TechniSat SkyStar HD: Problems scaning and zaping

Hello,

I have a Skystar HD. I use multiproto from here (checked it out yesterday):
http://jusst.de/hg/multiproto
It compiles fine, all needed modules are loaded automatically upon boot.
This is my dmesg:
saa7146: found saa7146 @ mem f88f2000 (revision 1, irq 21) (0x13c2,0x1019). saa7146 (0): dma buffer size 192512
DVB: registering new adapter (TT-Budget S2-3200 PCI)
adapter has MAC addr = 00:d0:5c:67:dd:91
input: Budget-CI dvb ir receiver saa7146 (0) as /class/input/input3 stb0899_write_regs [0xf1b6]: 02
stb0899_write_regs [0xf1c2]: 00
stb0899_write_regs [0xf1c3]: 00
_stb0899_read_reg: Reg=[0xf000], data=82
stb0899_get_dev_id: ID reg=[0x82]
stb0899_get_dev_id: Device ID=[8], Release=[2]
_stb0899_read_s2reg Device=[0xf3fc], Base address=[0x00000400], Offset=[0xf334], Data=[0x444d4431] _stb0899_read_s2reg Device=[0xf3fc], Base address=[0x00000400], Offset=[0xf33c], Data=[0x00000001]
stb0899_get_dev_id: Demodulator Core ID=[DMD1], Version=[1]
_stb0899_read_s2reg Device=[0xfafc], Base address=[0x00000800], Offset=[0xfa2c], Data=[0x46454331] _stb0899_read_s2reg Device=[0xfafc], Base address=[0x00000800], Offset=[0xfa34], Data=[0x00000001]
stb0899_get_dev_id: FEC Core ID=[FEC1], Version=[1]
stb0899_attach: Attaching STB0899
stb6100_attach: Attaching STB6100
DVB: registering frontend 0 (STB0899 Multistandard)...

I use scan from here:
http://jusst.de/manu/scan.tar.bz2
It does not compile, I use the binary I found in the archive.

I use szap2 from here:
http://www.free-x.de/vdr/patches/hdtv/szap2.tgz
It does compile.

My kernel is:
2.6.22-gentoo-r9

My problems are:
- When I scan the same transponder more than once, sometimes I get some services, sometimes I get no services.
- When I zap (with szap2) to a station more than once, sometimes I get a lock, sometimes I get no lock.
- I have also tried with vdr 1.4.7(old DVB API) and 1.5.14 (new dvb API). Both versions can tune to only one channel, the "Active" channel from previous vdr shutdown. When you change channel the signal is lost. I suppose it relates to the zap problem.

I would appreciate your help / suggestions.

Thank you very much.
Vagelis

_______________________________________________
linux-dvb mailing list
[email protected]
http://www.linuxtv.org/cgi-bin/mailman/listinfo/linux-dvb

.

Quoted 2 lines

My problems are:
- When I scan the same transponder more than once, sometimes I get some services, sometimes I get no services.

Quoted 1 line

- When I zap (with szap2) to a station more than once, sometimes I get a lock, sometimes I get no lock.

Quoted 1 line

- I have also tried with vdr 1.4.7(old DVB API) and 1.5.14 (new dvb API). Both versions can tune to only one channel, the "Active" channel from previous vdr shutdown. When you change channel the signal is lost. I suppose it relates to the zap problem.

Quoted 5 lines

Quoted 1 line

Quoted 1 line

- When I zap (with szap2) to a station more than once, sometimes I get a lock, sometimes I get no lock.

I have the same problem. I was trying to tune to ProSieben 100 times today. And I got only 11 successful locks. The signal strengt should be ok, as I have no problems to tune this
channel with other DVB-S cards...

no anymore Prosiben (hd version) since last week :(
http://www.lyngsat.com/hd/astra19.html

Hi,

Quoted 2 lines · Vangelis Nonas

Quoted 2 lines

My problems are:
- When I scan the same transponder more than once, sometimes I get some services, sometimes I get no services.

I have never tried to scan with the SkyStar HD.

Quoted 1 line

Quoted 1 line

- When I zap (with szap2) to a station more than once, sometimes I get a lock, sometimes I get no lock.

I have the same problem. I was trying to tune to ProSieben 100 times today. And I got only 11 successful locks. The signal strengt should be ok, as I have no problems to tune this
channel with other DVB-S cards...

Here is a little script to do such tests:
http://www.makhutov.org/downloads/dvb/tune.sh

You need this szap (http://abraham.manu.googlepages.com/szap.c) and this patch (http://www.makhutov.org/downloads/dvb/szap.patch) to make the script work.

