Re: CVS commit - no version change

On 8 May 2007, at 03:46, Jan Depner wrote:

> On Mon, 2007-05-07 at 19:58 -0500, Jack O'Quin wrote:
>> On 5/7/07, Jan Depner <[email protected]> wrote:
>>
>>>     Yeah, I was just bored ;-)  So, when is 1.0?
>>
>> Any time you like.  You seem to be on a roll right now.
>>
>> If we think we're functionally complete at this point,
>> I'd recommend building a release candidate and putting
>> out a message on LAU requesting that people test with it.
>>
>> Thanks for doing the RMS feature, Jan.  I felt bad about
>> promising to work on that almost six months ago, then
>> not doing a single thing in all that time.
>
>     That's OK.  I hadn't done anything either and you were busy with
> JACK.  I hurt my back last week and I was home for a few days and got
> bored.  I was actually surprised at how easy it turned out to be.   
> I was
> expecting much more difficulty with it.
>
>     I'd still like to be able to have up to 2ms delay on the low band
> and .5ms on the mid like you can do with a Rane crossover or BBE.   
> It's
> amazing what a difference that makes.  I have no idea how to implement
> that though (other than the interface of course).

That should be easy enough in theory, but there are some subtleties  
around bypassing.

The code would need to be instead of

             out[port][pos] =
                 out_tmp[port][XO_LOW][pos] + out_tmp[port][XO_MID] 
[pos] +
                 out_tmp[port][XO_HIGH][pos];

There would just need to be a set of 6 ringbuffers that had out_tmp[] 
[] written into them, and read out with an appropriate delay into out 
[][].

So, something like:

	dpos++;
	delay_buf[port][XO_LOW][dpos & DELAY_MASK] = out_tmp[port][XO_LOW] 
[pos];
         // rept for MO_MID and XO_HIGH

         out[pot][pos] = delay_buf[port][XO_LOW][(dpos - delay 
[XO_LOW]) & DELAY_MASK] +
                   // add other XOs


DELAY_MASK is one less than the length of the delay_buf, which must  
be a power of two. delay[] is the delays of the 3 channels in  
samples, and dpos is just a static unsigned int or similar.

I'm not sure what you do about bypass yet. It might just be that if  
you use channel delays then you get clicks when you switch bypass in/ 
out.

- Steve

-------------------------------------------------------------------------
This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE version of DB2 express and take
control of your XML. No limits. Just data. Click to get it now.
http://sourceforge.net/powerbar/db2/