1.3.0 Final and Digital Channel Scans (1 Viewer)

[mm1352000]

Just a quick post as it is 4:30am here...

1. Signal levels not showing for cable scan.
2. Unknown channel names marked as xxxx.yyy, not understandable.
3. Encrypted channels not marked as encrypted.
4. LCNs not set.
5. xx.yy LCNs not supported.
6. Increasing SDT/VCT timeout or rescanning may find more channels.
7. Physical channel numbers off by one.

Attached are solutions for issues 2, 4 and 7.

Solution to two is set name as "Unknown <physical>-<service ID>" as requested.
Solution to four is set LCN = (<physical> x 1000) + service ID. It is the best we can do at this stage. If/when we add support for 5 then we would obviously update this.
http://mantis.team-mediaportal.com/view.php?id=4406

Solution to seven was errors in tuning detail files. The errors would only have caused the offset for channels 23..95 (or something like that)... which is why it had me so confused.
http://mantis.team-mediaportal.com/view.php?id=4405

Instructions for installing and testing:

1. Open TV Server configuration.
2. Go to the "manual control" section and click "stop service".
3. Close TV Server configuration.
4. Open Windows Explorer and navigate to the folder you installed TV Server in. Default is C:\Program Files\Team MediaPortal\MediaPortal TV Server on a 32 bit version of Windows or C:\Program Files (x86)\Team ... on a 64 bit version of Windows.
5. Take a backup of the TVLibrary.dll.
6. Download and extract the attached TVLibrary.dll into the same folder.
7. Copy and paste "C:\ProgramData\Team MediaPortal\MediaPortal TV Server\TuningParameters\atsc\" into the address bar and press enter.
8. Replace the tuning detail XML files with the attached versions.

After this, please test scanning again.
You'll need to delete the cable channels before scanning as MP intentionally does not update the names or LCNs of existing channels.
Assuming you've done a significant amount of renaming etc. already, you can save your exising channel list using the import/export section of TV Server configuration.

Please use the QAM Standard tuning detail file if at all possible.I know you said you prefer the IRC one, but most people use the QAM Standard one. I can't get these changes into the next release unless they're tested... so QAM Standard really needs to be tested if at all possible.

Still to look at the duplicate channel issue...

@RonD @2BitSculptor
Any chance either of you can help with the testing?
Ron, I note that I've used [something like] your convention for assigning the LCN for both ATSC and cable.
Where available:
(<major channel> x 1000) + <minor channel>
...otherwise...
(<physical channel> x 1000) + <service ID>

I know you'd possibly prefer shorter numbers, but apparently for for cable the minor channel can run up to 999. I see @breese has examples that are definitely over 100 in his log files. Hence the need for multiplying by 1000 instead of 100.

 

[RonD]

mm,

I tried the new tvlibrary.dll and QAM tuning details and I don't NOT see any auto LCN channels, just get default 10000 for all channels. Do yo also need to upload a new Tswriter file that supplies updated mapped LCNs/channels.

I did 2 tests, test1 stop tv server, copy new files to folder serer, leave my original channels, and do a scan. test2, delete the old channels, do a new scan. In both cases I just get default 10000 for channels. The attached files are the tv.log and TsWriter.log files

EDIT, another comment, before I did the tests I did an export of all my channels and groups using TV Server Config > export, then did my my testing. To restore my original channels, I deleted all the new scanned channels and did an import to restore the channels/groups.

a typical channel scan event is pasted below

2013-04-30 13:30:38.243000 [(5)]: Scan: tuner locked:True signal:100 quality:83
2013-04-30 13:30:43.336000 [(5)]: 0) 0x0 0x18D 0x3 0x30 KNTV HD type:2
2013-04-30 13:30:43.337000 [(5)]: atsc:Found: ATSC:tv:unknown KNTV HD Freq:-1 ONID:0 TSID:397 SID:3 PMT:0x30 FTA:True LCN:10000 phys:12 maj:11 min:1 mod:Mod8Vsb
2013-04-30 13:30:43.338000 [(5)]: 1) 0x0 0x18D 0x4 0x40 Cozi Television type:2
2013-04-30 13:30:43.339000 [(5)]: atsc:Found: ATSC:tv:unknown Cozi Television Freq:-1 ONID:0 TSID:397 SID:4 PMT:0x40 FTA:True LCN:10000 phys:12 maj:11 min:2 mod:Mod8Vsb
2013-04-30 13:30:43.340000 [(5)]: Scan Got 2 from 2 channels
2013-04-30 13:30:43.417000 [ATSC scan thread(9)]: ATSC tune: QAM checkbox not selected... using Modulation 8Vsb
2013-04-30 13:30:43.418000 [ATSC scan thread(9)]: ATSC tune: PhysicalChannel: 13 Frequency: -1 Modulation: Mod8Vsb
2013-04-30 13:30:43.451000 [(5)]: atsc:Tune:ATSC:tv:  Freq:-1 ONID:-1 TSID:-1 SID:-1 PMT:0xFFFFFFFF FTA:False LCN:10000 phys:13 maj:-1 min:-1 mod:Mod8Vsb

 

[mm1352000]

Thanks for testing Ron!

You're right - I made the kind of stupid mistake I only make when I work at 4am in the morning.
Try the attached again please... and sorry for that previous waste of time. I'm confident with this one.

 

[RonD]

mm double check the zip you just uploaded, I still just see the tvlibrary file

 

[mm1352000]

mm double check the zip you just uploaded, I still just see the tvlibrary file

...as intended. There is no change to TsWriter.

 

[RonD]

mm, tried the new^2 tvlibrary file, if leave the old channels in my channel list, do a new scan the channel numbers do NOT get updated. I guess this is expected. This is "test3" in the zip file

I then deleted all my channels, did another scan, "test4" in the zip file, I see the new auto-generated LCN/channel numbers in the TV Server Config > Channels. But there are a few weird cases.

