суббота, 22 августа 2009 г.

SA Update 6.0.6.3 version.

SA Updateing to 6.0.6.3 version.

We have added two fields in the window of results.


  • Br - it is the last (it is importent) measured value by Measurement of Freq function in the working/operating window
  • Sc - it is the last (it is importent) measured value by Measurement of Freq function in the exponnantion in degree/reciveneing of harmonics Module.
It allows to automate a little operation with the program, or at least to make it easier.

The remembered values are used as a default settings, everywhere where it is possible. The Function of values storage is built in for a long time already, but now it is simply visualised. As for a long time, the mechanism of usage of these remembered values was described also, nevertheless, it is considered completely not superfluous to repeat, where and how these values are used.


  • Br - is used as value of the frequency of the cut LPF after phase/frequency detector (it can't be specified lesser then 30 Hrz)



  • Br - is used as value of the clock frequency of synchronization (clock of synchro) for Phase Plane module
  • Sc - is used as value of carrier frequency for Phase Plane module.
In this case, after Phase Plane module launching, it is necessary to load only an initial file, to put out n-ary and to press button Start.
The last measured value of time interval (dT) in WF module is used by default (if it possible depends of situation) for VMW unit. The order of operations can be any, it is only important to remember that measurement of all listed parametres, for their usage as by default, should be the last. That is any other new measurement will automatically re-write the previous one. It is also necessary to understand that initialization by these values occurs only by call/launching of those or other modules. In the already launched modules, these values are not vary automatically.

The ability to change all these values in regular order in any time is ramained, in limits of the usual work with the program.

Reset of all values occurs only at closing of current session of the analysis. If you close files manually, then all remembered values are saved, it is also necessary to remember. The number of minor problems, which were detected is removed and the problems are solved.

Special Thanks to everyone who have time to inform us about problems and help us to solve them.

Good luck!

среда, 19 августа 2009 г.

Signals analyzer version 6.0.6.0/6.0.6.1

SA updating to 6.0.6.1 version.

(screenshots are made in 6.0.6.0 version and it is not error, 6.0.6.1 version does not have little errors, which were detected after 6.0.6.0 version was published).

It was decided to take out results of operative measurements in the semi-transparent window in the right upper corner of the form, it will allow to increase easily quantity of derivable parametres if it is necessary. For example: FC value is added, FC - it is the central frequency between horizontal markers. This addition seems more convenient and modern to us.


On the sonogram:

  • Fu - approximate value of frequency of the upper horizontal marker.
  • Fc - central frequency between upper and lower horizontal marker.
  • Fl - approximate value of frequency of the lower horizontal marker.
  • dF - frequency band, which is limited by horizontal markers.
  • t1 - time position of the left vertical marker in the file.
  • t2 - time position of the right vertical marker in the file.
  • dT - time fragment, which is limited by vertical markers.
  • S - 1/dT time fragment which is limited by vertical markers and which is re-calculated into frequency.
Generally, names of parametres are standard, and there is no sense to repeat them. In the mode of measurements, after the phase/frequency detector the value Hz/Pixel is showing up in addition.
In the waveform browse mode, the horizontal markers do not carry any usefull work, that is why the values of horizontal markers are not outputed/displayed at all.


In VMW mode the next values are displayed/outputed:

  • dT - value of the time interval between vertical markers.
  • S - 1/dT inverse value, which is equal to frequency.
  • TSs - 1/TS inverse value of the decomposition line. This value is also equal to the frequency.
In the exponentiation module standard values are outputed/displayed. But, it is very important to understand that in this module all values, which are linked with direct frequency measurements are not
real! These measurements are wrongly interpreted often by many people. For example from the represented image, it does not follow that there are frequencies Fu = 2854 Hz, Fc = 446 Hz and Fl = -1962 Hz in the signal, these are frequencies of the complex signal, that is raised to a power in the eighth degree. There are also no obvious frequencies 2100 Hz, 1800 Hz and 1500 Hz in this signal, these are the complex frequencies enumerated/re-calculated for Phase Plane module after exponentation in the eighth degree.


Warning: be attentive, do not confuse measurements on the complex signal in this
module with the real signal.

