Miscellaneous News

1. The BAS is looking for a new webmaster. The current webmaster will train you and hand it over to you, and provide support as necessary. The website is written in simple HTML. You will need a computer and a high speed internet connection (you will need to download a 6GB backup in a reasonable amount of time). $300 is the annual payment. You may be asked to support admin of the BAS Facebook pages as well.

2.  V46n1 of the BAS Speaker has been published.  It features the January 2024 meeting, Open forum and Flea Market, Low Noise Design for MC Preamps by David Hadaway and Over the Record Counter with Moeran’s Symphony in G minor.  Also a note on speaker protection circuits by Tom Perazella.

3. Compression at the BSO

It is generally believed that compression as used in FM broadcasts is non-reversible—you can’t get back the original.  However the BSO (Boston Symphony Orchestra) stream on WCRB uses a simpler form known as limiting.  The output follows the input up to a threshold, then infinite compression is applied—for all input levels above the threshold the output is constant.  You can see this in the waveforms where the signal envelope looks as if a hedge trimmer has clipped them off.  (BASS 43n1p4)  (However there is no actual clipping, only gain reduction). It’s easy to imagine extrapolating the signal to recover the lost peaks using a judicious amount of increased gain, say 4-8 dB.  Of course the overall level has to be lowered to allow for the higher peaks.  An attack time of 50 msec and release time of 1 msec might be reasonable. Since the threshold level is always the same, just below full scale, the process can be automated.  Comments?

[I thought that an expansion program would work but then realized that it can’t work when there is no change of amplitude with time, as in this case where it is flattened off.   I then thought of Izotope’s “Declip” feature which looks for clipped waveforms and extrapolates upward in amplitude to give a rounded waveform.  This works very well for its intended purpose, but in this case it had no effect.  It is looking for sharp edges and there are none here.  I then did it manually (4 dB) and you can hear the result, before and after, at https://dbsystemsaudio.com/tchai.wav.  I’m sure there are editors out there that will do everything I would want, but the money and time involved are beyond my interest.]