Some time ago I fell into an interesting topic on the Ableton forum; someone was wondering how to make a Siren in Ableton Live. Unfortunately not too many people picked up on the subject, but I have a very weak spot for stuff like that. Especially because you may actually learn something from trying to solve such an “easy” problem; Making a siren, how hard can it be?
Well, as always, that which may seem easy enough can very well pack quite a punch.
First I started with Live itself. Because I have the Suite version it basically allows me to do whatever I want. I decided to put the setup into an Instrument rack so that regular Live users can enjoy it too.
I decided to start off with the Analog instrument because in my opinion it would be much better suited for a siren. I was looking for a full “phat” sound, fully focusing on a sawtooth.
Yet the first problem I faced: a siren, especially air raid sirens, tend to build up to full sound and when you want to stop they slowly decay. But the pitch build and release is something very specific for that sound.
This was yet another reason to resort to Analog because it allows you to setup a pitch build up on the fly, as you can see in this screenshot. All I did was setup the first oscillator and then defined the ramp it should use. Ignore the green dots for now; those only indicate that some options are being controlled by one of the rack macro dials.
Next I also wanted to try and give a little more body to the sound. I managed to achieve that by increasing the number of voices and also setting up the detune so that one oscillator can produce a slightly “out of sync” signal, which is also very specific for an air raid siren.
While the result isn’t too bad at all, in my opinion anyway, it does have one limitation… The sound is only building up, but when you release your key then the sound will stop instantaneously. A real siren however should also be able to more slowly “lower” the sound again.
Now, the real fun with challenges like these lies in trying to get the same (kind of) result with several environments. Several DAWs have several workflows, which can make it pretty interesting to compare. However, it turned out that Reason wasn’t really up to the challenge. That is; it is perfectly possible to create some kind of siren sound with Reason, that’s not the problem. But my aim was to create one specific device which you could then implement in whatever project you were working on…
Knowing where to start was easy; the Thor polyphonic synthesizer. It has many different oscillators and an extensive filter section which makes it the ideal basis for our siren sound. The only obstacle here is that we want the pitch of the sound to build up. The first attempt to solve this can be seen here; Thor’s own step sequencer. I think the problem should be obvious at once: with only 16 steps and no real way to define a range of some sort it won’t be enough for what we want.
Luckily for us Reason has many more tricks up its sleeve.. As such I ended up trying to hookup the Matrix device to Thor:
As you can see here a Matrix is a sort of step sequencer, but one which provides a little more specific control than a regular one. I’ve setup a ramp going from low to high and created the necessary connections on the back to make the Matrix device control Thor’s pitch wheel. Now while this works decently well there is one more problem here… As soon as the Matrix device reaches the last step (number 32) it will jump back to 1 and start all over again. That is not what we want since that would turn our siren into a looping sound; one which constantly builds up from 0.
I ended up concluding that there is no good way in Reason to create a siren like effect while using a single device. The main problem I came across was timing. While Reason has plenty of devices which provide “ramps” (like LFO’s, or the Matrix I mentioned earlier) they cannot be easily controlled themselves.
For example; I can’t control the Matrix using another device or tell it to only do a single run. As such my solution in Reason ended up with manual automation, as can be seen above.
I basically pulled up the sequencer and simply started setting up a pitch ramp myself, as you can see in the screenshot. This makes it perfectly doable to generate a siren sound in Reason, but it just doesn’t give you the same kind of control I have in Ableton Live. However this does solve the problem I have with Live’s Analog; now I can make sure that my siren will build up and more slowly fade off as well.
Absynth is one of the big favorites in my synth collection. While it comes with a rather heavy pricetag (even when I got it during N.I.’s last discount period) every time I work with it you immediately notice the high quality you’re working with.
It shouldn’t come as a surprise that I used a single sawtooth oscillator which I then tweaked a little by setting up a unison sound. 3 voices, a little differentiating and… Here is one of those many “high quality” examples: A random factor.
In other words; the amount in which the three voices are detuned isn’t a static setting in my patch, instead it slightly varies every time you play it.
As you can see I also setup a cloud filter to give my siren a little more body; making it fuller than the default sound generated by the oscillator.
And now to tackle the real problem: how to make the sound slowly build up ? Better yet, if possible how to make sure that it also will slowly decay again ?
One of Absynth 5’s key features is its envelope section, which can be seen above. By default you have 3 envelopes; one for each of the oscillator’s amplifiers. And with using the regular “ADSR structure” you can basically control whatever you want by merely adding extra envelopes as required, resulting in sound control by using your keyboard. In other words; when I push a key Absynth will start with the Attack section (the first round up curve which ends with the letter ‘A’). Then it follows the ‘Decay’ section (the straight line in the middle) and then remain there for as long as you keep the key pressed, obviously, considering that I’ve also not done anything special with the Sustain section.
Finally, when I release the key again it will then start following the Release section. Which is the last curve downwards. So in other words; it will then lower the pitch of the siren again until we’re back to 0.
Of course I had to alter the default “Oscil A Amp” envelope a bit to make its release cycle a little less steeper so that you could clearly hear the slowly lowering pitch.
Max for Live
Last but not least; what if we were to simply program a siren ourselves? M4L excels at that task obviously and although it will take you a little more effort to build such a device, the obvious advantage here is that it will also give you a lot more control than any of the above synths can give you.
Now, I’m still learning to find my way around M4L’s “MSP section” (this is the section which can be used to generate and process sound signals) so I kept the patch very simple. A mere 2 sawtooth oscillator without trying to apply detuning or filtering the end signal in any way.
I parse the (raw) incoming midi data using “midiparse” and end up grabbing the pitch and velocity values. I then separate the list and concentrate on the velocity value, which tells me if there’s a key pressed or not.
As soon as a key is pressed I tell the line~ object to generate a ‘ramp’ from 0 to 184 within 10 seconds. This signal is then send to both sawtooth oscillators (the saw~ objects), resulting in a slow pitch build up. After 10 seconds the full pitch is reached and the siren sounds as intended. When a key is released again a second ramp is build; this time its from the current playing value back to 0 in 7.5 seconds.
As you can see for yourself it’s a very simple and sub-standard patch, but it should give you a rough idea as to how you could build something like a Siren using M4L.
SO, does this mean that one product is better than the other? Of course not; it simply goes to show you that separate products have their own advantages and disadvantages. This is also the main reason why I started to collect a few more synths besides Ableton Live, even though I have quite a few instruments available with Suite and am pretty happy with Live the way it is now.
But that doesn’t mean it’s the ultimate solution for every sound-related problem out there. A little diversity can go a long way.
If you want to download the presets and play with ‘m yourself then you can grab an archive with all presets here.
(thanks to Sondance for warning me about the broken link!).