I'd be curious to hear your opinions on audio? There were some changes in the receive path which may or may not make a difference.
I've noted the received audio sounds somewhat sharper, but still "muddy" due to the limited 2450bps vocoder bitrate. This is obviously a contention which cannot be fixed without major changes to the DMR standard, both in terms of the air interface (better compression --> more bandwidth) and the bitrate of the vocoder. Fidelity will never really improve on DMR, unless changes occur.
One other matter which local users constantly remark about, is the delayed AGC attack time for TX audio. There are still very noticeable "spikes" of excessive amplitude at the beginning of a transmission, before the AGC attack/limit circuit can kick in. This should be re-examined by the same crew who implemented the AGC/level fixes in R02.50.
One method of limiting these spikes is to modify the codec codebook so as not to allow the spikes in the RX or TX radio in the first place. This is achieved by defining a maximum allowable amplitude, so the codec does not encode or reproduce the excessive amplitude.
If the engineers are interested in pursuing this, have them take a look at what changes were made in the original ASTRO (Saber/3000/Spectra) DSP firmware, in release 8. In this release, the amplitude spikes were eliminated. Historically, I believed it was the inclusion of the audio AGC feature (as of R07.01.xx and R11.01.xx), but it was not -- as I have forced DSP 8 firmware in to a 512K radio, and completely eliminated the TX amplitude spikes. The radio is unable to encode them with the new(er) DSP firmware.
Sorry for the rambling. I want to see improved audio in the TRBO lineup, and I'm telling you how to do it. Thanks Wayne.
