Blindscan on linux with tbs cards

[deeptho]

Satesco,
you wrote that you said you can lock these muxes with another card. That proves that the signal is strong enough.

Here the SNR is 12.8dB on a wavefrontier t90, but last week with snow on the dish, I could still receive it with a much lower snr.
The new drivers seem to work well. No problems so far, except a small one that prevented some other cards from working,
but that has been fixed.

The reason that the specta look different is that these cards were very poorly supported by crazyscan 5 years ago and the spectra
had serious artefacts. Now they look more realistic.

 

[satesco]

Only a signal of at least 9~10dB with the 6903x/6909x cards would give me a chance to be able to lock and run the T2-Mi stream on 12606 V on both operating systems (W&L).

Now the Transalpin spot, according to the maps, does not touch my area. But I still get signal from it and can only lock it with 6902se, and sometimes 6983.

It is excluded with stid135 cards. And even at the limit of being able to lock, these cards are not able to keep it even for testing, both under Windows and Linux.
For this reason, with regret, I won't be able to do tests with these cards with your latest driver on 12606 V.

Of all the DVB cards, only with the tbs6902 can I run the T2-MI stream well on 12606 V.

You have all my esteem and appreciation for the effort to remove the bug on the stid135 cards.

 

[deeptho]

Only a signal of at least 9~10dB with the 6903x/6909x cards would give me a chance to be able to lock and run the T2-Mi stream on 12606 V on both operating systems (W&L).

Now the Transalpin spot, according to the maps, does not touch my area. But I still get signal from it and can only lock it with 6902se, and sometimes 6983.

It is excluded with stid135 cards. And even at the limit of being able to lock, these cards are not able to keep it even for testing, both under Windows and Linux.
For this reason, with regret, I won't be able to do tests with these cards with your latest driver on 12606 V.

Of all the DVB cards, only with the tbs6902 can I run the T2-MI stream well on 12606 V.

You have all my esteem and appreciation for the effort to remove the bug on the stid135 cards.

Satesco,
as you don't want to believe me I am not going to insist, but here is my result. I have used a 1m dish
on which I normally get 15dB (first picture), but I pointed it at 4.0W so that the SNR is only 7.2dB (2nd picture).
This locks perfectly with tbs6909V2. No picture breakups (3rd picture).
This is at the limit (as the color bar shows: SNR is almost in the red) but it works.

There can be other reasons for not locking (e.g interference from other satellites), but then your other card would
also not tune the mux and it does work.

The stid135 cards are quite good. The only reason to prefer other cards is to tune very low symbol rates,
but even there the card performs well in many cases. An other reason would be that
diseqc did not and perhaps still not works properly on windows, but that is a driver issue. And another
reason is that they cost a bit of money.

If 6909x does it does not work for you then the most likely reason is that you have not read the provided information well
and have not configured it configured it correctly. I have explained several times now....

6903x should work as well, but I rarely test this card.

 

[satesco]

I apologize for giving the impression that I have doubts about your tests on the 12606 V. I assure you that I didn't mean that. Feedback from users with 6903x/6909x cards and upgrades to your latest driver versions is almost entirely lacking.

I'm not blaming them for that, but I would have liked more reports from them to help understand what's going on with your application in case of problems. Such a request from me seems legitimate for each user in part, but especially for you, the one who wants to know how your latest updates work, both about neumoDVB and drivers, are up to your expectations. We (users and developers) are in the same game and want to reach the same result.

I reinstalled your latest drivers again, and this time, 6909x sees the signal on 12606 V, but also, what do you know, it scanned 12627 H with programs, which didn't happen the other days. From this, I conclude what I was saying about the signal obtained with this card.

That means that the drivers I installed are not installed correctly. I wanted to remove and reinstall them completely, but the message I received was critical and referred to a system compromise. That's why I rebuilt only media_bulid, now neumo_media_build, but the result is only partially satisfactory, i.e., the signal seen on 12606 V of 6909x and the scanned programs on 12627 H.

So, I'm convinced that the problem I'm having at 5.0W is due to incorrectly installed or compromised drivers. A further argument in this regard is the 5927 tuner driver tuner does not install at all, although I tried several times. Another argument is that the app freezes and crashes when I tune TV programs,

But even in such conditions, I don't understand why the driver finds a frequency from the bottom of the band, 12719 H, 45000,1/2, when activated on 11169,45000,1/2. This is no longer a novelty, as I have reported it on other positions in the spring. This scan error also happens with Crazyscan under Windows, which gives me the same mux at 12719 H.

And I don't think you or @CrazyCat are to blame for this error. Who knows, maybe the result of the physical design of these cards and the Linux(?) drivers.
With other dvb cards, I don't find these repeated frequencies at the top and bottom of the band.

