|
From: Martyn S. <Mar...@ao...> - 2007-07-07 01:02:32
|
Hi there I'm sure that the response of meters is probably an emotive issue with at least a few people out there, so I thought I'd bring it up, since I've been looking at what we currently have and thinking about how to make it even better. A few questions: We mention 'VU' meters in the code but are not implementing a VU meter (IEC 60268-17, I believe, although I don't have a copy to hand). Evidence? VU meters are supposed to use the average of a rectified signal, we use rms. Question: to make it right would take some effort; is it worth it, given that VU (may) stand for 'virtually useless' in a digital environment? We implement a peak meter but it isn't PPM (BS 6840-17:1991). I believe that Steve Harris has done some successful work on this for 'JACK Meterbridge', but I'm on windows. Is this worth persuing / using? (I have not looked into it.) libbaptools tries to implement accuratly a BBC-type PPM meter, but I can't see an easy way of using this in Audacity. And the BBC PPM dates from 1938 and would not show the odd short peak, more relevant (perhaps) in our digital age. I believe that BS 1770 is the latest thing, should we try and implement that? I'm all for standards, there just seem to be so many of them! And I have not read this one. There are a few interesting papers on the subject, this on mentions Audacity www.kgw.tu-berlin.de/~lac2007/papers/lac07_cabrera.pdf Do you know of others I might like to read? So for my contribution to the debate (if you are on Windows), please have a look at the extra bit on the meters at http://members.aol.com/martynshaw/audacity.zip, inspired by a personal communication with Andrew Mason, of libbaptools. Try it undocked and larger! So, ideas? Martyn |