Subject: Re: [ecasound] best way to account for DC offset?
From: Kai Vehmanen (k@eca.cx)
Date: Fri Mar 16 2001 - 11:54:49 EET
On Mon, 12 Mar 2001, Jon Nall wrote:
> is it better to post-process a file to find DC offset (a la ecafixdc) or
> should i "record" about 30 seconds of "blank" audio with -ezf enabled to
> find the DC offset, and then use that real-time when recording? i've heard
> arguments both ways.
Adding the DC-offset in realtime does makes sense, but you'd need to do
this at the soundcard level (ie. soundcard/card-driver-sw would have to
apply the offset before the A/D conversion). After the data has been
converted into digital form (for instance signed 16bit), applying the
offset won't do any good. After the signal reaches ecasound, it is
converted into ecasound's internal format (32bit float), which can handle
large DC-offsets without loss of dynamics or precision.
So shortly-put, realtime -ezf isn't any better that ecafixdc. The only
exception is when the soundcard is sampling data with a much higher
precision than the produced output file (32bit>16bit for example). But if
you have a highquality soundcard, you rarely want to throw away the extra
precision...
-- . http://www.eca.cx ... [ audio software for linux ] /\ . . http://www.eca.cx/aivastus ... [ aivastus net radio ] /\ .-- To unsubscribe send message 'unsubscribe' in the body of the message to <ecasound-list-request@wakkanet.fi>.
This archive was generated by hypermail 2b28 : Fri Mar 16 2001 - 21:19:50 EET