Autodafe Module Pack for VCV Rack

[StudioDave]

Is the source code available for these modules ?

not at the moment, I didn't have time to check the code, clean it up and pubich them on a repository on github, but I plan to do it ASAP...

Thanks for the note, I'll watch for an announcement.

Best,

dp

[autodafe]

I have released a bunch of new modules, my Drum Kit modules and my first "commercial" module, the "Blank Panel". All modules are FREE, only the Blank Panel costs 5 euros, you'll undertsand why

https://www.autodafe.net/virtual-instru ... dules.html

[RPH]

FYI, I registered on your site and got an activation e-mail which contained my account details including my generated password visible in the mail ( security wise not very handy: it's plain text ).
Normally I would expect just a link to activate the account, not showing any details or maybe just the username.
If you have access to the sent e-mails on the mailserver, you can see my password.

[autodafe]

I don't have access to the mailserver. I see what I can do, it's a standard Joomla (website CMS) feature...

[autodafe]

FYI, I registered on your site and got an activation e-mail which contained my account details including my generated password visible in the mail ( security wise not very handy: it's plain text ).

thanks for pointing that out. I updated the setting in Joomla so that password are not sent to users.

[RPH]

No problem, thanks for updating so quickly.

Am trying your foldback module, it's a shame the volume gets really low while lowering the threshold point.
For Synthedit I once built a foldback module which has a gain knob to increase the fold and a switch for the dB response point.
That way the volume doesn't go so low when folding. Maybe an idea for your module or a second version?

Screenshot: http://www.rhmodules.nl/Graphics/screen ... dback2.gif
C++ Source: http://www.rhmodules.nl/Source/source-RH-Foldback2.rar

[autodafe]

Maybe an idea for your module or a second version?

yeah, it could be, thanks for the links. I now need to slow down a bit because I have been working almost day and night on them in the past two weeks or so...

[shroom81]

Great work Autodafe, thanks so much

Bought the blank panel cause I found the description hilarious
and added 5 euros extra just because

[autodafe]

Great work Autodafe, thanks so much

Bought the blank panel cause I found the description hilarious
and added 5 euros extra just because

ahaha glad to see someone understood my kind of humour

[GrokSynth]

This looks great! I also would like to compile the source files myself. Any idea when that will be available?

[autodafe]

I need to calm down a bit, then clean the code and make some teeaks and optimization, then I think I will publish the code on github

[autodafe]

OK, I think I managed to get my code on gitHub
so far My Module pack only

https://github.com/antoniograzioli/Autodafe

this version works for Rack 0.3.2 I am currently upgrading the code to work in 0.40.

[RPH]

clean the code and make some teeaks and optimization

FYI, you seem to oversample the LFO 16 times when I check your source.
The Biquad source could use a more efficient tan approximation.

[autodafe]

FYI, you seem to oversample the LFO 16 times when I check your source.
The Biquad source could use a more efficient tan approximation.

for the LFO i basically "cloned" the VCO in fundamentals, reduced the speed and adde a switch.
WIll have a look at what you suggest, thanks!

[RPH]

for the LFO i basically "cloned" the VCO in fundamentals, reduced the speed and adde a switch.
WIll have a look at what you suggest, thanks!

I see. The VCO needs the oversampling for FM and other audiorate modulation to keep it sounding good ( no aliasing ).
LFO doesn't need oversampling, so my suggestion would be to remove it completely ( or provide a switch on your panel to be able to lower it ).

Here is the fast tan function I got from a fellow Synthedit dev:

Code: Select all

double Pade_Tan(double x)   // pade-approximation, credits: Andrew @ SE SDK usersgroup
{
    double x2,x3,x4,a,b;
    x2 = x*x;
    x3 = x2*x;
    x4 = x2*x2;
    a = 5.0*((-21.0 * x) + (2.0 * x3));
    b = (105.0  - (45.0 * x2)) + x4;
    return -(a/b);
}

And a fast Pow function for usage in the peak type you use:

Code: Select all

double fastPow(double a, double b) // martin.ankerl.com
{
  union {
    double d;
    int x[2];
  } u = { a };
  u.x[1] = (int)(b * (u.x[1] - 1072632447) + 1072632447);
  u.x[0] = 0;
  return u.d;
}

Furthermore you can improve the biquad calcs by reducing the amount of multiplies:

Code: Select all

double norm, V;
double omega   = M_PI * Fc;
double K       = Pade_Tan(omega);
double K2      = K * K;


case bq_type_peak: //Peak
            V       = fastPow(10.0, fabs(peakGain) / 20.0);
            if (peakGain >= 0.0) {    // boost
                norm = 1.0 / (1.0 + 1.0/Q * K + K2);
                a0 = (1.0 + V/Q * K + K2) * norm;
                a1 = 2.0 * (K * K - 1.0) * norm;
                a2 = (1.0 - V/Q * K + K2) * norm;
                b1 = a1;
                b2 = (1.0 - 1.0/Q * K + K2) * norm;
            }
            else {    // cut
                norm = 1.0 / (1.0 + V/Q * K + K2);
                a0 = (1.0 + 1.0/Q * K + K2) * norm;
                a1 = 2.0 * (K2 - 1.0) * norm;
                a2 = (1.0 - 1.0/Q * K + K2) * norm;
                b1 = a1;
                b2 = (1.0 - V/Q * K + K2) * norm;
            }
            break;