Message-ID: <4EFDE37F.5080103@freenet.de>
Date: Fri, 30 Dec 2011 17:14:55 +0100
From: wolf_dl4yhf <dl4yhf@freenet.de>
User-Agent: Mozilla/5.0 (Windows NT 5.1; rv:8.0) Gecko/20111105 Thunderbird/8.0
MIME-Version: 1.0
To: RSGB LF Group <rsgb_lf_group@blacksheep.org>
Subject: LF: Re: SpecLab &  Ros  data  mode
Content-Type: text/plain; charset=ISO-8859-15; format=flowed
Content-Transfer-Encoding: 7bit
Sender: owner-rsgb_lf_group@blacksheep.org
Precedence: bulk
Reply-To: rsgb_lf_group@blacksheep.org
X-AOL-SCOLL-URL_COUNT: 0  
x-aol-sid: 3039ac1d601a4efde44a23f1
X-AOL-IP: 195.171.43.25
X-AOL-SPF: domain : blacksheep.org SPF : none


Hello Graham,

Sorry for having to post this response on the reflector... the usual 
annoyance with certain email providers.
Here is what your provider just "said" :

>  host mx4.hotmail.com [65.55.37.120]: 550 SC-001 (COL0-MC4-F14)
>  Unfortunately, messages from 195.4.92.92 weren't sent.
>  Please contact your Internet service provider since part of their network is on our block list.

In other words, freenet.de is blacklisted by hotmail.com . Aargh.

Here is what I originally wanted to reply:


Of course Spectrum Lab cannot affect the sampling rate delivered by 
Windows to another application.
The problem is, as so often, windows itself. It tries to be "very 
helpful" when multiple programs use the same audio hardware.
The soundcard's A/D converter (hardware) can only run one sample rate. 
This is ok as long as only one program uses it.
But when two programs try to use it, with *different* sampling rates 
(especially when they are not integer multiples) some bad things happen 
- called resampling. Windows does an extraordinarily poor job at this... 
for example, it may decide that to deliver 11.025 kSamples/second to 
application "A" and 48.000 kSamples/second to application "B", it's 
guesswork at which sampling rate the ADC is actually run. (I often get 
the impression that in these cases windows rolls the dice). It may be 
48000 S/s, but then the 11025 S/s output will be difficult. It may also 
be 44100 S/s, then the 11025 S/s output receives "every 4 th sample" (in 
simple terms), but constructing the 48000 S/c output will require some 
rather tricky (and poor-quality) fractional resampling.

That's why it's the operator's duty to configure all programs to use the 
same sampling rate. 48000 Samples/sec seems to be a good choice these days.

Also all the best for 2012,
    Wolf .

----
Am 30.12.2011 15:31, schrieb G ..:
Wolf ,
Over the  last days  , I have  run  tests  with  Jose  , Ros , with  
Speclab running  on the  same  PC ,
During  some  qso's  I noted   Ros  lost  decode  after a  short time  
with  good   signal  level , after  some  repeats  , I found that  
,only  when  Speclab  was  running , was this  problem..  I don't  
think  this  is a  new  problem as I have  some  lock  problems  with 
the  Ros mode  on MF , which could not  be  explained
Jose  coded a  test  version  of Ros  and this  showed that  when  
Speclab  was  running , then  Ros needed  to  compensate the  sample  
rate ..the  new test  version  then  decoded 100%  with  Speclab  or  
without  Speclab
Speclab  was  set to the  default    11KHz  or 12KHz   sample  rate  .. 
later  I found that  if  SL  was  set to  48KHz  or 41KHz   then  there  
is  -no-  problem , so it seems  to  be  linked to the  low  sample 
rate  setting's
Q
When the  low  sample  rates  are  selected , 11 or 12 Khz , how  can SL 
affect the  sound card  sample  rate  ? ,
Jose   tells me  Ros  needs    48KHz   sample rate, Is  there  
something  I have  not  set  in SL  set  up  ?
Many thanks  , best  for  2012
Graham
G0NBD