[digitalradio] Re: Comparison of RTTY software sensitivity
> By synchronous detection, Vojtech, do you mean treating the first start bit > as the beginning of a synchronous multi-character sequence, thereby > providing some protection against "broken" start and stop bits within that > sequence? Brian K6STI referred to his decoding technique as employing a > "flywheel", which I interpreted as a means of adjusting the synchronous > timing with high-quality start bits decoded within the sequence. Dave, what I mean is to consider all edge into sync recovery. Most software I know is using pretty stable clock and fills the spaces with idle characters. The decoder needs to know how long the stop bit is, which may be estimated on the air or configured by the user. I suppose 1.5 bit length is the most common? Then you may try to search for raising/falling edge at 2x bit speed and slowly adjust the sampling point. Yes, it is some kind of "flywheel", that any synchronous decoder like PSK31 uses. 73, Vojtech OK1IAK
AW: [digitalradio] Re: Comparison of RTTY software sensitivity
Comments in the text below .. _ Von: digitalradio@yahoogroups.com [mailto:digitalra...@yahoogroups.com] Im Auftrag von Dave AA6YQ Gesendet: Freitag, 22. Januar 2010 23:06 An: digitalradio@yahoogroups.com Betreff: RE: [digitalradio] Re: Comparison of RTTY software sensitivity MMTTY provides a choice of three different RTTY decoders, with the ability to shape the filters for each. There is also an optional bandpass filter and an optional notch filter, with user control of shape for each. As a first step in improving MMTTYs RTTY decoding performance, I am determining how to optimize performance given the capabilities Mako-san JE3HHT has already provided, using a setup similar to what Alex VE3NEA and Wes WZ7I have used. Note that on the chart Wes posted, WinWarbler running the HyperSensitive profile with both the bandpass and notch filters enabled yields sensitivity close to that of TrueTTY. WinWarbler uses MMTTY as its RTTY engine, so this performance is possible with MMTTY alone. I don´t have winwarbler but have mmtty . Can somebody make a text how to setup mmtty for the hyper sensitive profile that I can store it I mmtty as user profile MANY THANKS DG9BFC SIGI By synchronous detection, Vojtech, do you mean treating the first start bit as the beginning of a synchronous multi-character sequence, thereby providing some protection against broken start and stop bits within that sequence? Brian K6STI referred to his decoding technique as employing a flywheel, which I interpreted as a means of adjusting the synchronous timing with high-quality start bits decoded within the sequence. 73, Dave, AA6YQ From: digitalradio@yahoogroups.com [mailto:digitalra...@yahoogroups.com] On Behalf Of Vojtech Sent: Friday, January 22, 2010 8:27 AM To: digitalradio@yahoogroups.com Subject: [digitalradio] Re: Comparison of RTTY software sensitivity Here is another, similar chart: http://www.dxatlas. <http://www.dxatlas.com/RttyCompare/> com/RttyCompare/ I was comparing MMTTY with MultiPSK and gMFSK against RTTY in white noise. Interesting observation was that MMTTY was better than MultiPSK at better than marginal SNR, but MultiPSK was slightly better than MMTTY at very low SNR. My best bet is that MMTTY is doing some kind of signal processing after detector, which fixes some errors, but makes things worse in very low SNR. Both yours and Alex's graphs show superiority of TrueRTTY and MixW. I wonder whether TrueRTTY is doing synchronous detection. This is what I plan to try when I retire, hi. 73, Vojtech OK1IAK No virus found in this incoming message. Checked by AVG - www.avg.com Version: 9.0.730 / Virus Database: 271.1.1/2636 - Release Date: 01/21/10 15:34:00
RE: [digitalradio] Re: Comparison of RTTY software sensitivity