Has anyone any ideas why the tuning failes so often? Is it a problem of
szap or the driver itself? I heard, that this card works perfect with VDR for other users...

[...]
Can you guys please update from the multiproto tree and test again at the earliest and give me your feedback ?

Hi,

Quoted 3 lines · Manu Abraham

[...]
Can you guys please update from the multiproto tree and test again at the earliest and give me your feedback ?

No, still the same:

Try: 50
Failes: 37
Tunes: 13

Maybe the problem is in the szap.c and not in the driver, as Reinhard Nissl had no problems while tuning channels with VDR?

An other thing that I noticed is that only first tunes are successful. I continued the same tuning test to 100, but I got no more successfull locks. The last successfull lock was try 20:

Try: 100
Failes: 87
Tunes: 13

I am running a new test with a 15 seconds break after each tune now and will mail you the results when the test finises.

Quoted 25 lines · Artem Makhutov

Hi,

Quoted 3 lines · Manu Abraham

[...]
Can you guys please update from the multiproto tree and test again at the earliest and give me your feedback ?

No, still the same:

Try: 50
Failes: 37
Tunes: 13

Maybe the problem is in the szap.c and not in the driver, as Reinhard Nissl had no problems while tuning channels with VDR?

An other thing that I noticed is that only first tunes are successful. I continued the same tuning test to 100, but I got no more successfull locks.
The last successfull lock was try 20:

Try: 100
Failes: 87
Tunes: 13

Quoted 2 lines

I am running a new test with a 15 seconds break after each tune now and will mail you the results when the test finises.

Are you sure that you got the top level 2 changes changeset 7204 and 7203 respectively ?

Quoted 2 lines · Manu Abraham

Are you sure that you got the top level 2 changes changeset 7204 and 7203 respectively ?

Oh, I only got 7203. Will try with 7204 in a few minutes.

Hi,

Quoted 4 lines · Artem Makhutov

Quoted 2 lines · Manu Abraham

Are you sure that you got the top level 2 changes changeset 7204 and 7203 respectively ?

Oh, I only got 7203. Will try with 7204 in a few minutes.

Awesome! It fixed the problem:

Try: 100
Failes: 0
Tunes: 100

Great job!

Quoted 6 lines · Artem Makhutov

Hi,

Quoted 3 lines · Artem Makhutov

Quoted 2 lines · Manu Abraham

Are you sure that you got the top level 2 changes changeset 7204 and 7203 respectively ?

Quoted 8 lines

Quoted 1 line

Oh, I only got 7203. Will try with 7204 in a few minutes.

Awesome! It fixed the problem:

Try: 100
Failes: 0
Tunes: 100

Great job!

Quoted 13 lines · Manu Abraham

Quoted 11 lines · Artem Makhutov

Hi,

Quoted 7 lines · Artem Makhutov

Quoted 3 lines · Manu Abraham

Are you sure that you got the top level 2 changes changeset 7204 and 7203 respectively ?

Oh, I only got 7203. Will try with 7204 in a few minutes.

Quoted 8 lines

Quoted 8 lines

Awesome! It fixed the problem:

Try: 100
Failes: 0
Tunes: 100

Great job!

Also, can you please do a benchmark in lock timings between changeset 7205 and 7200 ?

The timing can be looked at by enabling the time stamps in the kernel config and looking at timestamps in the logs for start - stop (FE_HAS_LOCK) between the 2 changesets.

Regards,
Manu

_______________________________________________
linux-dvb mailing list
[email protected]
http://www.linuxtv.org/cgi-bin/mailman/listinfo/linux-dvb

Thank you all for your help,

With the latest changeset I can tune channels correctly. I tried vdr and szap (from Manu).

I dont know how to undo the last changesets and go back to 7200. Please tell me how to do it and I'll get the timings.

I also tried to scan Hotbird and it seems that scanning is now consistent when you scan the same transponder more than once.

A full scan produced 1722 channels which is not bad, but it needs improvement. There are about 2200 channels I think.

I attach the log of the scan. It fails to scan certain transponders and I believe (not 100% sure, but I did some testing) that this is consistent across runs.

Regards
Vagelis

Quoted 44 lines · Manu Abraham

Quoted 14 lines · Manu Abraham

Quoted 12 lines · Artem Makhutov

Hi,

Quoted 8 lines · Artem Makhutov

Quoted 4 lines · Manu Abraham

Are you sure that you got the top level 2 changes changeset 7204 and 7203
respectively ?

