SBLive 5.1 PCI or P5WD2 builtin Realtek ALC882D High Definition Audio 8-channel CODEC

[John]

Hi

What's going to give me the best sound quality? My current SBLive 5.1 PCI
sound card, or the built-in sound on the Asus P5WD2 (it's a Realtek ALC882D
Hi Def 8 Channel CODEC)?

I'm assuming the SBLive PCI card is going to consume less of my cpu time. Is
that correct?

Thanks

John

 

[Stephan Grossklass]

Hi

What's going to give me the best sound quality? My current SBLive 5.1 PCI
sound card, or the built-in sound on the Asus P5WD2 (it's a Realtek ALC882D
Hi Def 8 Channel CODEC)?

Judging from the datasheet of the ALC882, it could well be better than
the Live! 5.1 with its crappy old Sigmatel (STAC9708) codec. If you want
to be sure, install both and compare.

I'm assuming the SBLive PCI card is going to consume less of my cpu time. Is
that correct?

For games, it probably is.

Stephan

 

[Paul]

Judging from the datasheet of the ALC882, it could well be better than
the Live! 5.1 with its crappy old Sigmatel (STAC9708) codec. If you want
to be sure, install both and compare.


For games, it probably is.

Stephan

Whether a sound chip has 18 bit ADC/DAC or 24 bit ADC/DAC,
the noise floor of the PCB determines whether the resulting
dynamic range is useful or not. Anyone who has turned up the
volume on their motherboard sound chip, to be greeted by the
noises from their mouse, or the noises from the processor
as the computing load changes, knows how important the noise
floor is. Same thing with trying to record audio via the
motherboard sound chip - a bad noise floor will ruin your
attempts at recording.

Even if the SBLive 5.1 had 16 bit DACs, I bet it would sound
cleaner.

Doing an A/B comparison with the two solutions, is the only
way to quantify the differences between them. Some Asus
motherboards have terrible analog designs - like analog signal
wires running in parallel with Ethernet signal wires. No matter
how expensive an AC-97/Azalia chip is used, it cannot fix a
bad design.

Paul

 

[John]

Whether a sound chip has 18 bit ADC/DAC or 24 bit ADC/DAC,
the noise floor of the PCB determines whether the resulting
dynamic range is useful or not. Anyone who has turned up the
volume on their motherboard sound chip, to be greeted by the
noises from their mouse, or the noises from the processor
as the computing load changes, knows how important the noise
floor is. Same thing with trying to record audio via the
motherboard sound chip - a bad noise floor will ruin your
attempts at recording.

Even if the SBLive 5.1 had 16 bit DACs, I bet it would sound
cleaner.

Doing an A/B comparison with the two solutions, is the only
way to quantify the differences between them. Some Asus
motherboards have terrible analog designs - like analog signal
wires running in parallel with Ethernet signal wires. No matter
how expensive an AC-97/Azalia chip is used, it cannot fix a
bad design.

Paul

Thanks for the replies and the info. I'll have to test them. I don't
really trust my own ability to draw any meaningful conclusions
though.

 

[Paul]

Thanks for the replies and the info. I'll have to test them. I don't
really trust my own ability to draw any meaningful conclusions
though.

Try some high dynamic range classical music - a selection with
silent passages in it. Move the mouse around in the quiet passages
and see if you can hear mouse noise.

Some chips are ruined by the lousy DSP special effects, where
even when the control panel is set to "disabled", to turn them
off, they are still there. I had that problem with an AD1985
and SoundMax.

If you need a very basic audio waveform tool, try this one:

http://audacity.sourceforge.net/download/windows

What I tried, is using two programs (because I couldn't get
Audacity to record and playback at the same time). One program
was set to play test output, while the other was set up to
record. I connected output to input with a short 1/8" stereo
cable. What I discovered is, if I sent a pulse waveform on the
AD1985 output, two pulses would come back, one at t = 0 sec
and a much smaller one at t = 30mS. That means some driver
writer decided I needed "concert hall reverberation", so I
would feel I was "in a cave". Too much reverb makes music
sound "muddy". I should be able to drive a signal through
the sound chip, without the driver adding its share of
"garbage". If the driver offers a multiband equalizer,
that is not a bad thing, if you are correcting for coloration
in the room. If the equalizer is set to 0dB on all sliders,
there should be no effect at all on the output.

I'm no expert at this stuff, and I still don't know how to
draw waveforms in Audacity that are bandwidth limited. Using
sine waves (tones) to test, doesn't tell you too much,
except if maybe you are looking for harmonic distortion.

Given all of that, a classical music test should tell you
most of what you need to know. You can also try recording
from a microphone, as some users report that gives them an
instant reason to switch to a PCI audio card.

If you are playing CDs, the best way to do that, is use
DAE, where data is shipped digitally over the IDE cable.
The four wire CD analog cable that comes with a lot of
CD drives, works like an antenna and that can introduce
noise. Digital audio extraction (DAE) is a much better
option.

Paul

 

[Stephan Grossklass]

Some chips are ruined by the lousy DSP special effects, where
even when the control panel is set to "disabled", to turn them
off, they are still there. I had that problem with an AD1985
and SoundMax.

This one mostly seems to be plagued by bass expansion that is turned on
by default. One needs to enable the respective slider in the mixer, then
it can be muted. Another common problem with onboard sound solutions is
noise caught by additional inputs. Disabling all of those for playback
can clean up things considerably.

Given all of that, a classical music test should tell you
most of what you need to know. You can also try recording
from a microphone, as some users report that gives them an
instant reason to switch to a PCI audio card.

Actually mic inputs don't have the reputation of being very good on most
consumer-level cards. (CMedia based cheapies can be as bad as onboard
stuff.) The cheapest way to obtain a usable mic input seems to be
purchasing an inexpensive mixer that contains one.

If you are playing CDs, the best way to do that, is use
DAE, where data is shipped digitally over the IDE cable.
ACK.

The four wire CD analog cable that comes with a lot of
CD drives, works like an antenna and that can introduce
noise.

Actually it's mostly a ground loop that causes these problems, due to
the connection over both the PC case and audio cable. (You could call it
a "ground loop antenna".) Some sound chips like older Crystal AC97
codecs (e.g. CS4294) provide "pseudo-differential" CD inputs because of
this. Besides, the DACs in CD/DVD drives are not likely to be of great
quality, only fairly early (and thus expensive) models used TDA1541s and
such.

Stephan

 

[John]

Try some high dynamic range classical music - a selection with
silent passages in it. Move the mouse around in the quiet passages
and see if you can hear mouse noise.

Some chips are ruined by the lousy DSP special effects, where
even when the control panel is set to "disabled", to turn them
off, they are still there. I had that problem with an AD1985
and SoundMax.

If you need a very basic audio waveform tool, try this one:

http://audacity.sourceforge.net/download/windows

What I tried, is using two programs (because I couldn't get
Audacity to record and playback at the same time). One program
was set to play test output, while the other was set up to
record. I connected output to input with a short 1/8" stereo
cable. What I discovered is, if I sent a pulse waveform on the
AD1985 output, two pulses would come back, one at t = 0 sec
and a much smaller one at t = 30mS. That means some driver
writer decided I needed "concert hall reverberation", so I
would feel I was "in a cave". Too much reverb makes music
sound "muddy". I should be able to drive a signal through
the sound chip, without the driver adding its share of
"garbage". If the driver offers a multiband equalizer,
that is not a bad thing, if you are correcting for coloration
in the room. If the equalizer is set to 0dB on all sliders,
there should be no effect at all on the output.

I'm no expert at this stuff, and I still don't know how to
draw waveforms in Audacity that are bandwidth limited. Using
sine waves (tones) to test, doesn't tell you too much,
except if maybe you are looking for harmonic distortion.

Given all of that, a classical music test should tell you
most of what you need to know. You can also try recording
from a microphone, as some users report that gives them an
instant reason to switch to a PCI audio card.

If you are playing CDs, the best way to do that, is use
DAE, where data is shipped digitally over the IDE cable.
The four wire CD analog cable that comes with a lot of
CD drives, works like an antenna and that can introduce
noise. Digital audio extraction (DAE) is a much better
option.

Paul

Hi Paul,

Thanks for the comprehensive and informative reply.

I'll check those things out as you suggested.

Regards

John

 

[John]

This one mostly seems to be plagued by bass expansion that is turned on
by default. One needs to enable the respective slider in the mixer, then
it can be muted. Another common problem with onboard sound solutions is
noise caught by additional inputs. Disabling all of those for playback
can clean up things considerably.


Actually mic inputs don't have the reputation of being very good on most
consumer-level cards. (CMedia based cheapies can be as bad as onboard
stuff.) The cheapest way to obtain a usable mic input seems to be
purchasing an inexpensive mixer that contains one.


Actually it's mostly a ground loop that causes these problems, due to
the connection over both the PC case and audio cable. (You could call it
a "ground loop antenna".) Some sound chips like older Crystal AC97
codecs (e.g. CS4294) provide "pseudo-differential" CD inputs because of
this. Besides, the DACs in CD/DVD drives are not likely to be of great
quality, only fairly early (and thus expensive) models used TDA1541s and
such.

Stephan

Thanks for the additional info Stephan