The KiwiSDR 2 online store is open for orders! Please visit kiwisdr.nz
Spectrum & other improvements
Hi John,
Thanks for adding the peak hold function to the Spectrum display, I'm finding it to be really useful and it's much appreciated.
However on the basis that "No good deed goes unpunished" and not wishing to disappoint you............
First a bit of background.
I make a lot of use of various KiWI's (including my own) to make antenna measurements. This either consists of switching between various antennas on my own, or other KiWi's that have the antenna switching option installed, or by transmitting a signal and measuring the received level on remote KiWis whilst switching between the antenna under test and a 'reference' antenna.
Both methods rely upon being able to accurately measure the level in dB, and in many cases observe the peak level over a period of several minutes, so that any changes / enhancements / fades due to propagation can be compared 'like for like'.
So far I have tended to use the S-Meter extension (incidentally why does the S-Meter control panel say 000 1024 at the top), but using it for this purpose has a few problems. The display doesn't have 1dB scaling when you set narrow amplitude limits and also even the slowest scroll time is too short if you are trying to measure the differences between two antennas and you want to leave it running for more than a few minutes on each test.
Here's one example of a comparison between two transmit antennas with a pause in-between when I was swapping the cable over. (Yeh, I know it's not a KiWi)
You can clearly see the changes in propagation, and I have to screen grab and then draw a line to find the peak levels of each test run, then try to read off the difference in dB from the scale.
Ideally to be able to use this method more effectively, I'd like to be able to have 1dB scale resolution or better, or alternatively be able to download an amplitude / time .csv file that I could manipulate externally.
However, now that the peak hold function and other filter options are available on the spectrum screen it looks like it may be easier to use this for my tests, but a few extra's would make it better still.
I'm used to using spectrum analysers for various measurements, and these tend to have quite comprehensive marker functions that can be overlaid on to the plotted screen.
So, I was thinking - always a dangerous thing - especially for John :-)
--------------------------------------------------
Would it be possible to add a second peak hold trace in a different colour, so that along with the first peak hold trace, it could be toggled between off, peak hold and freeze.
This would allow one trace to be grabbed and frozen without further updates, then the second trace could be used to grab the next set of measurements and frozen, leaving both peak hold traces visible on the screen for comparison purposes. The GND button could be used to clear the active trace if required.
The screen cursor could then be used to measure the level of the frozen traces to my desired accuracy of 1dB.
---------------------------------------------------
With the improved functionality of the spectrum display, I now find that I am using it more than the waterfall.
However even when I remove the DX markers and F11 for a full screen display, I find that the height of the spectrum display vs waterfall is a bit on the small side.
Would it be possible to have an option to choose, or drag the boundary to set the height ratio between the two screens.
Alternatively could there be an option to toggle between Waterfall or Spectrum or both, instead of the current rather lonely / misplaced (IMHO) current 'Spec' button.
----------------------------------------------------
Would it be possible to add a 'grab to .csv option so that the values of the spectrum traces could be dumped to a file for download (or similar), so that the values of all three traces, Live Max A and Max B could be manipulated externally if required, as I currently have to manually transcribe the grabbed screen traces to a table, so that I can plot it.
----------------------------------------------------
When the spectrum Max and Min limits are set very close to each other, would it be possible to add 5dB (or perhaps 6dB to match S-Points) and 1dB graticule lines (dependent upon the Max / Min limits) to the display instead of the default 10dB lines ?
----------------------------------------------------
Would it also be possible to toggle the current 'band' marker strip that is above the 'DX' label window with a further level added to the 'shift-ctrl-Z', 'S' 'yyy' and 'x' " functions. I find that I don't really use the 'band' marker strip, but I can see it's value, and if it was possible for admins to add / edit to this I'm sure it would get more use.
e.g. adding the ITU Aero on and off route allocations and the marine bands or changing it when used with VHF / UHF down convertors etc.
----------------------------------------------------
How do folks feel about these ideas ?
I appreciate that some are much harder to implement than others, and I don't wish to burden John even more with stuff that isn't likely to be widely used.