It can seem that these values, in general, are conditional/relative and
they do not have great importance, however it not so, and sometimes it is very useful and very necessary. Just remember that these are not real frequencies, and sense of these frequencies is a bit different, than usually.

Good luck!

воскресенье, 16 августа 2009 г.

SA update to 6.0.5.8 version

SA update to 6.0.5.8 version.

In the previous version
6.0.5.6, the filter LPF operation in WF module, at various frequency measurements was recognised as unsatisfactory.


The difficulties are linked with that usage of double filtering circuits, which is strongly displaces the frequency of filter's cut, and it causes some difficulties with choosing and setup of filter's parametres.


In the most of cases, it is necessary to take frequency of the filter's cut considerably above the measured frequency of manipulation, and as result it remains vague: how much it is possible to trust to the histogram. In general, these problems are standard, because it is depends on signal's quality of how much strong it is necessary to process initial signal by the filter, after the phase/frequency detector, for reception of reliable histogram of the frequencies allocation.

But, in the previous version, this problem was expressed too obvious. We have removed the double circuit.

As a result we have:

  • the filtering is became much accurate
  • the frequency of filtrum's cut corresponds to the value, which is specified manually or selected from the list,
  • results of filtering allow to reach an excellent quality of the histograms even on the signals with the very strong and high level of noise.

In hard cases there several cycles of filtering can be demanded, however as practice shows: two cycles are quite enough. Ofcourse, it demands some skills and experience, but current implementation of LPF, in any case gives considerably best results, than former one.


This update also contains a lot of hardly noticeable, but the important improvements.

Good Luck!

среда, 5 августа 2009 г.

New: Update to version 6.0.5.6

Updating of SA to 6.0.5.6 version.

The general changes and additions:


1) Separating of the waveform into colours. The colour shows where it is possible to spend frequency
measurements and where it is impossible. Simple call of WF module does not assume any measurements, in this case the colour of the graphic is light blue, and the colour of the histogram is green.


2) WF unit call, after operation of phase detecting, is intended and allows to spend frequency measurements: in this case colour of the schedule of waveform is green, and the histograms color is yellow. It allows to be oriented easily, what functions are accessible, and what are not present. Also in this mode the value Hz/Pixel is deduced, that allows to estimate and operatively to pick up the
necessary accuracy/presicion. Basically accuracy in 1-2 hertz is enough. The necessary accuracy is provided automatically in some sence, on the narrow-band signals the accuracy is higher, than on
broadband signals, in any case, the more the size of a working window and the more scale of the graphic the better and more exact results.


3) AOF module is blocked, this module is liquidated. All functions of AOF module are completely fulfilled in Expanded WF module with much more high speed, and with much better quality and efficiency. The About window is modified, now there are direct links on appropriate resources, it does look much better.

4) The number of problems, which I was informed about is solved. The part from them were serious enough, these are various incidents within very intensive work with several signals in the program.
Thanks to everyone who informed me about that problem.

5) There is the temp folder for temporary files in the working directory of the program. This folder is needed because, as it has appeared in the course of active operation with SA - the various files of sort 123323214890.wav are accumulating, and the result size of these temporary files can reach the impressive value. One of users has informed us that for week of work he got about 200 mbytes of such temprorary garbage. It is not good. From version 6.0.5.6 this folder is automatically cleared after work with SA is finished.

Good luck!

суббота, 1 августа 2009 г.

The Famous Russian Mazielka: X02, Mazielka 2 tone

X02, Mazielka 2 tone
SelCall(?) and/or tuning(?) system for CROWD-36

The signal is here

Author: SergUA6

Band Width: ~1000 Hz
Low Range: ~300 Hz
Baud Rate: 1 Hz
Count of Carriers: 2
Step between carriers: 500 Гц
RX mode: AM

Sonograms:

pic.1 The general look and consecution after frequency detector



pic.2 Parameters of trasmitted consecution



pic.3 Manipulation frequency spacing


Mazielka is the very bright and characteristic signal. The mixed modulation is used. The signal is transferred in AM mode, and represents the frequency modulation, in which the multi-tone sequence is trasmitted with speed of manipulation in 1 Hz, in the resulted signal from five frequencies.
Such transmission in AM mode of FM signal, which represents the multi-tone sequence, in result, basically is not forbidden, and pursues the aims of the guaranteed position of the resulting signal on the frequency axis.