Oh, I only got 7203. Will try with 7204 in a few minutes.

Quoted 8 lines

Quoted 8 lines

Awesome! It fixed the problem:

Try: 100
Failes: 0
Tunes: 100

Great job!

Also, can you please do a benchmark in lock timings between changeset 7205 and 7200 ?

The timing can be looked at by enabling the time stamps in the kernel config and
looking at timestamps in the logs for start - stop (FE_HAS_LOCK) between the 2
changesets.

Regards,
Manu

_______________________________________________
linux-dvb mailing list
[email protected]
http://www.linuxtv.org/cgi-bin/mailman/listinfo/linux-dvb

Quoted 5 lines · Vangelis Nonas

Thank you all for your help,

With the latest changeset I can tune channels correctly. I tried vdr and szap (from Manu).

I dont know how to undo the last changesets and go back to 7200. Please tell me how to do it and I'll get the timings.

clone the multiproto tree

from there, do a partial clone locally

hg clone -r 7200 multiproto multiproto_7200

Quoted 3 lines

I also tried to scan Hotbird and it seems that scanning is now consistent when you scan the same transponder more than once.

A full scan produced 1722 channels which is not bad, but it needs improvement. There are about 2200 channels I think.

If you can provide a log with verbose=5 for the channels which it doesn't scan it will be a bit more helpful. Will need the logs (/var/log/messages) to debug this as well. The module parameters the same for both the STB0899 and the STB6100

Quoted 1 line

I attach the log of the scan. It fails to scan certain transponders and I believe (not 100% sure, but I did some testing) that this is consistent across runs.

The log is just the output of the scan utility, will need the output from the driver and a bit of thoughts.

Regards,
Manu

Quoted 47 lines

Regards
Vagelis

Quoted 40 lines · Manu Abraham

Quoted 13 lines · Manu Abraham

Quoted 11 lines · Artem Makhutov

Hi,

Quoted 7 lines · Artem Makhutov

Quoted 3 lines · Manu Abraham

Are you sure that you got the top level 2 changes changeset 7204 and 7203 respectively ?

Oh, I only got 7203. Will try with 7204 in a few minutes.

Quoted 8 lines

Quoted 8 lines

Awesome! It fixed the problem:

Try: 100
Failes: 0
Tunes: 100

Great job!

Also, can you please do a benchmark in lock timings between changeset 7205 and 7200 ?

The timing can be looked at by enabling the time stamps in the kernel config and looking at timestamps in the logs for start - stop (FE_HAS_LOCK) between the 2 changesets.

Regards,
Manu

_______________________________________________
linux-dvb mailing list
[email protected]
http://www.linuxtv.org/cgi-bin/mailman/listinfo/linux-dvb

Hello,

I attach the timing logs for changsets 7205 and 7201. The funny thing is that when I use 7201 it does NOT have the problem with locking when tuning, as it had before. I tried it 3 times, the third after a full shutdown of the pc and having removed the power for a couple on minutes. It is really strange.

Anyway, both logs were taken with these commands:

../szap/szap -c channels.conf -r "bbc world"
../szap/szap -c channels.conf -r "bbc prime"
../szap/szap -c channels.conf -r "filmnet1"

I'll check how scanning performs with 7201 and let you know

Regards
Vagelis

Quoted 34 lines · Manu Abraham

Quoted 7 lines · Vangelis Nonas

Thank you all for your help,

With the latest changeset I can tune channels correctly. I tried vdr and szap (from Manu).

I dont know how to undo the last changesets and go back to 7200. Please tell me how to do it and I'll get the timings.

clone the multiproto tree

from there, do a partial clone locally

hg clone -r 7200 multiproto multiproto_7200

Quoted 5 lines

I also tried to scan Hotbird and it seems that scanning is now consistent when you scan the same transponder more than once.

A full scan produced 1722 channels which is not bad, but it needs improvement. There are about 2200 channels I think.

If you can provide a log with verbose=5 for the channels which it doesn't
scan it will be a bit more helpful. Will need the logs (/var/log/messages)
to debug this as well. The module parameters the same for both the STB0899 and the STB6100

Quoted 3 lines

I attach the log of the scan. It fails to scan certain transponders and I believe (not 100% sure, but I did some testing) that this is consistent across runs.

Quoted 63 lines

The log is just the output of the scan utility, will need the output from
the driver and a bit of thoughts.

