1. William Walton's stirring film music for Shakespeare's Henry V is well served on the budget Naxos
8.553343. The music and narration from the play is recorded with uncompressed dynamic range. The human voice has among the widest dynamic ranges of
any instrument; heavy compression is routinely done. Not here. Definitely not background listening. Also on NAXOS, his music for Hamlet, 8.553344.
2. From former BAS webmaster Bernard Kinsley: Gene Pitts of TAV sent me a few of the new dual-disc format CD's These are standard
CD on one side, the other side features both DVD 5.1 surround and DVD-Audio 5.1 surround. Now being sold in a few major markets including Boston. Discs
play well in all the players I have tried, but will not be pulled in to my single disc audio cd player in my car (2003 Nissan Maxima). Player tried
to pull them in and they get hung up. CD's seem to be thicker than others, though the PR guy from Silverline
says they meet specs. Has anyone any experience with these in cars? Thanks.
3. The Primera Signature Z1 provides an inexpensive way to make short runs of professional looking CD labels that are waterproof.
It uses a thermal process to print in one of 4 colors and costs about 10 cents to print a disc. Print quality is very good, graphics not so good. List
price is $140. www.primera.com
4. For measuring noise levels, "A" weighting is most commonly employed (NAB/ANSI S1.4-1986). It approximates the ear's
response at a level of 55 dB SPL. However it needs to be used with caution in certain types of measurement. When measuring low level ambience noise,
low frequency energy (such as traffic noise) is given undue prominence the ear's response falls off sharply at low frequencies and low levels.
If there is significant ultrasonic energy present, such as bias from a tape recorder or a pilot tone or digital noise, it can give a misleading reading
because the filter falls off asymtotically at 12 dB per octave at high frequencies whereas human hearing hits a brickwall at around 20 kHz. It turns
out the common network for an A filter, such as printed in the Audio Cyclopedia, falls off at only 6 dB /oct at the high end. At the low end the network's
rolloff is 18 dB per octave but the actual curve is 24. All this results in errors up to 4 dB in the curve, which is good enough for type 2 and 3 sound
level meters. An improved network is shown in Electronics World which,
with a few additional components, is accurate to +.25 -.12 dB and exceeds specs for type 0 and 1 meters. De04
President, Boston Audio Society