here are the same gnomes marching without extra dynamics on their drum bus:
Alternative content

The only difference between the two audio clips are the extra dynamic processing on the drums. The drums in the second clip sound like they're further away in the mix due to the reverb on them. Your first instinct might be to turn up the drums or take away most of the reverb AND then again turning up the drums a little bit. Then you'd be left with a track that's surely "in your face" sounding but would sound much more dull due to lack of dynamics.

I achieved the cutting and present sound of those drums by running the whole drum bus through Logic's Enveloper plugin, which is a dynamic enhancer plugin, kinda like Transient Designer I guess (haven't used it so I can't be 100% sure). See the attached image for reference on the parameters I used. All I did was set the initial transient boost to 64% and left rest of the parameters alone.

Ultimately what I'm doing here is making the transient relatively bigger to the rest of the drum sound. You can achieve this same effect by different methods:
  1. Enveloper / Transient Designer type of processors.
  2. The standard compressor trick on the drums:
    • used either separately on each drum or on the drum bus, which ever works for the track
    • you can use side chaining here to achieve different effects, for example compress reverberated drums with dry drums on the side chain bus
  3. Tweak the drum synth's/sampler's AMP envelope:
    • attack = 0
    • decay = 10-30ms
    • sustain = somewhere around minus 4-8dB. The lower you go, the bigger the transient is, of course
All these methods give you different sound flavors. Also different compressors can sound very different indeed so it's worth trying all the gear you can get your hands on. It would be interesting to try slamming my drums on some high end compressor like a pair of Distressors or API 2500. Maybe some day...

The final thing I want to say is that the more you have dynamics in your song, the more you have choices where to put your sounds in the mix. If for example the drums have a huge dynamic range, you can choose if you want to mix the bass or guitar up loud or put them more in the background and still sound good. Just some food for thought.

Saturday, February 5, 2011

The Quest for Dynamics: part 1

Lately as I've been trying to add more and more dynamics into my mixes, I've ran into interesting problem with how different musical genres "should sound". When you tune in to your favorite radio/TV channel, most of the time you hear this flat and overly compressed sound wall coming your way. No dynamics to speak of; just this huge "block of sound" making its presence known through your speakers.

Many dance floor productions don't achieve this sound at the mastering stage by over compressing / limiting the final mix. Usually the sound comes from using dynamically "lacking" sounds in the first place, which is fine and many dance floor stomping teens don't seem to mind at all. But I do. The hard part would be getting those dynamics into the mix and still sounding good and familiar to the regular listener. It may sound simple enough, but the whole feel of the dance tracks changes dramatically when you create more dynamically/transient rich sounds and mix them well together. The mix starts sounding musically more interesting, but does it appeal to the club frequenting audience? Probably not. The mix just doesn't have that hugeness to it anymore which is expected of every dance track out there. That's the problem.

Most of the time I can't be arsed about this stuff since I usually won't spend that much time working on my tracks anyway. But from time to time I feel like paying closer attention to the details of production and on getting better results. When I'm doing this, I usually learn something new. My intention is to blog these new things I've learned, so expect a couple of blog entries on the dynamic range tricks with some audio examples included.