Regards,
Manu

Quoted 51 lines

Regards
Vagelis

Quoted 44 lines · Manu Abraham

Quoted 14 lines · Manu Abraham

Quoted 12 lines · Artem Makhutov

Hi,

Quoted 8 lines · Artem Makhutov

Quoted 4 lines · Manu Abraham

Are you sure that you got the top level 2 changes changeset 7204 and 7203
respectively ?

Oh, I only got 7203. Will try with 7204 in a few minutes.

Quoted 8 lines

Quoted 8 lines

Awesome! It fixed the problem:

Try: 100
Failes: 0
Tunes: 100

Great job!

Also, can you please do a benchmark in lock timings between changeset 7205 and 7200 ?

The timing can be looked at by enabling the time stamps in the kernel config and
looking at timestamps in the logs for start - stop (FE_HAS_LOCK) between the 2
changesets.

Regards,
Manu

_______________________________________________
linux-dvb mailing list
[email protected]
http://www.linuxtv.org/cgi-bin/mailman/listinfo/linux-dvb

------------------------------------------------------------------------

_______________________________________________
linux-dvb mailing list
[email protected]
http://www.linuxtv.org/cgi-bin/mailman/listinfo/linux-dvb

Hello,

I tried scanning with the 7201 changeset and I believe the results are better than the 7205 changeset. In the former case I get 2082 services on hotbird, in the latter 1722.

I'll shortly send kernel logs for a failing transponder during scan.

And something else:
When I give to scan the parameter -o vdr it will not output the results after a complete scan.

Quoted 6 lines · Vangelis Nonas

Hello,

I tried scanning with the 7201 changeset and I believe the results are better than the 7205 changeset. In the former case I get 2082 services on hotbird, in the latter 1722.

Hmm.. Ok.

Can you please get 7205 and load the stb0899 and stb6100 modules with verbose=2 or 5 as module parameters and see whether it makes any difference ?
(ie you see more of the services) ie check whether changing the module parameters
(verbosity level) makes any difference in the number of services found.

Quoted 2 lines

I'll shortly send kernel logs for a failing transponder during scan.

Quoted 3 lines

And something else:
When I give to scan the parameter -o vdr it will not output the results after a complete scan.

Ok, this i had not added in, will look at this aspect.

Regards,
Manu

Hello,

I scanned using 7205 changeset. With verbose=5, I get 1801 services. With verbose=2 I get 1725 services.

Using the 7201 changeset with verbose=5 I get 2082 services. I can check for verbose=2 and changeset 7201 if you think it is useful.

I have the "feeling" that 7201 behaves better. It is much faster also during scanning( I don't have measurements but I am pretty sure).

Should you need kernel logs for failing transponders either 7201 or 7205 let me know.

Regards
Vagelis

Quoted 30 lines · Manu Abraham

Quoted 6 lines · Vangelis Nonas

Hello,

I tried scanning with the 7201 changeset and I believe the results are better than the 7205 changeset. In the former case I get 2082 services on hotbird, in the latter 1722.

Hmm.. Ok.

Can you please get 7205 and load the stb0899 and stb6100 modules with verbose=2 or 5 as module parameters and see whether it makes any difference ?
(ie you see more of the services) ie check whether changing the module parameters
(verbosity level) makes any difference in the number of services found.

Quoted 5 lines

I'll shortly send kernel logs for a failing transponder during scan.

And something else:
When I give to scan the parameter -o vdr it will not output the results after a complete scan.

Ok, this i had not added in, will look at this aspect.

Regards,
Manu

Quoted 5 lines · Vangelis Nonas

Hello,

I scanned using 7205 changeset. With verbose=5, I get 1801 services. With verbose=2 I get 1725 services.

Using the 7201 changeset with verbose=5 I get 2082 services. I can check for verbose=2 and changeset 7201 if you think it is useful.

It would be nice to see changeset 7201 with verbose=1 and 2, the results.

Quoted 1 line

I have the "feeling" that 7201 behaves better. It is much faster also during scanning( I don't have measurements but I am pretty sure).

You can enable timestamps being output in printk's in the kernel config to see the timestamps, with which you compare with the start - stop events in the logs.

Quoted 1 line

Should you need kernel logs for failing transponders either 7201 or 7205 let me know.

Will need the logs for the failing transponders in 7205, to check for the actual cause.

Regards,
Manu

Quoted 35 lines

Regards
Vagelis

Quoted 25 lines · Manu Abraham

Quoted 4 lines · Vangelis Nonas

Hello,

I tried scanning with the 7201 changeset and I believe the results are better than the 7205 changeset. In the former case I get 2082 services on hotbird, in the latter 1722.

Hmm.. Ok.