So are they just things that are only of use to me, or are they likely to be of wider interest, or perhaps a basis for further suggestions ?
Regards,
Martin - G8JNJ
Thanks for adding the peak hold function to the Spectrum display, I'm finding it to be really useful and it's much appreciated.
However on the basis that "No good deed goes unpunished" and not wishing to disappoint you............
First a bit of background.
I make a lot of use of various KiWI's (including my own) to make antenna measurements. This either consists of switching between various antennas on my own, or other KiWi's that have the antenna switching option installed, or by transmitting a signal and measuring the received level on remote KiWis whilst switching between the antenna under test and a 'reference' antenna.
Both methods rely upon being able to accurately measure the level in dB, and in many cases observe the peak level over a period of several minutes, so that any changes / enhancements / fades due to propagation can be compared 'like for like'.
So far I have tended to use the S-Meter extension (incidentally why does the S-Meter control panel say 000 1024 at the top), but using it for this purpose has a few problems. The display doesn't have 1dB scaling when you set narrow amplitude limits and also even the slowest scroll time is too short if you are trying to measure the differences between two antennas and you want to leave it running for more than a few minutes on each test.
Here's one example of a comparison between two transmit antennas with a pause in-between when I was swapping the cable over. (Yeh, I know it's not a KiWi)
You can clearly see the changes in propagation, and I have to screen grab and then draw a line to find the peak levels of each test run, then try to read off the difference in dB from the scale.
Ideally to be able to use this method more effectively, I'd like to be able to have 1dB scale resolution or better, or alternatively be able to download an amplitude / time .csv file that I could manipulate externally.
However, now that the peak hold function and other filter options are available on the spectrum screen it looks like it may be easier to use this for my tests, but a few extra's would make it better still.
I'm used to using spectrum analysers for various measurements, and these tend to have quite comprehensive marker functions that can be overlaid on to the plotted screen.
So, I was thinking - always a dangerous thing - especially for John :-)
--------------------------------------------------
Would it be possible to add a second peak hold trace in a different colour, so that along with the first peak hold trace, it could be toggled between off, peak hold and freeze.
This would allow one trace to be grabbed and frozen without further updates, then the second trace could be used to grab the next set of measurements and frozen, leaving both peak hold traces visible on the screen for comparison purposes. The GND button could be used to clear the active trace if required.
The screen cursor could then be used to measure the level of the frozen traces to my desired accuracy of 1dB.
---------------------------------------------------
With the improved functionality of the spectrum display, I now find that I am using it more than the waterfall.
However even when I remove the DX markers and F11 for a full screen display, I find that the height of the spectrum display vs waterfall is a bit on the small side.
Would it be possible to have an option to choose, or drag the boundary to set the height ratio between the two screens.
Alternatively could there be an option to toggle between Waterfall or Spectrum or both, instead of the current rather lonely / misplaced (IMHO) current 'Spec' button.
----------------------------------------------------
Would it be possible to add a 'grab to .csv option so that the values of the spectrum traces could be dumped to a file for download (or similar), so that the values of all three traces, Live Max A and Max B could be manipulated externally if required, as I currently have to manually transcribe the grabbed screen traces to a table, so that I can plot it.
----------------------------------------------------
When the spectrum Max and Min limits are set very close to each other, would it be possible to add 5dB (or perhaps 6dB to match S-Points) and 1dB graticule lines (dependent upon the Max / Min limits) to the display instead of the default 10dB lines ?
----------------------------------------------------
Would it also be possible to toggle the current 'band' marker strip that is above the 'DX' label window with a further level added to the 'shift-ctrl-Z', 'S' 'yyy' and 'x' " functions. I find that I don't really use the 'band' marker strip, but I can see it's value, and if it was possible for admins to add / edit to this I'm sure it would get more use.
e.g. adding the ITU Aero on and off route allocations and the marine bands or changing it when used with VHF / UHF down convertors etc.
----------------------------------------------------
How do folks feel about these ideas ?
I appreciate that some are much harder to implement than others, and I don't wish to burden John even more with stuff that isn't likely to be widely used.
