falk@sun.uucp (Ed Falk) (02/15/86)
Just a few notes about digital audio etc... Before I moved to CA, I worked at radio station (WRPI, Troy, NY) that prided itself on the quality of its audio signal. One of the things the station did in recent years was to obtain a digital tape recorder for concerts etc. The digital machine we used was a Sony PCM-F1 which worked by taking 1/30 sec worth of audio at a time and converting it to a television picture full of black and white dots (along with a lot of error-correcting code) which you could record on any cheap video-tape recorder (the PCM-1 and Sony's portable Betamax deck are a matched set). The end result was a portable recording setup that weighed about thirty pounds, cost about $2500 and could record about three hours of 16-bit stereo audio on a tape that cost $8. WRPI's top of the line analog recorder weighed about 200 pounds, cost $7000 and recorded 45 minutes of audio on a $25 tape. On top of all that, the PCM audio quality was *much* better than the analog -- twice the dynamic range, NO wow and flutter, no hiss or intermodulation or any of the other problems inherent in the analog recording technique and it never needs to be aligned (an analog machine needs to be properly aligned -- a one-hour job -- before each recording). When we combined all of this with a pair of high-quality mikes (Neuman U-87's) and a really superb concert hall (Troy Music Hall was the protype for Carnegie Hall), we got some really spectacular recordings. More than once I walked into the station to the sound of piano music and asked "who tuned the piano in the back room?" only to find out that someone was playing back a concert recording -- the reporduction was so accurate that you literaly could not tell that you were listening to a recording. All this leads to the question of what makes a good CD recording. The best that any recording medium can provide is an exactly accurate playback of the audio that was fed into it in the first place. If a recording is made from a bad analog master, or bad mikes or bad miking technique or a just plain bad performance, you'll get a bad CD. In fact, you're probably worse off than if you use vinyl or cassette to distribute the recording because the flaws become painfully clear. For example, the recordings made in the Troy Music Hall are often marred by someone's digital watch beepin during a performance. Coughing and foot-scuffling also come across crystal-clear. The best CD's are those made directly onto digital tapes and which never see an analog stage before they're transcribed onto disk. Most classical music is recorded digitally nowadays. Most rock music released on CD was recorded on analog master in the first place (in such a case, you might just as well buy a vinyl recording and save your money). Miking technique is also critical. My friend from the station who was the expert on these matters said that he considered a recording acceptable if he could close his eyes and point to the location of every instrument on stage. I don't have a very good ear, and was skeptical that this could even be done, but he once had me listen to a digital recording on his home system, which had very good speakers (the other really critical component of sound reproduction). The result was that I could not only point out where every instrument on stage was, but I could also say how far away it was, how wide the stage was, how far back the back wall of the stage was and where the people in the audiunce were sitting (none of these features matched the shape of his living room by the way). I was astounded to find out that I had it all correct. My main complaint with CD's are the cost, both of the disks and the equipment. A decent CD player will run you at least a thousand dollars. This is not because the player is all that expensive -- they run about the same as comparable analog turntables -- but because as soon as you get the thing home and listen to it on your home system, you'll realize that you need much better speakers than you currently own. Whenever I listen to a digital recording at the radio station or my friend's house, I can't stand to listen to my own system for the next few days. As far as miking goes, my friend always had his best results with a pair of Neuman U-87's placed in what's called "coincident cardioid" pattern, which is to have two cardioid mikes right on top of each other, facing outward at 90-degree spacing along with a pair of PZM microphones on the stage for ambience and to pick up the performers when they're not at their instruments. Sometimes he hung a couple of omni mikes from the ceiling for ambience and audience reaction. There were other techniques tried as well. There was one guy at the station who was doing his PhD on concert-hall acoustics who experimented with binaural recording -- this is where you plant mikes in the ears of an acoustically correct maniken and place the maniken in the audiunce. This worked amazingly well when you wore headphones. One time my friend was sitting at the recorder (in a back room) with headphones on before the concert and he heard footsteps behind him walking up. Then a voice said "pardon me, could you -- oh my goodness" and the steps went away quickly in another direction. This was, of course, someone who had mistakenly walked up to the dummy to ask it a question. To my friend, it sounded like there was someone in the room with him. David Albrecht asks about miking piano. At WRPI, we miked pianos by placing two mikes (preferably Neumans) in the piano with the strings -- one over the base strings and one over the high strings. I don't think PZM's are the way to go for instrument miking (don't quote me on this) as they are meant to be laid flat on the stage (they work quite well for this too). Try to use condensor mikes if you can. Never use high-impedance mikes -- these are invariably cheap toys, not to be used for serious recordings. Normally I would also say never, never, never use Radio Shack mikes (or anything else), however the Radio Shack PZM has the potiential for being a good mike. That's because the Radio Shack PZM is actually made by a real manufacturer (I forget who at the moment). What you buy at Radio Shack are rejects that didn't quite make spec. If you're lucky, you can get a reasonable mike this way, for about a quarter of what the manufacturer charges for the real thing. As for overload when miking a piano, this is most likely in the electronics and not the mike itself. Try placing a pad (attenuator) between the mike and the pre-amp. If the mike electronics are overloading, you're probably out of luck. Edward A. Falk