On 03/08/2016 05:53 PM, David Rowe wrote:
> Yes I've noticed that too. It still sounds "coded", but the wideband
> presence is def there. Be nice to do that over narrow band HF channels :-)
It wouldn't actually be so costly to scale codec2 to wideband. Back in
2000, I was taking narrowband
Thanks Matt,
Yes on FreeDV 1600 the microphone and audio source seems to make a
difference. Perhaps also the speaker. Not sure why yet. A good area
for experimentation and research.
Cheers,
David
On 09/03/16 10:26, Matthew Cook wrote:
> Hi David,
>
> Definitely an improvement between the
Hi David,
Definitely an improvement between the Newamp modes and the 700B.
Intelligibility across all new modes was good, only minor artifacts like
"tings" and general "sounded heavily processed" was all that jumped out at
me. The Newamp modes seem to encode short vowels like "e" better, I found
On the low-passed "sound" :
Alot of the time I just find that it is 'perception' and the HF just
needs a bit of an energy boost on re-production.
This is certainly the case when using LMS style noise reduction.
On 9/03/2016 9:53 AM, David Rowe wrote:
>> Another thing I notice now, at least in
Hello David,
well i think the samples sound much better then 700B, but still not as
good as 1300, especially mmt1 sounds quite bad as if the new algorithm
has a problem with the noise suppression?
Also i wonder what a female voice sounds like, because ve9qrp_10s sounds
a bit lowpassed.
greetz
> Another thing I notice now, at least in Australia- Most 3G
> Telstra-Telstra mobile phone calls are all now wideband (AMW-WB)
> (generally the network loading permits it). There is a noticeable
> quality improvement going from land line (3k) to mobile phone...
Yes I've noticed that too. It
Hi Adrian,
Just for completeness, the modem we are planning on using for
FreeDV-over-analog-FM, is more like ASK+Manchester Encoding over FM. We're
calling it FMFSK or MEFSK because it ends up looking like a dirty FSK over
the air. The mod works by Manchester encoding the bits and firing them
(based on my audio background) , I think up to 5.2 or 5.5 kHz is where
the best bang for the buck is. My experiments with subband coding in the
90s (divide up 0to 6kHz spectrum with quadrature mirror filters) was
that I only needed a couple of bits for 4.5 to 6k (plus some scaling
bits which
Another intriguing possibility is add another few samples and do
wideband, but still low bit rate coding. For example a 6 kHz audio
bandwidth at a channel bit rate of 1000 bit/s, and a RF bandwidth of say
500Hz (QSPK with tones spaced at Rs).
It takes very few bits to represent the higher
Maybe try some female voices to see how that tears it up.
To tell the truth I didn't notice much difference between the old 700b
and new 700.
Maybe get rid of the fourth frequency and use the bits for 3 more
voicing bits...
It would seem like 300-3000 should be wide enough??
"The
4th freq frequency histogram has a mean of 3.2kHz. "
I think we need the high frequency support- it adds , i think ,
to the intelligibility in a high noise listening environment.
Especially for younger ears. There is a significant intel
Hi Adrian,
Great work - you should write it up (or YouTube). What noise (channel)
bandwidth are your SNRs measured in? Are all the modems tested below
measured using the same bit rate?
That's why Eb/No is useful - it's normalized to a bandwidth of 1Hz and
bit rate of 1 bit/s. It's like the
Hi all,
Motivated by David's work with VHF modems, I have started to
investigate various Tx-Rx chains that might work with legacy narrow FM
radios. My goal is to develop a completely portable system composed of
a development board running open-source software paired with an FM
transceiver. The
Hello Lists,
I'm working on some new ideas for Codec 2. The aim is to get quality
similar to FreeDV 1600 (which uses Codec 2 at 1300 bit/s) at 700 bit/s.
I have posted some samples here:
http://www.rowetel.com/wordpress/?p=5015
(table of samples part way down the blog post).
Would be
14 matches
Mail list logo
