I have some further plans for additional units. One I'd like to test is a biquad filter that does limiting or saturation on the resulting values before copying them to become y[n-1]?
2:00 PM
 
would be interesting to hear what kind of distorted filter effects it migh produce when driven hard.
Tim Place
2:00 PM
yes!!!
2:01 PM
 
I'm interested to hear the results of that
Trond Lossius
2:01 PM
I'll test it when I get the time. If I find it to produce something interesting, I'll let you know.
2:02 PM
 
The nice thing is that more or less all of the code required is present already in existing units.
Tim Place
2:02 PM
right
2:02 PM
 
you might even be able to instantiate the existing units in the new unit, rather than copying the code
2:02 PM
 
for example,
2:02 PM
 
the limiter (I think) has a dc blocker built into it
2:03 PM
 
by just instantiating the TTDCBlock? unit
2:03 PM
 
(If I remember correctly)
Trond Lossius
2:05 PM
I was thinking of it. The only thing is that all units do their processing on all of the vector. I'll need to do the saturation inside of the filter calculations:
y = a1*x0 + a1*x1 + a2*x2 + b1*y1 + b2*y2
y = saturation(y)
y2 = y1
y1 = y
Tim Place
2:05 PM
sure
2:05 PM
 
There are two things here
2:05 PM
 
one is that you could pass a vector with 1 sample in it
2:06 PM
 
the other is that I've been wondering if units should have an additional method which processes 1 sample
2:06 PM
 
y = f(x)
Trond Lossius
2:06 PM
Could be interesting.
Tim Place
2:06 PM
it could be that the base class would provide this automatically if the class didn't provide the method directly
Trond Lossius
2:06 PM
Then TTBlue? could be used for the FunctionLib as well, in fact!
Tim Place
2:06 PM
ah yes!
2:06 PM
 
I hadn't thought of that
2:06 PM
 
but it's true
2:07 PM
 
In fact, the FunctionLib in the experimental branch does use TTBlue? -- it inherits from TTDataObject? instead of TTAudioObject? though
Trond Lossius
2:07 PM
So you could have control rate filters, even if one would not necsessarily have obvious access to how to translate parameters such as freq in Hz into meaningful range descriptions within the control rate context.
Tim Place
2:10 PM
This also suggests that any TTBlue? unit could be used as a function in the FuncitonLib?. The overdrive or limiter for example.
2:11 PM
 
That could be strange, but it could be interesting to experiment with
Trond Lossius
2:12 PM
Would be a challenge to describe sampling freq, but I could imagine that it could be quite useful in fact.
2:13 PM
 
Imagine e.g. continuous input from a sensor. A limiter could keep the signal "hotter" while remaining responsive to short term changes.
Tim Place
2:14 PM
good example