Can you please get 7205 and load the stb0899 and stb6100 modules with verbose=2 or 5 as module parameters and see whether it makes any difference ? (ie you see more of the services) ie check whether changing the module parameters (verbosity level) makes any difference in the number of services found.

Quoted 4 lines

I'll shortly send kernel logs for a failing transponder during scan.

And something else:
When I give to scan the parameter -o vdr it will not output the results after a complete scan.

Ok, this i had not added in, will look at this aspect.

Regards,
Manu

.

Hello,

I have tested again changesets 7201 and 7205 with verbose 5, 2 and 1 scanning 101 transponders on Hotbird (I attach my transponders file -- it is taken and adopted from ProgDvb).

Here are my statistics:

Changeset Verbose channels
--------------------------------

7201         1        2152
7201         2        2105
7201         5        2081
7205         1        1760
7205         2        1608
7205         5        1578

Quoted 8 lines · Vangelis Nonas

Hello,

I have tested again changesets 7201 and 7205 with verbose 5, 2 and 1 scanning 101 transponders on Hotbird (I attach my transponders file -- it is taken and adopted from ProgDvb).

Here are my statistics:

Quoted 8 lines

Changeset Verbose channels
--------------------------------

7201         1        2152
7201         2        2105
7201         5        2081
7205         1        1760
7205         2        1608
7205         5        1578

Are you "really" sure that 7201 behaves better than others. I ask this, since
there was a bug in 7201 which caused many people not to have a LOCK, the bugs which was fixed in 7203 and 7204. I am at a loss now, as to understanding this strange phenomena, how a lock was achieved with no communication to the tuner.

If it were 7200, i could have still believed, there was a possibility, but 7201
i am terribly confused.

Can you please verify whether you didn't get mixed up with the changeset numbers or the logs that were produced ?

Regards,
Manu

.

Hello,

Here is the output of hg log|head -n 5 for two different directories (multiproto and multiproto_7200)

for multiproto:

changeset:   7205:9bdb997e38b5
tag:         tip
user:        Manu Abraham <[email protected]>
date:        Sun Feb 24 02:10:56 2008 +0400
summary:     We can now reduce the debug levels, just need to look at 

errors only.

for multiproto_7200:

changeset:   7200:45eec532cefa
tag:         tip
parent:      7095:a577a5dbc93d
parent:      7199:0448e5a6d8a6
user:        Manu Abraham <[email protected]>

So I guess I was referring to 7200 and not to 7201.
I am very positive about the results because I have tested it many times. It is just that it is 7200 instead of 7201.
So as a concluesion, 7200 behaves better than 7205. My corrected little table follows below just for clarification.

Changeset Verbose channels
--------------------------------

7200         1        2152
7200         2        2105
7200         5        2081
7205         1        1760
7205         2        1608
7205         5        1578

I apologise for the confusion, I may have caused.

Hello,

Here is the output of hg log|head -n 5 for two different directories (multiproto and multiproto_7200)

for multiproto:

changeset:   7205:9bdb997e38b5
tag:         tip
user:        Manu Abraham <[email protected]>
date:        Sun Feb 24 02:10:56 2008 +0400
summary:     We can now reduce the debug levels, just need to look at 

errors only.

for multiproto_7200:

changeset:   7200:45eec532cefa
tag:         tip
parent:      7095:a577a5dbc93d
parent:      7199:0448e5a6d8a6
user:        Manu Abraham <[email protected]>

So I guess I was referring to 7200 and not to 7201.
I am very positive about the results because I have tested it many times. It is just that it is 7200 instead of 7201.
So as a concluesion, 7200 behaves better than 7205. My corrected little table follows below just for clarification.

Changeset Verbose channels
--------------------------------

7200         1        2152
7200         2        2105
7200         5        2081
7205         1        1760
7205         2        1608
7205         5        1578

I apologise for the confusion, I may have caused.

I can confirm that current hg-tree breaks SkyStarHD. Mostly unusable, many transponders not working.
I changed back to 45eec532cefa and everything is working fine.

regards
Oliver

-----Original Message-----
From: [email protected] [mailto:[email protected]] On Behalf Of Manu Abraham
Sent: Dienstag, 26. Februar 2008 22:45
To: [email protected]
Subject: [linux-dvb] STB0899 users, please verify results was Re:
TechniSat SkyStar HD: Problems scaning and zaping

