EERIG07@TECHNION.BITNET (04/23/91)
i am looking for the cheapest hardware available for sampling voice data into an ibm pc. i need fixed rate 8khz sampling, linear (not mu-law or such), and no d/a capabilities. preferably connecting to the parallel port, but internal boards are okay too. any ideas out there? thanks, noam amir EE dept. Technion Haifa Israel email: noam@techunix.technion.ac.il eerig07@technion.ac.il
yhchung@aix01.aix.rpi.edu (ying-hung chung) (04/24/91)
The Sound blaster from Creative lab can sample data from 4KHz up to 13Khz (I check the mannual but it never mentions anything about the method it uses I think it uses just PCM with no companding) By the way, I just saw an ad somewhere in forsale.comp someone is selling a SB for $90. A new SB is about $150
tmottonen@cc.helsinki.fi (04/25/91)
In article <1ksgbd-@rpi.edu>, yhchung@aix01.aix.rpi.edu (ying-hung chung) writes: > The Sound blaster from Creative lab can sample data from 4KHz up to 13Khz > (I check the mannual but it never mentions anything about the method it uses > I think it uses just PCM with no companding) My SB manual says: "1 Digitized voice channel (DAC), programmable variable sampling 4KHz to 23KHz, compression schemes: - 8 bit data, no compression - 2 to 1 compression: 4 bit ADPCM, hardware decompression - 3 to 1 compression: 2.6 bit ADPCM, hardware decompression - 4 to 1 compression: 2 bit ADPCM, hardware decompression" and "8 bit A/D conversion of sound signal, variable sampling rate: 4KHz to 12KHz". Could someone explain to me what effect the sampling rate has on the frequency (i.e does 12KHz sampling mean 20-12000Hz, 4KHz 20-4000Hz etc) ? Teemu. -- Teemu Mottonen / Department of Computer Science / University of Helsinki tmottonen@cc.helsinki.fi | TMOTTONEN@FINUH | tmottone@kruuna.helsinki.fi address: Vapaalantie 26B, SF-01650 Vantaa, Finland / tel: + 358 0 844938
hchung@aix01.aix.rpi.edu (yinghung chung) (04/25/91)
The sampling frequency is dependent on the frequency your of signals. normally, the sampling frequency has to be twice the highest frequency of your input signal. For example, if the highest frewquncy is 4KHz then your sampling freqeuncy must be at least 8Kz. If the sampling freq is less than 2x, you will have aliasing problem For a male voice, a smapling freq at about 8Khz should be good enough However, for a female voice, you will need higher than 8kz as female voice has a higher pitch than male voice Yhchung