Subject: Re: [ecasound] alsa device as input
From: Kai Vehmanen (email@example.com)
Date: Mon Feb 05 2001 - 10:51:49 EET
On Mon, 29 Jan 2001, Jeremy Hall wrote:
> I have found that in the past, /dev/dsp was more efficient than alsa
> support. This didn't make sense to me at the time, but I found a bytes to
> frames error in the alsa source (audioio_alsa3.cpp)
> I haven't checked to see if it still exists.
The bytes-frames confusion was created when ALSA people decided to change
all sample counts into frame counts (during 0.6.x developemnt). This
doesn't directly affect performance, but can mess up the audio output. For
instance, for normal cd-quality audio, 1 frame consists of a pair of
2-byte (16bit) samples (left+right). So confusing sample and frame
values causes lots of trouble.
It's actually quite hard to compare the performance of different sound
drivers. And especially with consumer cards, where the data stream
bandwidths are relatively small (cd-audio 176kB/s). You can't really use
xruns for comparison, as only ALSA reports them. You can suffer from
occasional xruns with OSS drivers without ever knowing about the
-- . http://www.eca.cx ... [ audio software for linux ] /\ . . http://www.eca.cx/sculpscape [ my armchair-tunes mp3/ra/wav ]
-- To unsubscribe send message 'unsubscribe' in the body of the message to <firstname.lastname@example.org>.
This archive was generated by hypermail 2b28 : Mon Feb 05 2001 - 11:28:57 EET