[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[linrad] RE: Linrad 01.09: Soundcard Test Mode
- Subject: [linrad] RE: Linrad 01.09: Soundcard Test Mode
- From: Leif Asbrink <leif.asbrink@xxxxxxxxxxxxxxxxxxxx
- Date: Mon, 1 Dec 2003 20:40:52 +0100
Hi Gerhard,
> Using this function for further investigations concerning my cpu-load
> (overrun errors at high input levels) problem, I found some new details.
>
> On my 2GHz CPU with Delta44 Card the cpu-load ("Graph:") rises up to
> 99.99%,
> as the input signal reaches a certain level. Signals lower than this
> level produces a cpu-load of approx. 3-5% ("Graph:").
Oooh! This is no error! Your data arrives at a rate of 96kHz and you can
not place random pixels at such a high speed without fast graphics.
The setpixel function is fast as long as the pixels are in the same
memory block but when the block is different from pixel to pixel, the screen
is too slow.
Linrad avoids putting pixels like this except when you use the oscilloscope
functions with too much gain so just do not worry.
You may use the "+" and the "-" keys to change the gain so you can
study the noise floor. If you use high gain, there may be warnings
or overrun errors. Perfectly normal but it indicates somewhat slow
graphics. My 2 GHz laptop has this problem with only 44 kHz sampling
rate. With data covering more than a narrow band I get 99.98% time
for the graphics and overrun error.
The good old 600 MHz PIII uses about 60% cpu load to fill the entire
screen with random points in soundcard test mode.
73
Leif / SM5BSZ
LINRADDARNIL