[vorbis-dev] mdct.c

[Dave Hooper]

[This doesn't answer all of the questions, but only because I'm too stupid
to figure out all the answers]

> Hmm, you are right... But is the smallest element of data read from this
> stream a kind of integer in a fixed-length amount (let's say 8 or 16
> bits)? Or it is not at byte boundary at all...

This kindof depends on your definition of 'smallest element'... given that
the stream is compressed bitwise (i.e. a bitstream) the smallest element is
really a bit - the individual pieces of data making up an encoded block are
not on a byte boundary. However, given that you pass whole blocks to a
decoder, the smallest element is more usefully the smallest valid blocksize
in a Vorbis file.  I can't remember if a blocksize needs to be an integral
number of bytes.

> Then it is probably faster to run the decoder with floating point and FPU
> than the integer version without FPU given that both circuits work at the
> same frequency?

In general, on modern computer hardware, then probably, yes :-)
For embedded applications, the specifics of the target device will
implicitly govern this sort of decision.  Without access to an FPU, then
obviously the integer version is the only choice.  If you have full control
of a FPU, then
the floating-point algorithm would usually be a more sensible option.

Dave

<p>--- >8 ----
List archives:  http://www.xiph.org/archives/
Ogg project homepage: http://www.xiph.org/ogg/
To unsubscribe from this list, send a message to 'vorbis-dev-request at xiph.org'
containing only the word 'unsubscribe' in the body.  No subject is needed.
Unsubscribe messages sent to the list will be ignored/filtered.