So are they just things that are only of use to me, or are they likely to be of wider interest, or perhaps a basis for further suggestions ?
Regards,
Martin - G8JNJ
Comments
I also don't want to burden John and am very appreciative not only of all he has done and does but also of how extremely quickly he often responds to a new idea. Having said that, I want to leave him free to be sure he is doing things he wants to do and receives return and enjoyment from. I suspect that the answer to most of your "would it be possible" questions is "yes" but I'm not certain that it would necessarily be desirable.
I'm like you in using the Kiwi spectral path basically as a spectrum analyzer more than the waterfall. I do recognize the great value of the band/waterfall has to many users but I have always almost exclusively use the spectrum display in preference to the waterfall. It is indeed sometimes useful to look at the waterfall for a time record vs. frequency, particularly when trying to understand the effect of changes to antenna and preamplifier systems in light of improving noise floor and Kiwi sensitivity, but I think that more space for a spectrum up above it and more features there a larger benefit for me in my pursuits.
I worry a little bit though when too much SA (or other) functionality is placed inside the Kiwi itself. The S-meter extension is nice but should we perhaps not be using an external tool with, say, kiwirecorder, to access the data and then process and present it as we see need? Extensions, whether internal or external, are fine but they are rather an unbounded and endless domain and resource sink so that their development perhaps should be pushed toward the particular user who wants them. I guess I'm saying that I don't want to see creeping-featurism burn John or anybody else out from creating and maintaining core capability and wider KiwiSDR acceptance and use.
Specifically in regard to spectral display, one of the dangers I see is in trying to implement full trace math and marker capabilities such as are on a 'real' spectrum analyzer at the expense of the KiwiSDR project's finite resources and risk of burn out. If the additions you mention are truly fun for anyone, that's great, but for some of the particular extensions and applications, both hardware and software, that I personally might think of or want, I hold myself accountable to not only describe them but to implement them, at least to the degree that I can. If I'm not able to I'll do my best to get someone else interested to help and I hope I don't burden John or other core contributors unnecessarily. I'm definitely going to need some help in integrating some of my own favorite potential 'extensions' such as DC-microwave frequency extension with SA and VNA functionality. I expect to have to do hardware and much firmware myself but if/when I do this I'm going to need help integrating it into the KiwiSDR.
Maybe I'm misunderstanding how open source HW/SW works best, I confess to not having much experience with it. But I do appreciate the KiwiSDR and the effort and results that it represents.
I absolutely agree with you, I also don't wish to burden John, as he has got enough on his plate.
You are probably correct that something like kiwirecorder is the way to implement some of these ideas. Maybe I should 'bite the bullet' and try to get to grips with its capabilities, which at the moment are a bit of a mystery to me.
As you may guess, I'm not a software guy, so I don't know how much or how little some of my thoughts would take to implement, or how many other people would ever make use of them should they be incorporated.
However I thought that it may be worth raising the points, as even if it is not practical or desirable to include them, they may lead to further discussion and ultimately the adoption of other ideas.
As John gradually provides the KiWi with more functionality, it opens up more possibilities regarding how it can be used, and new features such as the peak hold facility, may perhaps make it more attractive for users to choose the spectrum display rather than prefer the waterfall. Consequently it may be that with changes in user behaviour, the GUI requires occasional 'tweeks' in order to keep track with operating preferences.
Looking back at the KiWi GUI and functionality when it was first released, it is absolutely amazing how John has managed to squeeze so much out of the hardware, and has been able to implement so many of the suggestions that have been made. I'm certainly in awe of his abilities and I am personally extremely grateful of his hard work and commitment to making the KiWI such a success.
I think the KiWi has definitely been one of my best radio purchases of all time. So much so that I'm just waiting for my fourth to arrive so that I can start building it up ready for deployment on yet another site.
Regards,
Martin - G8JNJ
Translation [parental advisory: offensive language, NZ humor]
Ha yes - Well it's your party :-)
But, if you don't ask you don't get, and if you always ask you sometimes get .......
Thanks again for all your hard work and consideration.
Regards,
Martin - G8JNJ