Although I would like to know why the 6909x doesn't see the signal on French muxes (12648 V and 12731 V), for now, I blame it on the drivers installed in my system, precisely on frequencies with a stronger signal than the Italian MIS.
Thanks

 

[Llew]

@satesco
r
I have no problem with 12648V or 12731V on 5W. Or most of the other TPs. What driver are you talking about? I haven't needed to update anything for a while in neumodvb or blindscan_kernel.

I'm on Fedora 39 kernel 6.11.9, and using my Gibby 1.2 dish, which is perfectly adequate for the 5W Transalpine TPs.

 

[satesco]

@satesco
r
I have no problem with 12648V or 12731V on 5W. Or most of the other TPs. What driver are you talking about? I haven't needed to update anything for a while in neumodvb or blindscan_kernel.

I'm on Fedora 39 kernel 6.11.9, and using my Gibby 1.2 dish, which is perfectly adequate for the 5W Transalpine TPs.

...

If you don't have the 6903x/6909x cards, there is no need to update the drivers. @deeptho has improved the driver for these cards to lock the 5 T2-MI stream on the 12606 V mux, which was/is a reception problem, at least for me. For this reason, I also tried to check this mux, but due to a weak signal and other causes, probably a faulty update on my part, I cannot lock this mux, although, with different cards, there are no signal problems.

On this occasion, I discovered that neither muxes 12648 V nor 12731 V can be locked with stid135 cards. That's the problem raised, which is surprising because the signal with other cards is strong; only with these do I fail.

The rest of the problems were encountered after the driver update, which I mentioned in the previous post.
I'm working with these cards under Ubuntu 24.02 and haven't had any problems.

Thanks for your intervention, also welcome.

 

[Llew]

If you don't have the 6903x/6909x cards, there is no need to update the drivers.

l'm using a 6903x. I don't have a 6909x, so I can't compare the two.

 

[satesco]

Forgive me for not seeing you working with 6903x, which is very good. What's valid on one card is valid on the other; the only difference is that 6909x is an octa DVB-S/S2/S2X tuner.

Now that I know you have this card, try to lock stream 5 with ROOT code 16416 on 12606 V,35300, 2/3, 8PSK.
It's also an extra check that you don't have to do anything to neumoDVB. That's what @deeptho wanted, of course, me too: unfortunately, the latest driver is a problem, possibly, also a weaker signal on this mux in my case.

 

[Llew]

Now that I know you have this card, try to lock stream 5 with ROOT code 16416 on 12606 V,35300, 2/3, 8PSK.

It's PARTIAL, not OK, so doesn't appear in the Services List, so not listed in Channels.

 

[deeptho]

Forgive me for not seeing you working with 6903x, which is very good. What's valid on one card is valid on the other; the only difference is that 6909x is an octa DVB-S/S2/S2X tuner.

Now that I know you have this card, try to lock stream 5 with ROOT code 16416 on 12606 V,35300, 2/3, 8PSK.
It's also an extra check that you don't have to do anything to neumoDVB. That's what @deeptho wanted, of course, me too: unfortunately, the latest driver is a problem, possibly, also a weaker signal on this mux in my case.

Just to not confuse other readers

1. There was a problem with 12606V on 5.0W. The new drivers fix that problem. They do not cause it
2. The problem is not a problem of signal strength. The signal is many dB stronger than needed
3. It what is written on this forum or in the documentation I made. All of that text is still there an can be read, using a device called "a mouse" that you can be used to magically display old post. :->

 

[deeptho]

It's PARTIAL, not OK, so doesn't appear in the Services List, so not listed in Channels.

Your screenshot is for 12648V. ...

Below is what users will get on 12606V stream 5 after they install the new drivers, read the forum, and follow the instructions.
The others (including windows users) will get a partial lock (stream 5) or nothing at all (4). This problem occurs only on stid135 based
cards.

Note that stream 5 is a t2mi stream embedded in a transport stream in a multistream, so a lot of complex things are going on.
The first screenshot shows 12 channels from this mux. Many of them only have an info screen.
If you look closely, one of the channels has screen corruption, but that is unrelated to neumoDVB (rather a bug in some
other remote desktop software).

Also: there may be a bug that removes the pls code in some cases. I still need to fix that, but
it is unrelated.

 

[CrazyCat]

Although I would like to know why the 6909x doesn't see the signal on French muxes (12648 V and 12731 V), for now, I blame it on the drivers installed in my system, precisely on frequencies with a stronger signal than the Italian MIS.

You have some high signal in horizontal pol on this freqs. You have 6909xV1 or 6909xV2 (with VGLNA) ?

 

[satesco]

You have some high signal in horizontal pol on this freqs. You have 6909xV1 or 6909xV2 (with VGLNA) ?

