LinuxQuestions.org
Help answer threads with 0 replies.
Go Back   LinuxQuestions.org > Forums > Linux Forums > Linux - Hardware
User Name
Password
Linux - Hardware This forum is for Hardware issues.
Having trouble installing a piece of hardware? Want to know if that peripheral is compatible with Linux?

Notices

Reply
 
Search this Thread
Old 06-09-2008, 05:05 AM   #1
blackhole54
Senior Member
 
Registered: Mar 2006
Posts: 1,896

Rep: Reputation: 61
Interpreting Front End Data (Signal Strength, SNR, BER) for DVB devices


Hi All,

This isn't about getting a device working per se but this still seemed like this might be the right forum. I am trying to use a Hauppauge HDVR950 ATSC/NTSC USB tuner (lsusb results below). It's basically working but I am having some reception issues in ATSC mode. As I result I have been trying to look at low level data like signal strength, signal-to-noise (SNR), etc. Specifically I have been using and modifying dtvsignal (download: here) from pcHDTV to extract information from the dvb front end. Specifically this software does the following calls to pull the data:

Code:
        ioctl(device, FE_READ_STATUS, &status);
        ioctl(device, FE_READ_SIGNAL_STRENGTH, &snl);
        ioctl(device, FE_READ_SNR, &snr);
        ioctl(device, FE_READ_BER, &ber);
        ioctl(device, FE_READ_UNCORRECTED_BLOCKS, &uncorrected_blocks);
I have modified the code to report the values of the last four, but now I would like to find out exactly how to interpret the raw values. I have searched the internet, including finding this page, but still haven't found the info I am seeking.

One possibility is that this data just doesn't follow a standard format but is vendor (and maybe model) specific. In which case, knowing that would be useful. Even more useful would be knowing where one might get the vendor specific info. However,the previously mentioned software has hints that there might actually be a standard, at least for signal strength and SNR. But the documentation is almost non-existent and I hate reading too much into the the output of these small programs. Both signal strength and SNR are returned as unsigned 16 bits. The dtvsignal program hints that full scale (65535) is 30db, perhaps with the value being linear wrt the signal strength. Another program included (but not even mentioned in the README) is dtvsnr. In this case the program seems to be hinting that the SNR output is linear wrt decibels (of the ratio) and that 0 corresponds to -10 db and full scale (65535) corresponds to 40 db. But this program is even more problematic because, as I read it, it is not internally consistent. (Since its not mentioned in the README I've wondered if maybe the program was never finished/debugged.) But taking the signal strength as linear and the SNR as decibels with the indicated scaling does seem to have some internal consistency:

The following data was taken on various channels with a modified dtvsignal program. The signal strength is the raw data (after smoothing) expressed as percent of full scale. The SNR (also smoothed) is raw data scaled as indicated above (0 = -10 db; FS = 40 db).

Code:
channel = 15 freq = 479000000Hz    SNR: -5.1   Signal:   70%   
channel = 16 freq = 485000000Hz    SNR: -4.9   Signal:   74%   
channel = 17 freq = 491000000Hz    SNR: -5.6   Signal:   64%   
channel = 18 freq = 497000000Hz    SNR: -6.1   Signal:   57%   
channel = 19 freq = 503000000Hz    SNR: -5.9   Signal:   59%   
channel = 32 freq = 581000000Hz    SNR: -5.0   Signal:   73%   
channel = 34 freq = 593000000Hz    SNR: -5.4   Signal:   66%   
channel = 38 freq = 617000000Hz    SNR: -6.6   Signal:   49%
If it is assumed (yes, it is an assumption that may not be true) that all channels have similar noise, then picking any pair of channels, and comparing ten times the common log of the ratio (definition of db) does in fact give aproximately the difference of the numbers reported as SNR. So this does suggest an internal consistency with the interpretation of the meaning of the raw SNR value. But if those SNR and signal strength values are correctly interpreted, the values seem rather dismal, with signal buried in noise. (Which could account for my problems.)

The remaining issue is the bit error rate (BER). And maybe for me this is not an issue at all since I consistently get a value of zero. Which doesn't seem likely. Particularly considering the reception issues I have. Any thoughts on this?

I appologize for the fact that this post has been a little rambling. (Thanks for bearing with me.) I wanted to lay out what Iknow, what I suspect and where my thinking is on this, and this was the most orderly I could figure out to do so. Any pointers about where I might find actual documentation telling me how to interpret the raw data would be greatly appreciated. Thanks.



In case it is of any use, here is the output of lsusb:

Code:
Bus 002 Device 001: ID 0000:0000  
Bus 005 Device 001: ID 0000:0000  
Bus 004 Device 001: ID 0000:0000  
Bus 003 Device 001: ID 0000:0000  
Bus 001 Device 006: ID 2040:6513 Hauppauge 
Bus 001 Device 001: ID 0000:0000
P.S. The (GPL) programs I mentioned above have a copyright notice for pcHDTV, but no contact info.

Last edited by blackhole54; 06-09-2008 at 05:11 AM.
 
  


Reply


Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is Off
HTML code is Off


Similar Threads
Thread Thread Starter Forum Replies Last Post
Getting AP signal strength via C program ... toships Linux - Wireless Networking 5 01-07-2009 03:35 AM
Signal Strength Meter shows low signal? falcon56215 Linux - Networking 0 09-15-2006 08:39 AM
Signal strength not showing hosler Linux - Software 3 03-31-2006 09:17 AM
Wireless Signal Strength and Quality tubatodd Slackware 5 08-07-2005 09:30 AM
interpreting iwconfig output: signal strength, etc ahz10 Linux - Wireless Networking 1 04-11-2005 07:55 PM


All times are GMT -5. The time now is 07:39 AM.

Main Menu
My LQ
Write for LQ
LinuxQuestions.org is looking for people interested in writing Editorials, Articles, Reviews, and more. If you'd like to contribute content, let us know.
Main Menu
Syndicate
RSS1  Latest Threads
RSS1  LQ News
Twitter: @linuxquestions
identi.ca: @linuxquestions
Facebook: linuxquestions Google+: linuxquestions
Open Source Consulting | Domain Registration