[gmpi] Re: Decision Time: 7.1.2
- From: Tim Hockin <thockin@xxxxxxxxxx>
- To: gmpi@xxxxxxxxxxxxx
- Date: Thu, 12 Jun 2003 23:50:10 -0700 (PDT)
> > > . Sony's "Super Bit Mapping", or Apogee's UV22
>
> > What the heck are those?
>
> When you convert floats to 16 bit int, to send to your Soundblaster Live, or
> burn to CD, you often 'dither' to preserve as much dynamic range as posible.
> Helps mask the 'grainy, digital' sound.
Thanks for that. I knew what dithering is, just not those two. Tehre is
also the patent-pending POW-R dither algorithm, which has won much acclaim.
YOu don't need to dither when converting between Float32 and Int24. You
might want to dither when converting to Int16. If you need a better dither
than the host provides in it's automatic conversion, You can buy a POW-R
plugin.
It takes Float32 input, converts and dithers to Int16, then converts back to
Float32. You won't lose any precision once it is converted and dithered.
You can then convert it to real Int16 by the most simple form, and get back
your dithered form.
Yeah, there is an extra conversion. Unless we allow in_type != out_type,
that is the sanest way to get binary dither algorithms into your graph,
unless your host does something more special.
As a GMPI comrade we could do a small API for hosts to load converter
plugins/libs. It would be a simple and useful API for things EXACTLY like
this. POW-R_convert.dll gets loaded and is used to convert Float32 to Int16.
If that API doesn't need all the flexibility of a plugin, why force it?
In fact, I think it is a wonderful idea for type conversion/dithering. The
GMPI_convert SDK would provide basic transparent, dithered, and noise shaped
algorithms.
So I now assert that (intype != outype) is a non-issue. The type-converter
API can handle loadable converters, and is outside the scope of GMPI.
----------------------------------------------------------------------
Generalized Music Plugin Interface (GMPI) public discussion list
Participation in this list is contingent upon your abiding by the
following rules: Please stay on topic. You are responsible for your own
words. Please respect your fellow subscribers. Please do not
redistribute anyone else's words without their permission.
Archive: //www.freelists.org/archives/gmpi
Email gmpi-request@xxxxxxxxxxxxx w/ subject "unsubscribe" to unsubscribe
Other related posts:
