Souncard Testing Methods

Polliwog · Oct 23, 2003, 01:46 PM

Greatings,

I went though the previous thread about sound quality and testing.
You folks have been working hard! Lots of interesting discoveries.

While looking into a new sound card, Daniel pointed me to the TOMS HARDWARE site. I was impressed with the sound quality graphs, having generated a few myself last year with shareware to test my speaker fidelity, recording room acoustics,...

So I did some searching to find REAL MARK. It is a shareware that is really intesresting for us. The documentation in the RMAA5.1 package explains the FFT math, human hearing sensitivity (hence A-weighting for SNR testing) and many other issues, including a bit about 48kHz locked cards (sound familiar?) and their difficulty with IMD (intermodulation distortion).

Also SNR testing tests with no signal, like you have been doing, however results are corrected to full scale signal (ah so!). Moreover, DR (Dynamic Range) measures noise with a weak signal since some cards cheat by clamping down volume at low levels (like a compressor), though I don't think this is the case with our cards.

What is really cool is that the software generates it's test signals and analyses them. I did several retests with the same card and results rae very consistent. You can test a card full duplex (plug output into input) or test one card with another. The results take into account BOTH PLAYBACK and RECORDING. You can also graph 4 tests.

One thing I noticed is that drivers have a lot to do with sound quality. Big differences in sound quality here between WAVE MAPPER and ESS AudioDrive (a cheap aopen card, don't comments pls, just testing soft). I intend to run several tests on my live ct4832 with different drivers (my NT runs creative drivers) and differnet DSP window configurations.

NOTE: Special setup for AUDIGY (with Creative Drivers) because of a feedback loop issue.
Performs very well when setup correctly (a default gamer setup issue).

Cheers,
Paul.

Polliwog · Oct 23, 2003, 02:01 PM

I did not search the site using the right keywords. I see that Right Mark has been used for quite some time now for kX project. Many postings in the GENERAL discussion area.

Found the links to the extensive testing of inputs and outputs with kX. You folks are very rigorous!

RMAA is quite interesting.

Cheers,
Paul

Oct 23, 2003, 01:46 PM	#1
Polliwog DriverHeaven Junior Member Join Date: Oct 2003 Location: Quebec City, Canada Posts: 25 Rep Power: 0	Souncard Testing Methods Greatings, I went though the previous thread about sound quality and testing. You folks have been working hard! Lots of interesting discoveries. While looking into a new sound card, Daniel pointed me to the TOMS HARDWARE site. I was impressed with the sound quality graphs, having generated a few myself last year with shareware to test my speaker fidelity, recording room acoustics,... So I did some searching to find REAL MARK. It is a shareware that is really intesresting for us. The documentation in the RMAA5.1 package explains the FFT math, human hearing sensitivity (hence A-weighting for SNR testing) and many other issues, including a bit about 48kHz locked cards (sound familiar?) and their difficulty with IMD (intermodulation distortion). Also SNR testing tests with no signal, like you have been doing, however results are corrected to full scale signal (ah so!). Moreover, DR (Dynamic Range) measures noise with a weak signal since some cards cheat by clamping down volume at low levels (like a compressor), though I don't think this is the case with our cards. What is really cool is that the software generates it's test signals and analyses them. I did several retests with the same card and results rae very consistent. You can test a card full duplex (plug output into input) or test one card with another. The results take into account BOTH PLAYBACK and RECORDING. You can also graph 4 tests. One thing I noticed is that drivers have a lot to do with sound quality. Big differences in sound quality here between WAVE MAPPER and ESS AudioDrive (a cheap aopen card, don't comments pls, just testing soft). I intend to run several tests on my live ct4832 with different drivers (my NT runs creative drivers) and differnet DSP window configurations. NOTE: Special setup for AUDIGY (with Creative Drivers) because of a feedback loop issue. Performs very well when setup correctly (a default gamer setup issue). Cheers, Paul. __________________ ASUS P4 1.7 512MB Ram, 80GB 7200rpm, Nvidia Quadro2, w98se (dual boot NT4), Audigy Platinum SB0090 SbLive Value CT4832

Oct 23, 2003, 02:01 PM	Thread Starter #2
Polliwog DriverHeaven Junior Member Join Date: Oct 2003 Location: Quebec City, Canada Posts: 25 Rep Power: 0	I did not search the site using the right keywords. I see that Right Mark has been used for quite some time now for kX project. Many postings in the GENERAL discussion area. Found the links to the extensive testing of inputs and outputs with kX. You folks are very rigorous! RMAA is quite interesting. Cheers, Paul __________________ ASUS P4 1.7 512MB Ram, 80GB 7200rpm, Nvidia Quadro2, w98se (dual boot NT4), Audigy Platinum SB0090 SbLive Value CT4832