[LAD] More about de-normals - some test results.

[Tim E. Real]

On May 4, 2010 10:10:13 am you wrote:
> On Mon, May 3, 2010 at 8:27 PM, Tim E. Real <termtech at rogers.com> wrote:
> > Haven't tested Ardour yet, but what do you think about this, Paul?
> > I suppose Ardour's processor flags option *must* cure it?
>
> should fix the problem if it was denormals, but i don't think that the
> values you are seeing there are denormals (too big). i'm not sure about
> that though.

Good point. The values may be a little bit too high for that. 
The system starts chugging right around when those values I showed 
 start to appear (around +/-1e-18 +/-1e-19 or so).

The problem may really still have been inside the plugin, 
 and not the fault of passing these values to Jack later.
Inside the plugin maybe much lower values are tossed around,
 before the higher values appear at the output.

It never occurred to me, but the plugin's AC coupling makes 
 perfect sense now, being models of real electronic devices, 
 complete with input coupling capacitors.
Some plugins would have it, others wouldn't.

Thanks. Tim.