In looking at the recipe for libspeex, I see that the
--disable-float-api is set.
Is there a reason for this? Is there a better way to do it? (I'm new to speex
and audio encoding/decoding).
Thanks,
Pete
In looking at the recipe for libspeex, I see that the
--disable-float-api is set.
Is there a reason for this? Is there a better way to do it? (I'm new to speex
and audio encoding/decoding).
Thanks,
Pete
Does it disturb you? Any algorithm implemented in integer works much faster then with floating point.
2010/9/23 Pete Flugstad <pete.flugstad@gmail.com>
Maxim Podbereznyy <lisarden@gmail.com> writes:
Any algorithm implemented in integer works much faster then with
floating point.
That is not true.
Well - I was thinking that the OMAP has the spiffy NEON/VFP units,
which speex lib
is compiled to use. We've switched our code to the integer interface,
so I'm mainly now
just curious as to why it was disabled.
Pete
It was indeed hardcoded to fixed-point eons ago, I just fixed it: http://cgit.openembedded.org/cgit.cgi/openembedded/commit/?id=e06553979d23531397af3dd71870abb80718c681
regards,
Koen