1. In MePo Client Config, I had 4-digit channel numbers enabled, but I see all 5-6 channel digits when I look at the MePo Client TV guide. Not a major problem but I expected MePo client to only show 4 digits.

2. My KGO channel did not get scanned correctly. Phy = 7, Major = 7, Minor = 1 (2 more subchannels 7.2, 7.3). This showed as "unknown 7-3", "unknown 7-4", "unknown 7-5". This may be a general MePo scan robustness limitation. This is normally a "good" channel, but sometimes MePo does not always scan. This is also 1 of my SF weird stations that has 2 PhyChannels 7 and 35. I get better results with 7 and do some tweaking. Look for TSID:381. In "test3" log files it scanned correctly with both 7 and 35, in "test4" it scanned correctly with 35.

3. Two channels show unexpected major channel numbers. both have several sub-channels but I only show the 1st one. Not sure why MePo finds strange major channel numbers.

KKPX ION, LCN:257001 phys:41 maj:257 min:1, on a US TV this shows as 65-1
KFSF DT, LCN:258001 phys:34 maj:258 min:1, on a US TV this shows as 66-1

if I use the HDHR config tool, enter phys:41 or phys:34, HDHR always shows the expected 65-1 or 66-1 Major.Minor channel numbers. With HDHR there is a way to run a manual scan and save the results in a text file. I'll look around the HDHR web pages to find how to save HDHR scan data for reference.

EDIT, added the HDHR scan0.log file not sure this helps, info from
http://www.silicondust.com/hdhomerun/hdhomerun_development.pdf
ss = signal strength. 80% is approximately -12dBmV.
snq = signal to noise quality (based on analog signal to noise ratio).
seq = symbol error quality (number of uncorrectable digital errors detected).

 

[mm1352000]

Hello again

mm, tried the new^2 tvlibrary file, if leave the old channels in my channel list, do a new scan the channel numbers do NOT get updated. I guess this is expected.

Yes this is expected, as per my comment above...

You'll need to delete the cable channels before scanning as MP intentionally does not update the names or LCNs of existing channels.

1. In MePo Client Config, I had 4-digit channel numbers enabled, but I see all 5-6 channel digits when I look at the MePo Client TV guide. Not a major problem but I expected MePo client to only show 4 digits.

AIUI that setting is not a control of how many digits show.
For example, if your channel numbers were all 2 digits long and the setting was 4, I wouldn't expect MP to pad the numbers out to 4 digits.
(Does it do that? At work, so can't check...)
What it is meant to be is an indicator for the zapping code so once you enter x number of digits MP will know to change channel without applying the zap delay.
For example, changing to channel "2" with the setting value "3" is faster when you enter "002".
Make sense?

2. My KGO channel did not get scanned correctly. Phy = 7, Major = 7, Minor = 1 (2 more subchannels 7.2, 7.3). This showed as "unknown 7-3", "unknown 7-4", "unknown 7-5". This may be a general MePo scan robustness limitation. This is normally a "good" channel, but sometimes MePo does not always scan. This is also 1 of my SF weird stations that has 2 PhyChannels 7 and 35. I get better results with 7 and do some tweaking. Look for TSID:381. In "test3" log files it scanned correctly with both 7 and 35, in "test4" it scanned correctly with 35.

Right, so it scanned as unknown on physical 7 but found the names on physical 35. Possibly this:
http://mantis.team-mediaportal.com/view.php?id=3707

3. Two channels show unexpected major channel numbers. both have several sub-channels but I only show the 1st one. Not sure why MePo finds strange major channel numbers.



KKPX ION, LCN:257001 phys:41 maj:257 min:1, on a US TV this shows as 65-1

KFSF DT, LCN:258001 phys:34 maj:258 min:1, on a US TV this shows as 66-1



if I use the HDHR config tool, enter phys:41 or phys:34, HDHR always shows the expected 65-1 or 66-1 Major.Minor channel numbers. With HDHR there is a way to run a manual scan and save the results in a text file. I'll look around the HDHR web pages to find how to save HDHR scan data for reference.

Hmmm, okay.
In binary terms those numbers are really close.
65 = 0b1000001
257 = 0b100000001

Let me see... what does TsWriter do:
https://github.com/MediaPortal/Medi...iter/source/VirtualChannelTableParser.cpp#153


...with reference to ATSC A/65:
http://www.atsc.org/cms/standards/a_65-2009.pdf
(page 32 table 6.4)

And there we see a bug. Code is:

  int major_channel =((buf[start ]&0xf)<<8) + (buf[start+1]>>2);
  int minor_channel =((buf[start+1]&0x3)<<8) + buf[start+2];

Code should be:

  int major_channel =((buf[start ]&0xf)<<6) + (buf[start+1]>>2);
  int minor_channel =((buf[start+1]&0x3)<<8) + buf[start+2];

[edit: bolding doesn't work with code! Note the 6 instead of the 8 in the first statement.]

And this bug would have precisely the effect that you described. It only affects major channel numbers greater than 63.

mm

 

[mm1352000]

Mantis:
http://mantis.team-mediaportal.com/view.php?id=4410

 

[Pat Clark]

I like to see good work! Way to go guys.

 

[RonD]

So looking at some large cities in www.rabbitears.info > listings > WebListings the highest ATSC major channel number is 69, the larger cities have 2-4 channels from 64 to 69. Click on a city to see a list of channels for that city.
http://www.rabbitears.info/market.php

In a city, you can click on "expand all" to see subchannel and additional info for the channels, for any channel you can click on "Technical Data and Screencaps" to get more info on the ATSC sub-channels.