Re: RRC-06 Maps

In article <W%Sng.1658$eQ.543@xxxxxxxxxxxxxxxxxxxx>, dab.is@xxxxxxxxxxxxxx
says...


Pierre PANTALÉON wrote:
steve41@xxxxxxxxxxxxxx a écrit :

Provide me with a precise definition of "near CD-quality". The main
problem with the term is how far is "near". It could be very close,
it could be a significant distance away.

How do you define "good audio quality"?
How do you define "excellent audio quality"?

Where do you draw between average / good / superb / excellent? Which
is best, superb of excellent? Is dire worse than awful?

Moreover, as I've already explained, I wrote the MPEG Coding page in
2002, and I HAVE learnt a lot since then and I HAVE listened to a lot
of 192 kbps MP2 audio since then - I listen to 192 kbps MP2 on
digital TV every day, and it doesn't sound as good as FM sounds.

What is FM quality caracteristic ?
Dynamic ?
sampling rate ?


I would say that FM-quality with well-engineered audio and good reception
would score very highly on a blind listening test.

Fm has three main problems:

1. The "sampling rate" of the system is 38 kHz, which limits the usable audio
bandwidth to around 16 kHz because lowpass filters are needed to protect the
19 kHz pilot tone and the 57 kHz RDS subcarrier. In fact, the 16 kHz bandwidth
limitation is usually of no consequence to adult listeners, so I consider this
the least of FM's problems.

2. FM uses pre-emphasis at 50us or 75us (50us in Europe and the UK), which
limits the HF headroom of the system. Avoiding overdeviating the TX requires
either reducing the average modulation (which compromises the S/N ratio at the
receiver) or using HF limiting, which is a form of dynamic HF filtering. It is
very difficult to make an FM signal be audibly indistinguishable from a modern
(i.e., so bright it makes your ears bleed) CD.

3. In practice, there is enough multipath at the receiver to cause
significantly audible nonlinear distortion. This is true even with a carefully
aimed outdoor antenna. I speak from experience, having tried to do audio
"proof of performance" measurements through remote radios and finding measured
THD typically in the 0.5% to 1% range due to multipath. This is why in the US,
these "Proofs" were always done via a demodulator connected via a sample loop
directly to the final output amplifier of the TX.

And there is one minor problem: The bandwidth of a modulated FM carrier is
infinite. Fortunately, the energy of the higher-order sidebands falls off
rapidly, which allows practical receivers (with RF selectivity) to be created.
However, as soon as the FM signal is bandpass-filtered, the output of the FM
detector exhibits nonlinear distortion. Indeed, the major drawback of FM is
that linear filtering between the output of the modulator and the input of the
demodulator causes nonlinear distortion upon demodulation. In the RF domain,
multipath is just a linear comb filter, after all.

.

Relevant Pages

  • Re: RRC-06 Maps
    ... How do you define "excellent audio quality"? ... which is a form of dynamic HF filtering. ... demodulator causes nonlinear distortion upon demodulation. ... Be all the above as it may FM for critical listening is still better ...
    (alt.radio.digital)
  • Re: Need ideas for 8-channel output
    ... scale the audio properly to avoid the 50% PWM duty cycle with no audio ... with no audio present is also 2.4 V. Wouldn't the PWM be at 50% if I ... The only problem here then is the filtering as the ...
    (sci.electronics.design)
  • no Auduio software?
    ... or is the ports collection completely devoid of any ... filtering of audio files? ... memory that could do audio wave analysis and filtering, ...
    (comp.unix.bsd.freebsd.misc)
  • Re: unsure: filter idea
    ... this one is related to signal filtering, eg: for images or audio ... typically been using png for images, at least for what cases I am not using ... I could, for example, implement adpcm, but adpcm both sounds poor imo and ... lossy filtering may be needed though if any decent compression is desired. ...
    (comp.compression)
  • Re: Slew Rates and Mic Pre performance
    ... > Don't take Phil too seriously. ... > reflected power. ... These days if we want to send audio any significant distance ...
    (rec.audio.pro)
Loading