The problem with similar signals - is that on the records they are look very different and vary, depending of mode of their reception АМ or SSB, and thats why these signals have absolutly different appearance from different sources, and that fact brings some mess.

As it was established, the name Mazielka 2 tone has gotten from foreign observers and co-called analysts-by-ear, because in reality these two tones are not transferred in the signal.


Maria from MSM Group


суббота, 25 июля 2009 г.

Updating of SA to version 6.0.5.4: new mode and the yellow historgam.

Version 6.0.5.4: new mode, the yellow historgam and more.

Preliminary tests of possibilities of the phase detector, have seemed to us very promising, in result the improved analogue of AOF module has been developed on the basis of the phase detector. I would not tell that the developing process of this module was easy, but the main technical problems were managed to be solved, the tool has become Functional enough, flexible and it has number of possibilities, which are unavailable in old AOF module. New module also solves The primary goals of definition ofparametres of frequency characteristics of MFSK signals.


The switch "SUM" is added in WF module. Activation of that switch leads to construction(in the displaying window) of the histogram of that part of the signal, which is currently displayed on the screen.

New fields are added:
  • 1 - switch of the histogram's represantion. It is swtiched off by default
  • 2 - frequency, on which the upper horizontal marker is located
  • 3 - frequency, on which the lower horizontal marker is located
  • 4 - difference of frequencies between upper and lower marker

Precision of measurments of the frequencies is 0,01 Hrz. It is the very serious value. Although in some cases such accuracy is not required, but nevertheless, it is provided in overwhelming majority
of cases.

Work is in the histogram mode, or in the accumulation mode. Per se these modes are same.


As it was already spoken,
swtiching ON of "SUM" mode leads to construction of the yellow histogram.


The visible part of signal is used as the data for histogram's construction. It seemed to us that there is a sense do not overlap the visible part of the Signal completely by the histogram, thus we only reduced its brightness. This allows to see, by moving the sliders, which part of the signal is currently represented as histogram. In general it is necessary to notice, that all elements of the window do work in absolutely regular mode that gives extreme flexibility and efficiency.

As it is better to see once, than hundred times to hear, we have prepared some videoclips from which, You will understand the main principles of work in "SUM" mode.

Some words about accuracy of measurements, and some additional possibilities which are accessible in the new Mode.

You must realize that if You are working in the small window, or reduce by scaling buttons the histogram image to one hundred pixels by vertical, then the declared accuracy in 0.01 hertz loses any
meaning. Anyhow the pixel consist several tens or hunreds of Hertz. Thus, the declared and provided accuracy is realised only on full-scale images, which are having the greatest possible size, at least by vertical.

Accuracy of measurements depends on the signal bar and sampling rate of the file. It is also necessary to understand that the frequencies, which are above then half of sampling rate and below zero do not exist. But the scale on vertical axis allows to reduce the graphic, at an intersection of these natural boundaries, Indications of markers do not vary, and are fixed on the level of boundary conditions.

Important: The histogram mode can be switched ON at any time and on any file, but, it is possible to adequate display the frequencies only on files, which were prcocessed by the phase detector, and these files should be received directly after operation of the phase detecting in the program.

In the mode when the histogram's construction is active, but the file does not grow out of operation of the phase detector, the frequencies are not displayed and fields for these values will be empty. The histogram, in this case, represents allocation of amplitudes in loaded wav file. Sometimes it is useful, because it allows to detect files with amplitude limitation, files, which are made with 8 bit coding and files with various anomalies.

Several screenshots:

Signal CHINESE-32


Signal WSPR by K1JT - it is indeed that case, when the declared presicion of measurements is fully realized.



Attention please, it is nessesary to watch these videoclips and most of things have become clear.

This new mode, basically assumes further development, but these are future questions. It seems to us that current implementation is more than adequate and take the place of extremely not fast AOF module,
although we do not remove AOF module for some time. Probably further it will be liquidated, time will show.

The second visible update addition, I underline visible (because internal solutions are vary very strongly) addition - is the displaying of the line's frequency into that line the signal is spreading out in VMW. Sometimes it is convenient, especially for raster methods of studying of phase-manipulated signals.


Good Luck!