Hello,
I'm trying to make an envelope controlled filter (centre frequency set by the loudness of my guitar, like a Q-Tron). I can't for the life of me get it to work. I've used the env~ object and got a value between 0 and 100 when playing. I tried multplying that by 1000 and running it into the middle inlet of both bp and vcf and nothing happens.

Not worth posting what I've made so far, but am I missing some basic principle here? All the online help stuff focuses on oscillators and ADSR's and not real signal envelopes.

Also what does the vcf mean by frequency controlled? Is it smart enough to detect a complex sound? Does it only take sine waves or something? It's not very clear.

Thanks.

[vcf~] allows you to use a signal to control the frequency instead of a message. This means it can be changed smoothly without clicks. If you do something like this, it should keep it smooth:

[env~]
|

• <-- set your frequency range
  |
  [+ 100] <-- set the minimum of the range
  |
  [$1 20(
  |
  [line~]
  \
  \
  [vcf~]

There are other ways to do it, such as using [lop~] instead of [line~], but this should give results. Just make sure you also set the Q value in [vcf~].

I didn't knew the q-tron but after listening some demos here's how I would emulate it in pd.
I always use the second inlet of fiddle~ for attack detection that gives better result than env~ with a guitar, then use line~ etc as Maelstorm has explained.

http://www.pdpatchrepo.info/hurleur/q-tron.pd

Thanks guys. So what does the second inlet for bp expect thats different to the inlet for vcf? They're both frequency in Hz aren't they? Also if you set the minimum freq outside of vcf do you still put q in there? Or do you set q in a separate object ?

Yes they're both in Hz but the second inlet of bp~ is expecting a control message and the second one of vcf~ is expecting a signal. That's why vcf~ is more adapted if you want to change the center frequency in real-time without getting any click. To respond to your first post, you'll never put a complex signal in it, rather a sinewave or line~ to control a simple signal (as in the example I gave you).

Q (or resonance) is set with the third of vcf~ inlet and you have to set it at least one time otherwise the filter won't do anything. It is useful to emphasize the filter sweeping so it is useful for this kind of effect.

Not sure if this is clear don't hesitate to ask again !

@coloscope said:

I didn't knew the q-tron but after listening some demos here's how I would emulate it in pd.
I always use the second inlet of fiddle~ for attack detection that gives better result than env~ with a guitar, then use line~ etc as Maelstorm has explained.

Great patch thanks!. I couldn't get it to work at first because things were connected to the right inlet (outlet?) on adc, so I moved them over to the left (Where my guitar is plugged in). It sounded great, but kept re-triggering if I held a note long enough. Am I doing something wrong?

You should set the "attack detection threshold" slider accordingly to prevent fiddle to output attacks too often. It depends on your setup and the input level of your instrument, usually I set it around 50.
You can also change the values of the [reattack( message sent to [fiddle~] to modify the behaviour of the attack detection, just check the help file of fiddle~ for details on this.
Glad you like it, and thanks to you for pointing to this effect, I'm now using it in my multi-effect to get some sitar-like sound, really fun to play with.

I actually own a Q-Tron plus. It has an fx loop so you can detect guitar signal envelope before an effect, and it applies the filter after the effect. It's great for filtering gain based pedals. Also have a look at the Bassballs, two filters in parallel set to different frequencies.