Hello Tony,
>The sound card was already calibrated, but the offsets were not locked. >Locking them seems to have made a difference. I repeated the Yes the locking of the sampling frequency, at least, must be done to compare. Note: I did this complex automatic shift sampling determination because MT63 has a problem which is inherent to the exagerated number of carriers (64). The more you have carriers the more the calibration must be precise. MT63 was done for almost perfect sound cards (Sound Blaster or equivalent). For the on-board sound cards, there is a problem due to the poor sampling accuracy of these cards, so you can't be sure that the other Ham has a well-calibrated sound card. It's the reason why I propose an automatic sampling autodetermination, by default, working only with good signals (as you saw in the previous tests). I think, aside to the poor average power/peak power ratio (about 0.1), this problem of calibration is an issue which limits the use of MT63 (Hams, in general, use their on-board sound card installed in their PC). >have. Minimum SNR (errorr-free throughput) is -8db for IZ8BLY and -6db for >Mutlipsk. -6dB for Multipsk seems correct for 10 bauds (I measure -5dB for 10 bauds). I agree that Nino's program has a better decoding program. TKS for all your interesting tests! 73 Patrick ----- Original Message ----- From: Tony To: digitalradio@yahoogroups.com Sent: Saturday, September 27, 2008 10:32 PM Subject: Re: [digitalradio] MT63 -- Mutlipsk, IZ8BLY, MixW Patrick, The sound card was already calibrated, but the offsets were not locked. Locking them seems to have made a difference. I repeated the same tests and decoding did improve with Multipsk (see below). As for one program having an advantage over another, I guess it could be due to the slight sensitivity advantage Nino's program seems to have. Minimum SNR (errorr-free throughput) is -8db for IZ8BLY and -6db for Mutlipsk. This seems to show-up in test #1 where the selective fading may be causing the signal to dip below the -6db decode threshold for Multipsk. Nino's software has a 2db "buffer" with this test. In test #3 where there's selective fading and lightning static, things improve with Multipsk when the SNR is inceased to +3db from -3db: Multipsk Test #3 SNR +3db THE QUIK BROW FOX JUMPS OVER THE LAZY DOG THE QUICK BROWN FOX JUMPS OVER THE LAZY DOG THE QUICK BRWN FOX JUMPS OVER THE LAZY DOG THE QUICK BROWN FOX JUMPS OVER THE LAZY DOG SNR -3db HE QUIK3,ROWS F'XexMn OVER THE L/ZY DOG THE QUICK BROWN FOX JUMPS OVER THE LAZY D+G THE QUICK BRWg FOX JMPS OVER THE LAZY DOG THE QUICK BROWN FOX JUMPS OVER THE LAZY DOG Thanks for all Patrick. Tony, K2MO ++++++++++++++++++++++++++++++++++++++++++++++++++ Test #1: SNR -6db selective fading IZ8BLY TH; QUICK BAOWN FOX JUOPS OVrR THE LAZY [OG THE QUICK BROWN FOX JUMPS OVER THE LAZY DOG THE QUICK BROWN FOX JUMPS OVER THE LAZY DOG THE QUICK BROWNFOX JUMPS OVER THE LAZY DOG MixW 2Hx QU/8qvB,4YNn}mGbSI*+OenJX|Y aO| TH3N$aaK RptNWFOX JUMPS OVER THE LAZY DOG THE QUICK BROWN FOX JUMPS OVER THE LAZY DOG THE QUIcK BROWN FOXFJUmPS "VER THE LAZY DOG Multipsk THN QUICKFB;OW ZX JP OVXR THErL,ZY RG THE [UICK BROW FOX JUMPS OVER THE LAZY DOG THE QUICK BROWN FOX JUMPS OVER THE LAZY DOG THE QUICK BROWN FOX JUMPS OVER THE LAZY DOG ________________________________________________ Test #2: SNR -3db selective fading IZ8BLY THE QUICK BROWN FOX JUOPS OVER THE LAZY DOG THE QUICK BROWN FOX JUMPS OVER THE LAZY DOG THE QUICK BROWN FOX JUMPS OVER THE LAZY DOG THE QUICK BROWN FOX JUMPS OVER THE LAZY DOG MixW THE AUICKROWN F:X JUMPS ,VER THE L/ZY [OG THE QUICK BROWN FOX JUMPS OVER THE LAZY DOG THE QUICK BROWN FOX JUMPS OVER THE LAZY DOG THE QUICK BROWN FOX JUMPS OVER THE LAZY DOG Multipsk THE QUICK BROWN FOX JUMPS OVER THE LAZY DOG THE QUICK BROWN FOX JUMPS OVER THE LAZY DOG THE QUICK BROWN FOX JUMPS OVER THE LAZY DOG THE QUICK BROWN FOX JUMPS OVER THE LAZY DOG _______________________________________________ Test #3: SNR -3db selective fading / lightning static IZ8BLY THE QUCIK BROWq FOX JUMPS OVrR THE LAZY DOG THE QUICK BROWN FOX JUMPS OVER THE LAZY DOG THE QUICK BROWN FOX JUMPS OVER THE LAZY DOG THE QUICK BROWN FOX JUMPS OVER THE LAZY DOG MixW THH mUICK0I\OWm iZM JUMPS ,VER THE 2AZYzDOG THE QUICK BROWN FOX xUMPS OVER