Sound Blaster 16 Value Edition PnP tested
Creative Labs has made it's Sound Blaster soundcard series very popular in multimedia PC. Nowadays Creative Labs sells the soundcard series: SB16, AWE32 and SB32. Those soundcards have different kind of capabilities, but the technical specifications. Those technical specifications could be more delailed, but this is a good start. One question arises: are those technical specifications listed real or are they just marketing hype ?
I wanted to make sure of that because a am looking for a good soundcard and I wanted to get unbiased test results. I tested SoundBlaster 16 Value Edition PnP soundcard installed to Octec computer (133 MHz Pentium computer with Rhino 6 motherboard in mini tower case). The card was installed to the card slot in the bottom of the case three slots away from other cards. I tested the soundcard using Nakamichi T-100 audio analyzer and WaveStudio sampling software which comes with the card. In the test I used stereo recording at 44.1 kHz sampling rate and 16 bit resolution to get best sound quality. All tone conrols were in center position all the time. I wanted to test what is the card in it's best. And here are the results.
Line level signals
I tested line level frequency response using following method: I used Nakamichi as signal generator to generate sine wave test signals, I used WaveStudio to digitize and play back those samples and finally I measured the output with Nakamichi. The signal levels I used were 0.78 Vrms. The frequency response for this type of recording and playback process can be seen in figure below.
The frequency response is straight from 20 Hz to 15 kHz with +-0.3 dB, which is very good value. The frequency response starts to drop at frequencies above 15 kHz and drops up to 3 dB at 20 kHz frequency. This frequency response is quite good. I also tested the THD distortion at 400 Hz frequency and signal levels on 0.78 Vrms input and output and I got THD+noise value 0.03% which is quite good. The FFT spectrum which shows out the harmonics could still be better.
I used the same method for microphone line testing as for line level signals. The difference is that I used lower signal level (0.078 Vrms) for microphone to avoid overloading the microphone preamplifier. Microphone input in Sound Blaster soundcards is mono and it has one special feature: there is some power (5.2VDC feed through 2.2 kohm resistor) available in microphone connector for electret microphone capsule. I did not need that power because I used Nakamichi as the signal source. Nakamichi had 600 ohm output impedance, which is also the value specified for soundblaster microphone. In the figure below you can see the results.
Microphone line frequency response is horrible. It is only a little better than ordinary telephone line frequency response. The microphone input is not suitable for anything alse than recording low quality speech. Any sample rate higher than 11 kHz is waste of storing page. Microphone input can't be used for anything serious.
If you are thingking of replacing the cheap microphone which cames with some Sound Blaster cards, I advice you to also buy and external microphone preamplifier also and connecting the microphone with it to line level input.
One thing which might interrest many users is the noise levels. Sound cards from Creative Labs have reputation fo being noisy and easily picking up all interferences from inside the computer. Nakamichi audio analyzer can be also used for noise level testing, so I measured also the noise levels. For measuring I set the ouptut level so that card gives about 0.78Vrms (0 dB reference) output level from sample playback sample channel at maximum level. I left output level to this position for all measurements.
First I turned all signal sources to minimum level and turned microphone input off. I got noise levels od -85 dB A-weighted and -76 dB in linear scale. Then I turned all signal source to maximum level, still leaving microphone input off. I measured noise level now being -77 dB A-wighted and -62 dB in linear scale. In last noise test I also turned the microphone input on and at maximum level. Not I got -70 dB A-weighted and -60 dB in linear scale.
Tomi Engdahl <firstname.lastname@example.org>