MMTTY provides a choice of three different RTTY decoders, with the ability to shape the filters for each. There is also an optional bandpass filter and an optional notch filter, with user control of shape for each. As a first step in improving MMTTY's RTTY decoding performance, I am determining how to optimize performance given the capabilities Mako-san JE3HHT has already provided, using a setup similar to what Alex VE3NEA and Wes WZ7I have used. Note that on the chart Wes posted, WinWarbler running the HyperSensitive profile with both the bandpass and notch filters enabled yields sensitivity close to that of TrueTTY. WinWarbler uses MMTTY as its RTTY engine, so this performance is possible with MMTTY alone. By synchronous detection, Vojtech, do you mean treating the first start bit as the beginning of a synchronous multi-character sequence, thereby providing some protection against "broken" start and stop bits within that sequence? Brian K6STI referred to his decoding technique as employing a "flywheel", which I interpreted as a means of adjusting the synchronous timing with high-quality start bits decoded within the sequence. 73, Dave, AA6YQ From: digitalradio@yahoogroups.com [mailto:digitalra...@yahoogroups.com] On Behalf Of Vojtech Sent: Friday, January 22, 2010 8:27 AM To: digitalradio@yahoogroups.com Subject: [digitalradio] Re: Comparison of RTTY software sensitivity Here is another, similar chart: http://www.dxatlas.com/RttyCompare/ I was comparing MMTTY with MultiPSK and gMFSK against RTTY in white noise. Interesting observation was that MMTTY was better than MultiPSK at better than marginal SNR, but MultiPSK was slightly better than MMTTY at very low SNR. My best bet is that MMTTY is doing some kind of signal processing after detector, which fixes some errors, but makes things worse in very low SNR. Both yours and Alex's graphs show superiority of TrueRTTY and MixW. I wonder whether TrueRTTY is doing synchronous detection. This is what I plan to try when I retire, hi. 73, Vojtech OK1IAK No virus found in this incoming message. Checked by AVG - www.avg.com Version: 9.0.730 / Virus Database: 271.1.1/2636 - Release Date: 01/21/10 15:34:00
[digitalradio] Re: Comparison of RTTY software sensitivity
Now I don't know about TrueTTY, but MMTTY has a number of things you can play with in terms of the filtering and detection, so I wonder if we could get some comparisons of those - or at least tell us what the setup was when it was tested. Jim W6JVE
[digitalradio] Re: Comparison of RTTY software sensitivity
--- In digitalradio@yahoogroups.com, "Vojtech" wrote: > > Both yours and Alex's graphs show superiority of TrueRTTY and MixW. I wonder > whether TrueRTTY is doing synchronous detection. This is what I plan to try > when I retire, hi. > There used to be the K6STI RITTY program which does synchronous detection. I wish we had a comparison of it with some of the modern RTTY software. Trouble is, RITTY requires a DOS environment and a SoundBlaster ISA sound card, and you don't find those much anymore. And it's no longer on the market. I was really excited when it came out, because I had wanted to do synchronous detection for a long time. However it didn't appear to be all that helpful, maybe a db or two. Then PSK31 for the sound card modem came out and most of the rag chewers switched to that from RTTY. Jim W6JVE
Re: [digitalradio] Re: Comparison of RTTY software sensitivity
> > > Both yours and Alex's graphs show superiority of TrueRTTY and MixW. I > wonder whether TrueRTTY is doing synchronous detection. This is what I plan > to try when I retire, hi. > > 73, Vojtech OK1IAK > TrueTTY also gave results better than any other package tested for MFSK16. Wes, WZ7I
[digitalradio] Re: Comparison of RTTY software sensitivity
Here is another, similar chart: http://www.dxatlas.com/RttyCompare/ I was comparing MMTTY with MultiPSK and gMFSK against RTTY in white noise. Interesting observation was that MMTTY was better than MultiPSK at better than marginal SNR, but MultiPSK was slightly better than MMTTY at very low SNR. My best bet is that MMTTY is doing some kind of signal processing after detector, which fixes some errors, but makes things worse in very low SNR. Both yours and Alex's graphs show superiority of TrueRTTY and MixW. I wonder whether TrueRTTY is doing synchronous detection. This is what I plan to try when I retire, hi. 73, Vojtech OK1IAK
[digitalradio] Re: Comparison of RTTY software sensitivity
Thank you for your tests and report Wes. The data is very enlightening. fldigi's detector is a simple pre-filter / frequency discriminator with hysterisis. I will build and test alternate detectors and would welcome the assistance of both designers and testers to improve the RTTY decoder in fldigi. 73, Dave, W1HKJ