THE LAZY DOG THE QUICK BRO9N FOX JUMPS OVER THE LAZY DOG THE QUICK BROWN FOX JUMPS OVER THE LAZY DOG Multipsk HE QUIK3,ROWS F'XexMn OVER THE L/ZY DOG THE QUICK BROWN FOX JUMPS OVER THE LAZY D+G THE QUICK BRWg FOX JMPS OVER THE LAZY DOG THE QUICK BROWN FOX JUMPS OVER THE LAZY DOG ___________________________________________________ Test #4 SNR -9db IZ8BLY TH QUICK BROWN FOX_UMPS OVER THE LAZY DOG THE QUICK BROWN FOX JUMPS OVER THE LAZY DOG THE QUICK BROWN FOX JUMPS OVER THE LAZY DOG THE QUICK BROWN FOX JUMPS OVER THE LAZY DOG MixW UHE qU4C? BROWN 9GX'JUMPS OVE THE LAZY DOG THE QUI^d BnCWNlFOX JUMPS OVERDTHE LAZY OGj HEs%#IC BRWN FOX JUMS OV5R THE LAZYBDOG THE QrCK BROWN FBX UXPS OVER {HE"LAZpHDOG Multipsk TH\ QUIF,RO\[ F(U'xM S OV:R THE LAZY DOG THE QUICKBRON FOX JUMPS OVER THE LAZY DOG THE QUICK BRKWN FO { MPS OVER THE LAZY DOG THE QUICK BROWN FOX JUMPS OVER THE LAZY DOG ----- Original Message ----- From: "Patrick Lindecker" <[EMAIL PROTECTED]> To: <digitalradio@yahoogroups.com> Sent: Friday, September 26, 2008 7:04 PM Subject: Re: [digitalradio] MT63 -- Mutlipsk, IZ8BLY, MixW Hello Tony, About Multipsk, could you first (if not done) the MT63 sampling calibration by clicking on "Sampling freq." and start the automatic calibration. Note: in Multipsk there is a double correction: * determination of the sampling frequency offset between the TX and the RX sound cards. This function does not belong to the original soft nor Mixw, * determination of the frequency offset. If you do a comparizon, click on the first "lock" to 0 (to avoid the determination of the sampling frequency offset, function which needs a good MT63 signal). However, the original soft could be much better (I wrote an independant MT63 decoding code and I don't know the way the original soft decodes). 73 Patrick ----- Original Message ----- From: Tony To: digitalradio@yahoogroups.com Sent: Saturday, September 27, 2008 12:35 AM Subject: [digitalradio] MT63 -- Mutlipsk, IZ8BLY, MixW All, There seems to be some difference between Mutlipsk, MixW and IZ8BLY's software when it comes to decoding MT63. I ran all three programs simultaneously with a path simulator in line and the SNR set above the decode threshold. The results were the same whether I ran the programs one at a time or simultaneously. Not sure why this is; would appreciate it if someone can shed some light on this. I plan to run each program on-the-air to see if the results change. Tony, K2MO Test #1: SNR -6db selective fading Test #2: SNR -3db selective fading Test #3: SNR -3db selective fading / lightning static Test #1 IZ8BLY MT63 Terminal TH QUICK BRAOWN FOX JUOPS OVER THE LAZY [OG THE QUICK BROWN FOX JUMPSOVER THE LAZY DOG THE QUICK BROWN FOX JUMPS OVER THE LAZY DOG THE QUICK BROWNFOX JUMPS OVER THE LAZY DOG MixW TH QUI4 M)9"Nd3/ANjV (UT$ExLZY UG THE QUICK BRmWN 1OX JUMPS OVER THE LAZY DOG THE QUICK BROWN FOX JUMPS OVER THE LAZY DOG THE QUI>K BROWN F9XFJUmPS "VER THE LAZY DOG Multipsk YhlU+ O~^ FlX JUiPSkOVER THE LAZY DOG THE QUICK{BOovFV!D.M]y5xVr| +v75UX7ØOpL{ z]p?Bx~i6GzV$cSNVDcTH: LØZY DOG]mOT* _________________________________________________ Test #2 MT63 Terminal THE QUICK BROWN FOX JUMPS OVER THE LAZY DOG THE QUICK BROWN FOX JUMPS OVER THE LAZY DOG THE QUICK BROWN FOX JUMPS OVER THE LAZY DOG THE QUICK BROWN FOX JUMPS OVER THE LAZY DOG MixW THF QUIaK BRORNdFOX J$MPS ?[7@ TWE LAZY DOG THE QUICK BROWN FOX JUMPS OVER THE LAZY DOG THE QUICK BROWN FOX JUMPS OVER THE LAZY DOG THE QUICK BROWN FOX JUMPS OVER THE LAZY DOG Multipsk THE QUICK BROWN FOX JM8; O5Es HE LAZY DG THe*QICK BROWN F+X JUMPS OVER THE LAZY DOG THE QUICK BROWN FOX JUMPS OVER THE LAZY DOG THE QUICK BROWN FOX JUMPS OVER THE LAZY DOG ______________________________________________________ Test #3 IZ8BLY MT63 TERMINAL RON FOX JUMPS OVER THE LAZY DOG THE QUICK BROWN FOX JUMPS OVER THE LAZY DOG THE QUICK BROWN FOX JUMPS OVER THE LAZY DOG THE QUICK BROWN FOX JUMPS OVER THE LAZY DOG MIXW HE QUIAn BROWtF?X:J+MPS v)rR TE L/ZY DOG THE QUICK BRKWN FOX JUMPS OVER THE LAZY DOG THE QUICK BRO9 FOX JUMPS OVER THE LAZY DOG THE QUICK BROWN FOX JUMPS OVER THE LAZY DOG MULTIPSK $Xiu(L Kka?SO?m)#-ap^jtOu{+^d4,-`Ø$P{&&8YAZY O) ,njl'KuBRja "_J(/S VØRU!H5N(1=YDOG THE QUIkK B}O9g OY4JUMPS OVER THE LAZY DOG THE QUICK BROWN FOX JUMPS OVER THE LAZYØDOG