search

gqrx-sdr / gqrx

Software defined radio receiver powered by GNU Radio and Qt.
http://gqrx.dk
GNU General Public License v3.0
3.11k stars 545 forks source link

"Mirrored" signals #207

Closed michelinok closed 7 years ago

michelinok commented 10 years ago

Strange problem here... I'm not good in english...i hope to explain the problem.... I get mirrored signals around the central frequency You can see the problem on the attached image... Already tryed to update to the latest version of gqrx. I've tryed the fftspectrum utility of gnuradio, and the problem is not there (llok at 144.800Mhz...it does not have the corresponding mirrored signal at 146.200Mhz).

testmio

schermata4

csete commented 10 years ago

We have just discussed this issue on the mailing list: https://groups.google.com/d/topic/gqrx/GkKHL2Gvtf8/discussion I assume you are the same person because you are posting the same screenshots?

I have confirmed that it is a bug somewhere and we have agreed that this is the same as issue #99, so why post again?

It is also not nice to start a thread on reddit without telling people there what you already know from the mailing list.

michelinok commented 10 years ago

Sorry, didn't see your reply to #99 of 6 days ago. Yes, i've opened a thread on reddit to ask for help and see if someone could help (so i could help here also if i'm not a linux developer.) Really really sorry....didn't meant to create such "heartquake" (and sorry again for my english!). I really would investigate on the problem, that's why i'm asking on reddit. The problem of #99 is not exactly my problem... I've tryed toggling everything...but the problem persists, so i don't think it's the same problem. "Toggling" everything should "solve"...but that's not my case :(

csete commented 10 years ago

I am quite sure it is the same issue even if you don't see any difference when you toggle the controls. When you toggle something in gqrx the flow graph will reset and that is when the issue appears if it is not already there. Toggling again may remove it but it is not a solution, it's just a clue where the problem may occur. In the osmocom-fft application there is no reset of the flow graph and that is why you don't see the issue there.

It's very good that you try getting help other places, including reddit. My point was that you should give people all the info you have, otherwise they will just be guessing blindly.

pavel-i-am commented 8 years ago

I thought I was going crazy or my new board was faulty. At least now that's settled. Bug is still relevant if anyone cares, @csete . If there's any info I can help to provide, let me know. My current setup is a Mac with El Capitan and Ettus Research b205 mini. Precompiled dmg doesn't work with the board, by the way, so I had to compile or install through macports.

csete commented 8 years ago

I never said this bug was irrelevant, just that I think it is a duplicate of another. But if you guys prefer to have two bugs open, fine with me - it's not going to be fixed twice as fast ;-)

pavel-i-am commented 8 years ago

@csete , I'm not complaining about statuses at all. Just letting you know that this thing might still be relevant and asking if I could be of help :)

p.s. there's a nice module gr-dsd and a fork with it embedded as a filter. Can you look at that once you have some spare time? I watched some youtube examples and it seems like a good thing to have. Cheers. (https://github.com/kantooon/gqrx_fork/tree/digital)

csete commented 7 years ago

Duplicate of issue #99

andimik commented 7 years ago

Same here. Sdrsharp has no mirrors, gqrx has indeed. But not always and not at all frequencies.

csete commented 7 years ago

@andimik Please see issue #99. It has been documented for a long time now that this is a problem in gr-osmosdr and there is a workaround. A patched version of gr-osmosdr is available - again please see issue #99. Give feedback there instead of poking this closed issue.

andimik commented 7 years ago

Google found that for me as this already known issue is NOT described in your readme. It only says one should see the issues.

There are lots of references in #99 but it's not clear for me why I should patch a 3 year old issue by myself.

So why not start a pull request for your patch?

Sorry for the questions, but I am just a user, not a programmer...

csete commented 7 years ago

I was just pointing out that a fix is known and you can apply it if you wish. If you build everything from source, then it makes sense to apply the patch yourself. If you are using binaries, you will have to wait until the patch is available in a binary release.

The patch comes from the gr-osmosdr mailing list. I do not know why it wasn't applied upstream, so the best I can do for now is to make it available in the binaries that we distribute.