Quoted 21 lines · Vangelis Nonas

Hello,

Here is the output of hg log|head -n 5 for two different directories (multiproto and multiproto_7200)

for multiproto:

changeset:   7205:9bdb997e38b5
tag:         tip
user:        Manu Abraham <[email protected]>
date:        Sun Feb 24 02:10:56 2008 +0400
summary:     We can now reduce the debug levels, just need to look at 

errors only.

for multiproto_7200:

changeset:   7200:45eec532cefa
tag:         tip
parent:      7095:a577a5dbc93d
parent:      7199:0448e5a6d8a6
user:        Manu Abraham <[email protected]>

Quoted 5 lines

So I guess I was referring to 7200 and not to 7201.
I am very positive about the results because I have tested it many times. It is just that it is 7200 instead of 7201.
So as a concluesion, 7200 behaves better than 7205. My corrected little table follows below just for clarification.

Ok, fine this is very much possible. Can other STB0899 users (all of them) verify this result, such that i can revert back the changes ?

Quoted 11 lines

Changeset Verbose channels
--------------------------------

7200         1        2152
7200         2        2105
7200         5        2081
7205         1        1760
7205         2        1608
7205         5        1578

I apologise for the confusion, I may have caused.

No worries, don't apologise, this is all a part of the testing phase.

Thanks,
Manu

_______________________________________________
linux-dvb mailing list
[email protected]
http://www.linuxtv.org/cgi-bin/mailman/listinfo/linux-dvb

I can confirm that current hg-tree breaks SkyStarHD. Mostly unusable, many transponders not working.
I changed back to 45eec532cefa and everything is working fine.

Seppo, You should use the group-mailer [email protected]

regards
Oliver

-----Original Message-----
From: Seppo Ingalsuo [mailto:[email protected]]
Sent: Donnerstag, 28. Februar 2008 07:08
To: Oliver Bardenheier (obardenh)
Subject: Re: [linux-dvb] STB0899 users, please verify results was Re:
TechniSat SkyStar HD: Problems scaning and zaping

Quoted 5 lines · Oliver Bardenheier (obardenh)

I can confirm that current hg-tree breaks SkyStarHD. Mostly unusable, many transponders not working.
I changed back to 45eec532cefa and everything is working fine.

I'm using multiproto-74c6c2ee613b grabbed yesterday and I was able to see only BBC World from Astra 19.2E. Other channels I tried to zap to didn't work. Also there seemed to be problems in driving reliably diseqc positioner (that the old TT budget-s handled fine).

It was my first experience with brand new TT S2-3200 and vdr 1.5.16 with Reinhard's H.264 patch set.

BR,
Seppo

Seppo, You should use the group-mailer [email protected]

regards
Oliver

-----Original Message-----
From: Seppo Ingalsuo [mailto:[email protected]]
Sent: Donnerstag, 28. Februar 2008 07:08
To: Oliver Bardenheier (obardenh)
Subject: Re: [linux-dvb] STB0899 users, please verify results was Re:
TechniSat SkyStar HD: Problems scaning and zaping

Quoted 6 lines · Oliver Bardenheier (obardenh)

I can confirm that current hg-tree breaks SkyStarHD. Mostly unusable,

many transponders not working.
I changed back to 45eec532cefa and everything is working fine.

I'm using multiproto-74c6c2ee613b grabbed yesterday and I was able to see only BBC World from Astra 19.2E. Other channels I tried to zap to didn't work. Also there seemed to be problems in driving reliably diseqc positioner (that the old TT budget-s handled fine).

It was my first experience with brand new TT S2-3200 and vdr 1.5.16 with Reinhard's H.264 patch set.

BR,
Seppo

No problem, :-)

You just check out the tree via the webpage:
http://jusst.de/hg/multiproto

then click on the needed changeset
http://jusst.de/hg/multiproto/rev/0448e5a6d8a6

and download the bzip with the complete changeset
http://jusst.de/hg/multiproto/archive/0448e5a6d8a6.tar.bz2

The rest is routine.... :-)

Oliver

-----Original Message-----
From: [email protected] [mailto:[email protected]] On Behalf Of Seppo Ingalsuo
Sent: Donnerstag, 28. Februar 2008 16:53
To: [email protected]
Subject: Re: [linux-dvb] STB0899 users, please verify results was Re:
TechniSat SkyStar HD: Problems scaning and zaping

Quoted 3 lines · Oliver Bardenheier (obardenh)

Seppo, You should use the group-mailer [email protected]

Sorry Olivier, I didn't mean to send you personal email!

I still have the same problem and I'm not familiar with various DVB driver sources discussed here (hg-tree, mantis). As I suppose I have the same issue as discussed in this thread, how do I retrieve an older version of the driver for TT S2-3200 that would work?