Sorry, I was out, so I couldn't reply sooner.
I'm using the 6909xV1 card, v 11
As you know, 12648 V and 12731 V are on Vertical, but as you can see on the screenshot the signal is above 10DB, the 6909x card doesn't lock it.
In fact, lately, quite seldom I have been able to lock some transponders with this card. The signal has dropped quite a lot lately in my area. Probably that also explains why I can't lock them with his card(?), not to mention the Italian ones.

 

[CrazyCat]

FBC tuner/demod require SNR +0.5-1dB for stable lock. So for 8PSK,8/9 need at least 10.69dB+ (STV091x based cards) and 11dB+ for stid135

 

[tasoss]

Hello again.
Let me summarize what i want to achieve.
I want to make stid135-blindscan work as neumodvb.The reason is that neumodvb finds
some transponders that stid135-blindscan(which uses the driver's default search method) doesn't.
This is info from @deeptho.
Until now i have spent a lot of hours trying to find what changes i should make.
I want to mention that i am not a driver or kernel programmer nor a C programmer.
So an example whould be , stid135-blindscan finds 5 peaks and neumodvb finds 20.
I suppose i have to find the location where peaks are discovered and change stid135-blindscan somehow ?

Correct me if i am wrong but i think neumo-scan already uses the logic of spectrum_algo5.
Files
/linux_media/drivers/media/dvb-frontends/neumo-scan.c
/neumodvb/src/receiver/spectrum_algo5.cc
( @deeptho told me that spectrum_algo5.cc is the key ).

So. How can i start ? If it needs a lot of changes then i suppose i will have to ask
from @deeptho to create a modified stid135-blindscan version for me.If this is possible of course.
Thanks again.

 

[deeptho]

Hello again.
Let me summarize what i want to achieve.
I want to make stid135-blindscan work as neumodvb.The reason is that neumodvb finds
some transponders that stid135-blindscan(which uses the driver's default search method) doesn't.
This is info from @deeptho.
Until now i have spent a lot of hours trying to find what changes i should make.
I want to mention that i am not a driver or kernel programmer nor a C programmer.
So an example whould be , stid135-blindscan finds 5 peaks and neumodvb finds 20.
I suppose i have to find the location where peaks are discovered and change stid135-blindscan somehow ?

Correct me if i am wrong but i think neumo-scan already uses the logic of spectrum_algo5.
Files
/linux_media/drivers/media/dvb-frontends/neumo-scan.c
/neumodvb/src/receiver/spectrum_algo5.cc
( @deeptho told me that spectrum_algo5.cc is the key ).

So. How can i start ? If it needs a lot of changes then i suppose i will have to ask
from @deeptho to create a modified stid135-blindscan version for me.If this is possible of course.
Thanks again.

It seems that this is indeed already included in the kernel code. Maybe you have some other problem like not actually using the newest blindscan drivers due to some installation problem or perhaps a switch that switches too slowly.

What you should first check if the spectrum that can be output by stid135-blindscan is ok when you have the problem.

 

[tasoss]

It seems that this is indeed already included in the kernel code. Maybe you have some other problem like not actually using the newest blindscan drivers due to some installation problem or perhaps a switch that switches too slowly.

This is not the case.
I would like to ask a simple question. What is the equivalent of the following screenshot to a command line util(stid135-blindscan or neumo-blindscan) ? Maybe i am waiting for a result that it shouldn't be returned and going crazy for no reason.
You can use "--adapter=2 --rf-in=0" in the command.



Thanks!

 

[deeptho]

This is not the case.
I would like to ask a simple question. What is the equivalent of the following screenshot to a command line util(stid135-blindscan or neumo-blindscan) ? Maybe i am waiting for a result that it shouldn't be returned and going crazy for no reason.
You can use "--adapter=2 --rf-in=0" in the command.

View attachment 160662

Thanks!

Did you read the documentation?

GitHub - deeptho/blindscan: User space code for DVB blind scanning

User space code for DVB blind scanning. Contribute to deeptho/blindscan development by creating an account on GitHub.

github.com

The spectrum is automatically saved in /tmp/

You could also use neumo-blindscan instead. It will be slower, but if it works or also does not work, then it helps to find the problem.

 

[tasoss]

I don't know how neumodvb works. I might be using the command line app in a wrong way.
In any way stid135-blindscan returns 3 peaks and neumodvb whatever you see in the picture i have attached.

stid135-blindscan -cblindscan --adapter=2 --rf-in=0 --pol=1 --start-freq=12500000 --end-freq=12700000

I have attached the spectrum of this command.

I am using the latest drivers/apps etc. I don't know what else to say in order to find a solution.
I can give you access if you want to.

 

[tasoss]

@deeptho
Those are the spectrums(only H) from the two different apps.

Neumodvb spectrum


stid135-blindscan spectrum