Does Onboard Sound Stress The CPU?

[Rob Stow]

You're one brave puppy to change priority for any application to
"realtime" as this may crash your OS.

When I've got an app that needs priority, I set it too high
and if that doesn't do the job I set it too realtime. I've
never once had a crash. If the app that you give realtime
too is a heavy-duty app it will make other apps less responsive
or perhaps even non-responsive to mouse/keyboard input, but it
will never cause a crash. Setting a lightweight app like
WMP or WinAmp to realtime to prevent jerkiness in MP3 playback
is pretty risk-free.

 

[Jim Davis Nature Photography]

Why would you care about CPU usage to play an mp3?

The only time it matters is in games, when the CPU has other stuff to do.

Are you going to change the setup every time you play games. Didn't
think so.

I tested the onboard sound on my P4P800 and it did take some CPU. My
old SBLive doesn't.

Simple math.

 

[Kylesb]

message | On Fri, 26 Mar 2004 00:16:36 +1100, "Darkfalz"
|
| >>
| >> You might get better sound using the SBLive 5.1, but you won't
lower CPU
| >> usage. In fact, the SBLive will probably have higher CPU usage
than the
| >> onboard sound.
| >
| >Why would you care about CPU usage to play an mp3?
| >
| >The only time it matters is in games, when the CPU has other stuff
to do.
|
| Are you going to change the setup every time you play games. Didn't
| think so.
|
| I tested the onboard sound on my P4P800 and it did take some CPU. My
| old SBLive doesn't.
|
| Simple math.
|


Maybe you have one of those "magical" SBLive cards that doesn't use
any CPU but mine sure ate some CPU time.

For example, to render a stereo audio stream at 44khz, 32 voice using
directsound, winbench audio testing showed my SBLive 5.1 used 13% of
available CPU (tested with a 1.4G Tbird CPU with 266 MHz FSB) and 4.6%
to render 16 voices at the same rate.

When I upgraded my system to the A7N8X-DLX mobo and a new CPU (xp2600
at 333 MHz FSB) I reran the tests, so these numbers do not directly
correspond in view of hardware changes, but here's the same test data
on the Nforce Soundstorm system CPU Usage: 44kHz 32 voice directsound
CPU usage: 3.7%; 16 Voice: 1.88% (sure wish I had tested with the old
CPU just for grins, but I had a new case also and did not have much
desire to swap CPUs around much my case PS is very close to the mobo
and it requires me to remove the PS to get the HSF off the CPU).

CPU and FSB changes taken into account, the Soundstorm (onboard audio)
looks to be using less CPU time versus an SBLive as I doubt the
hardware upgrades produced a true 3x improvement in system speeds.
Nevertheless, it's well known the SBLive drivers are not the most
efficient sound solution for PCs.

 

[Jim Davis Nature Photography]

CPU and FSB changes taken into account, the Soundstorm (onboard audio)
looks to be using less CPU time versus an SBLive as I doubt the
hardware upgrades produced a true 3x improvement in system speeds.
Nevertheless, it's well known the SBLive drivers are not the most
efficient sound solution for PCs.

Well, I just used the System Graph app. built into XP to see CPU
usage. Then I used Winamp to play a MP3. Like I said, almost no CPU
power was used. And it's working like a champ in games for me.

 

[Phil]

Well, I just used the System Graph app. built into XP to see CPU

usage. Then I used Winamp to play a MP3. Like I said, almost no CPU
power was used. And it's working like a champ in games for me.

I may be waaaay off here, but when hardware like that uses the CPU it
doesn't seem to show up on the graph, e.g. with my onboard sound when I used
it the temperatures rose, but the % of processor "use" didn't change....just
an observation, but it could be right.....

 

[Jim Davis Nature Photography]

I may be waaaay off here, but when hardware like that uses the CPU it
doesn't seem to show up on the graph, e.g. with my onboard sound when I used
it the temperatures rose, but the % of processor "use" didn't change....just
an observation, but it could be right.....

Well, when I used my onboard sound, the CPU usage graph rose, but the
SBLive didn't make it rise.

I'm not sure how to measure it during a game, but I simply played a
MP3 in Winamp.

I'm pretty happy though with the old SBLive in gaming.