BR,
Seppo

Quoted 27 lines

regards
Oliver

-----Original Message-----
From: Seppo Ingalsuo [mailto:[email protected]]
Sent: Donnerstag, 28. Februar 2008 07:08
To: Oliver Bardenheier (obardenh)
Subject: Re: [linux-dvb] STB0899 users, please verify results was Re:
TechniSat SkyStar HD: Problems scaning and zaping

Quoted 9 lines · Oliver Bardenheier (obardenh)

I can confirm that current hg-tree breaks SkyStarHD. Mostly unusable,

many transponders not working.
I changed back to 45eec532cefa and everything is working fine.

I'm using multiproto-74c6c2ee613b grabbed yesterday and I was able to see only BBC World from Astra 19.2E. Other channels I tried to zap to didn't work. Also there seemed to be problems in driving reliably diseqc positioner (that the old TT budget-s handled fine).

Quoted 8 lines

It was my first experience with brand new TT S2-3200 and vdr 1.5.16 with Reinhard's H.264 patch set.

BR,
Seppo

_______________________________________________
linux-dvb mailing list
[email protected]
http://www.linuxtv.org/cgi-bin/mailman/listinfo/linux-dvb

_______________________________________________
linux-dvb mailing list
[email protected]
http://www.linuxtv.org/cgi-bin/mailman/listinfo/linux-dvb

I think this page is very useful, because it explains the procedure quite well (also tells you about modified scan and zap):
http://www.linuxtv.org/wiki/index.php/Multiproto

No problem, :-)

You just check out the tree via the webpage:
http://jusst.de/hg/multiproto

then click on the needed changeset
http://jusst.de/hg/multiproto/rev/0448e5a6d8a6

and download the bzip with the complete changeset
http://jusst.de/hg/multiproto/archive/0448e5a6d8a6.tar.bz2

The rest is routine.... :-)

Hello,

I tried scanning with the 7201 changeset and I believe the results are better than the 7205 changeset. In the former case I get 2082 services on hotbird, in the latter 1722.

I'll shortly send kernel logs for a failing transponder during scan.

And something else:
When I give to scan the parameter -o vdr it will not output the results after a complete scan.

Regards
Vagelis

Quoted 112 lines · Vangelis Nonas

Hello,

I attach the timing logs for changsets 7205 and 7201. The funny thing is that when I use 7201 it does NOT have the problem with locking when tuning, as it had before. I tried it 3 times, the third after a full shutdown of the pc and having removed the power for a couple on minutes. It is really strange.

Anyway, both logs were taken with these commands:

../szap/szap -c channels.conf -r "bbc world"
../szap/szap -c channels.conf -r "bbc prime"
../szap/szap -c channels.conf -r "filmnet1"

I'll check how scanning performs with 7201 and let you know

Regards
Vagelis

Quoted 31 lines · Manu Abraham

Quoted 7 lines · Vangelis Nonas

Thank you all for your help,

With the latest changeset I can tune channels correctly. I tried vdr and szap (from Manu).

I dont know how to undo the last changesets and go back to 7200. Please tell me how to do it and I'll get the timings.

clone the multiproto tree

from there, do a partial clone locally

hg clone -r 7200 multiproto multiproto_7200

Quoted 5 lines

I also tried to scan Hotbird and it seems that scanning is now consistent when you scan the same transponder more than once.

A full scan produced 1722 channels which is not bad, but it needs improvement. There are about 2200 channels I think.

If you can provide a log with verbose=5 for the channels which it doesn't
scan it will be a bit more helpful. Will need the logs (/var/log/messages)
to debug this as well. The module parameters the same for both the STB0899 and the STB6100

Quoted 3 lines

I attach the log of the scan. It fails to scan certain transponders and I believe (not 100% sure, but I did some testing) that this is consistent across runs.

Quoted 60 lines

The log is just the output of the scan utility, will need the output from
the driver and a bit of thoughts.

Regards,
Manu

Quoted 49 lines

Regards
Vagelis

Quoted 43 lines · Manu Abraham

Quoted 14 lines · Manu Abraham

Quoted 12 lines · Artem Makhutov

Hi,

Quoted 8 lines · Artem Makhutov

Quoted 4 lines · Manu Abraham

Are you sure that you got the top level 2 changes changeset 7204 and 7203
respectively ?

Oh, I only got 7203. Will try with 7204 in a few minutes.

Quoted 8 lines

Quoted 8 lines

Awesome! It fixed the problem:

Try: 100
Failes: 0
Tunes: 100

Great job!

