To: rsgb_lf_group@blacksheep.org
References: <6DB8451D7F3D3947A5918808A59621EA14E20EF0@servigilant.vigilant.local>
From: John Andrews <w1tag@charter.net>
Message-ID: <da0c8b2b-800c-21eb-9627-ee18557b360a@charter.net>
Date: Thu, 22 Nov 2018 18:51:09 -0500
User-Agent: Mozilla/5.0 (Windows NT 6.1; WOW64; rv:60.0) Gecko/20100101
 Thunderbird/60.3.1
MIME-Version: 1.0
In-Reply-To: <6DB8451D7F3D3947A5918808A59621EA14E20EF0@servigilant.vigilant.local>
Content-Language: en-US
Subject: Re: LF: WSPR-15 in WSJT-X
Content-Type: text/plain; charset=windows-1252; format=flowed
Content-Transfer-Encoding: 8bit
Sender: owner-rsgb_lf_group@blacksheep.org
Precedence: bulk
Reply-To: rsgb_lf_group@blacksheep.org

Joe has raised those questions before, and I've assumed that he has not 
been happy with the responses, including mine. Maybe a greater number of 
opinions would be of help.

Regarding #2, the averaging or summing of results worked very well in 
WOLF. Of course, since the message length was 96 seconds, you could 
collect quite a few of them before running out of patience (or FETS). 
Since Joe is hinting at shorter message times in his proposed new mode, 
don't reject the summing aspect until we know the parameters.

John, W1TAG

On 11/22/2018 4:49 PM, VIGILANT Luis Fernández wrote:
> Hi LF
> 
> Regarding this issue there is an answer from Joe, K1JT
> Original message follows
> 
> 73 de Luis
> EA5DOM
> 
>  1.
>     ###########################################################################
>  2.
>  3.
>     I have received similar requests from a few others. We should probably
>  4.
>     address this perceived need before too long. I would like to retire
>  5.
>     WSPR-X, anyway, and do further development within WSJT-X.
>  6.
>  7.
>     I am not persuaded that WSPR-15 is really the best way to go. Here are
>  8.
>     some potentially important questions:
>  9.
> 10.
>     1. Is it clear that in practice WSPR-15 provides LF/MF decodes at lower
> 11.
>     S/N than WSPR-2? If so, ho much lower?
> 12.
> 13.
>     2. Could an equivalent gain in performance be achieved by having the
> 14.
>     decoder average several consecutive, properly synchronized WSPR-2
> 15.
>     transmissions?
> 16.
> 17.
>     3. If a more sensitive WSPR-like mode is truly needed for LF/MF
> 18.
>     experimentation, would it be better to create something that for now
> 19.
>     I'll call "WSPR-MSK", which (like MSK144) uses OQPSK (Offset Quadrature
> 20.
>     Phase-Shift Keying), a constant-envelope waveform, coherent
> 21.
>     demodulation, and an LDPC code? Steve (K9AN) and I have discussed such
> 22.
>     a possible mode, and we might be more motivated to develop that rather
> 23.
>     than going "back" to WSPR-15. I suspect WSPR-MSK could be made as
> 24.
>     sensitive (or better) than WSPR-15, even with transmissions shorter than
> 25.
>     15 minutes.
> 26.
> 27.
>     -- 73, Joe, K1JT
>