Alberto:
Here's my reverse engineering of the auxi chunk contents for an SDR-IQ file centered at 1269 kHz. The center frequency is stored to the 1 Hz level so is really "1269000". - 1269000 = 135d08 hex - I used Riffview to look at the SDR-IQ file contents. It shows values in decimal form. Expressing the auxi bytes as [0..n], byte 32 = 8, byte 33 = 93, byte 34 = 19 and byte 35 = 0 So it looks like we have a long variable stored with the MSB in the high address. Chuck _____ From: soft_radio@yahoogroups.com [mailto:[EMAIL PROTECTED] On Behalf Of i2phd Sent: Tuesday, January 01, 2008 8:28 AM To: soft_radio@yahoogroups.com Subject: [soft_radio] Re: Winrad 1.30 --- In [EMAIL PROTECTED] <mailto:soft_radio%40yahoogroups.com> ups.com, "chutton12000" <[EMAIL PROTECTED]> wrote: > > (1) with SDR-IQ wav files, there's no way to know what frequency is > where in the display. That can be quite a mess when the recording is > full of stations on both 9 and 10 kHz plans. This could be fixed by > reading the center frequency (bytes 32, 33, and 34) (LSB to MSB > respectively) from the auxi chunk of the wav file and using it > to "grid" the display every 5 or 10 kHz. Presently I set the LO frequency when reading a WAV file only if that file was recorded with Perseus, as it uses a special chunk for that information. If you, or somebody else can provide me the exact format of the auxi chunk used by SpectraVue, I will do the same for those files. > > (2) The ECSS is quite often not usable for me when the desired > carrier is weak and close to a strong carrier. The ECSS seems to > lock onto the strong carrier as if it is still in the capture range. > I found it almost impossible to get ECSS tuned to 1089 (in LSB mode) > with a strong station on 1090, even though the 1089 station was > strong enough to produce usable audio in plain LSB mode. > Yes, the ECSS locks to the strongest signal present in its loop passband. I do already an automatic narrowing of the loop passband after lock acquisition, but apparently it is not narrow enough. But it is, as always, a matter of compromise... too narrow a loop, more easy to lose the lock, especially in conditions of strong fading... Will do some more experiments. > > Chuck > > P.S. I promised you some Matlab files for a Costas loop. I haven't > forgotten. Thanks indeed, they will be appreciated. 73 Alberto I2PHD