Also, can you please do a benchmark in lock timings between changeset 7205 and 7200 ?

The timing can be looked at by enabling the time stamps in the kernel config and
looking at timestamps in the logs for start - stop (FE_HAS_LOCK) between the 2
changesets.

Regards,
Manu

_______________________________________________
linux-dvb mailing list
[email protected]
http://www.linuxtv.org/cgi-bin/mailman/listinfo/linux-dvb

Quoted 6 lines

------------------------------------------------------------------------

_______________________________________________
linux-dvb mailing list
[email protected]
http://www.linuxtv.org/cgi-bin/mailman/listinfo/linux-dvb

_______________________________________________
linux-dvb mailing list
[email protected]
http://www.linuxtv.org/cgi-bin/mailman/listinfo/linux-dvb

Also, can you please do a benchmark in lock timings between changeset 7205 and 7200 ?

[...]

Hi,

Quoted 2 lines · Manu Abraham

Also, can you please do a benchmark in lock timings between changeset 7205 and 7200 ?

Do you mean changeset 7200 or 7204

Quoted 7 lines · Artem Makhutov

Hi,

Quoted 2 lines · Manu Abraham

Also, can you please do a benchmark in lock timings between changeset 7205 and 7200 ?

Do you mean changeset 7200 or 7204

yep. 7205 is head, but just a minor change.

Hi,

Quoted 6 lines · Manu Abraham

Quoted 5 lines · Artem Makhutov

Hi,

Quoted 2 lines · Manu Abraham

Also, can you please do a benchmark in lock timings between changeset 7205 and 7200 ?

Quoted 2 lines

Quoted 1 line

Do you mean changeset 7200 or 7204

yep. 7205 is head, but just a minor change.

I did the benchmark betweet 7205 and 7204. The modules were loaded with no additional parameters.

Tuning time in 7204: ~0.83 seconds (min 0.81191 sec ; max 2.428967 sec) Tuning time in 7205: ~0.26 seconds (min 0.24789 sec ; max 1.994216 sec)

Quoted 15 lines · Artem Makhutov

Hi,

Quoted 8 lines · Manu Abraham

Quoted 6 lines · Artem Makhutov

Hi,

Quoted 2 lines · Manu Abraham

Also, can you please do a benchmark in lock timings between changeset 7205 and 7200 ?

Do you mean changeset 7200 or 7204

yep. 7205 is head, but just a minor change.

I did the benchmark betweet 7205 and 7204. The modules were loaded with no additional parameters.

Quoted 2 lines

Tuning time in 7204: ~0.83 seconds (min 0.81191 sec ; max 2.428967 sec) Tuning time in 7205: ~0.26 seconds (min 0.24789 sec ; max 1.994216 sec)

Can you test the time between 7200 and 7204/7205.
(7201 - 7204 contains an optimization, compared to 7200 and hence the need to check the results. So need to exclude 7201 and 7204)

Hi,

Quoted 7 lines · Manu Abraham

Quoted 6 lines · Artem Makhutov

Hi,

Quoted 2 lines · Manu Abraham

Also, can you please do a benchmark in lock timings between changeset 7205 and 7200 ?

Do you mean changeset 7200 or 7204

I did the benchmark betweet 7205 and 7204. The modules were loaded with no additional parameters.

Tuning time in 7204: ~0.83 seconds (min 0.81191 sec ; max 2.428967 sec) Tuning time in 7205: ~0.26 seconds (min 0.24789 sec ; max 1.994216 sec)

I have attached the logfiles of the tests.

I made 60 tunes for 7204 and 45 tunes for 7205.

Quoted 19 lines · Artem Makhutov

Hi,

Quoted 7 lines · Manu Abraham

Quoted 6 lines · Artem Makhutov

Hi,

Quoted 2 lines · Manu Abraham

Also, can you please do a benchmark in lock timings between changeset 7205 and 7200 ?

Do you mean changeset 7200 or 7204

I did the benchmark betweet 7205 and 7204. The modules were loaded with no additional parameters.

Tuning time in 7204: ~0.83 seconds (min 0.81191 sec ; max 2.428967 sec) Tuning time in 7205: ~0.26 seconds (min 0.24789 sec ; max 1.994216 sec)

I have attached the logfiles of the tests.

Quoted 1 line

I made 60 tunes for 7204 and 45 tunes for 7205.

From the 7205 logs, it looks no lock failure at all. Is this too good to believe ?

but i do see that you have a lock loss case in 7204. is that 7204 and 7205 logs got interchanged ?