Xref: utzoo sci.electronics:18222 comp.dsp:1342 Path: utzoo!news-server.csri.toronto.edu!cs.utexas.edu!usc!samsung!uakari.primate.wisc.edu!dali.cs.montana.edu!milton!uw-beaver!ubc-cs!sol.UVic.CA!sirius!hedstrom From: hedstrom@sirius.UVic.CA (Brad Hedstrom) Newsgroups: sci.electronics,comp.dsp Subject: Re: A question about the Nyquist theorm Message-ID: Date: 5 Mar 91 19:18:31 GMT References: <20408@shlump.nac.dec.com> <625@ctycal.UUCP> <11515@pasteur.Berkeley.EDU> <1991Mar5.155748.29328@eagle.lerc.nasa.gov> Sender: news@sol.UVic.CA Distribution: comp Organization: University of Victoria, Victoria, BC, Canada Lines: 61 In-Reply-To: tohall@mars.lerc.nasa.gov's message of 5 Mar 91 17:50:02 GMT Nntp-Posting-Host: edwin.uvic.ca In article <1991Mar5.155748.29328@eagle.lerc.nasa.gov> tohall@mars.lerc.nasa.gov (Dave Hall (Sverdrup)) writes: >>jbuck@galileo.berkeley.edu (Joe Buck) writes: >>When I bought my CD player, it said on the front panel 'Dual D/A >>converters'. For fun, I asked the salesperson what that meant. >>The reply was rather funny, and of course completely inaccurate. >> >>What does this really mean ? (I figured maybe two distinct D/A's rather >>than one D/A and two sample and holds). > OK, I may be advertising my ignorance of CD player design here, > but it seems to me it would be very difficult to produce right and left > channel (stereo) outputs without 2 separate D/A's. My lack of expert > knowledge leads me to believe that the 'Dual D/A converters' logo > is like building a car with a V-8 engine and advertising 'Dual Quad > Cylinder Heads' as a unique technical advancement! What is the real > story? Let's hear from some CD technology wizards. I don't profess to be a CD (or any other kind of) guru but here goes. The samples are stored on the CD serially alternating between L and R channels. This means that for each samples time interval, two samples are read from the CD sample L R L R L R ... ----------------------------------------------------------- time t = 0 t = t1 t = t2 ... where the sampling rate, t(n) - t(n-1) = 1/(44.1 kHz) = 22.7 usec. The L and R samples were taken at the same time (in parallel) but can only be read from the CD one at a time (serially). Assuming the L is read first, that means that the corresponding R is delayed by 1/(44.1 kHz)/2 = 11.3 usec. Since the L and R are available in a serial stream, only one D/A is required. At the output of the D/A are two parallel analog sections composed of sample and holds, filters, amplifiers, etc. This is where the "stereoness" in created. Now since CD player manufactures are always looking for that little thing that distinguished their product from the rest, they started offering CD players with 2 D/A's. The reason: as shown above there is a time delay between L and R which translates to a phase difference between the two channels. By using 2 D/A's and delaying the L sample by 11.3 usec, the two channels could be put back in phase. Sounds very impressive. Of course there are numerous "audiophiles" who claim to be able to audibly distinguish single and dual D/A players. To put this phase difference in perspective, assume that your speakers are *optimally* placed in an acoustically perfect room. Further assume that you, the only listener (or object for that matter) in the room, have placed yourself equidistant from the two optimally placed speakers. Now move the R speaker (the one delayed by 11.3 usec) about 1/2" closer than its optimal position. Now you have compensated for the time delay. But don't you dare change your location in the room; you'll throw everything totally out of wack! -- _____________________________________________________________________________ Brad Hedstrom Electrical and Computer Engineering Department University of Victoria Victoria, British Columbia, Canada UUCP: ...!{uw-beaver,ubc-vision}!uvicctr!hedstrom ``I don't think so.'' Internet: hedstrom@sirius.UVic.CA ``Homey don't play that.''