Automatically Tune to Strongest Duplicate Channel (1 Viewer)

[vapourEyes]

Please,take a look at the red rectangle, it's possible to have signal for channel.

View attachment 204809

Thank you @framug.

I scoured the code high and low for per-channel/MUX strength.

Can you point me to the right direction in the code.

I may have missed it being a bit of a newbie.

 

[framug]

Well, it's very old (2009) but, if you look at pictures and search in classes I modified for numeric signal values there, I'm sure you will find easily how to.

 

[vapourEyes]

Well, it's very old (2009) but, if you look at pictures and search in classes I modified for numeric signal values there, I'm sure you will find easily how to.

Thanks @framug,

These patches are all UI patches ?

Ideally I need to get to the point in the code where I insert the StrengthXQuality measure.

To be fair to all channels on all MUX'es I need to multiply quality by strength to arrive at a value that can feed the comparitor which decides which channel / MUX wins overall during duplicates.
We are here in the code, at which point we need to store qualityXstrength number.

So, I'm wondering if I need to enrich an existing object with the strength, for use later.

If I can get this done, I can re-use these stengths to satisfy an implementation for Smart-Autotune hinted at here:
Automatically Re-Tune Periodically.

 

[framug]

To be fair to all channels on all MUX'es I need to multiply quality by strength to arrive at a value that can feed the comparitor which decides which channel / MUX wins overall during duplicates.

From my personal research, quality impact more video/audio than strength but multiply could be a good way.

So, I'm wondering if I need to enrich an existing object with the strength, for use later.

If I can get this done, I can re-use these stengths to satisfy an implementation for Smart-Autotune hinted at here:
Automatically Re-Tune Periodically.

For me yes, you should store it, your second part could then, be done by an external process plugin, for example.
Of course, you will need a free tuner of your card(s) to do that.

 

[CyberSimian]

I need to multiply quality by strength to arrive at a value that can feed the comparitor

From my personal research, quality impact more video/audio than strength

So far in this discussion we have been talking about "signal strength", but now that you mention it, I think that "signal quality" is the value that we should be using to compare channels. Both strength and quality are arbitrary measures that have no absolute meaning, and TV Server scales each to fit in the range 0-100. Suppose during a scan you encountered the following:

• Channel "BBC2 SD" in MUX A: strength=80, quality=30
• Channel "BBC2 SD" in MUX B: strength=20, quality=70

Which channel should be retained? Surely it should be the one with the higher quality rating (MUX B). If you use strength*quality as the measure, you would choose MUX A (value=2400) instead of MUX B (value=1400), even though MUX A is likely to give worse reception than MUX B.

-- from CyberSimian in the UK

 

[ajs]

Which channel should be retained? Surely it should be the one with the higher quality rating (MUX B). If you use strength*quality as the measure, you would choose MUX A (value=2400) instead of MUX B (value=1400), even though MUX A is likely to give worse reception than MUX B.

Better use Increasing factor for the desired value. For example 3 for quality, ie strength + quality * f, for example from above: 80 + 30 * 3 < 20 + 70 * 3

 

[framug]

Channel "BBC2 SD" in MUX A: strength=80, quality=30
Channel "BBC2 SD" in MUX B: strength=20, quality=70

Which channel should be retained? Surely it should be the one with the higher quality rating (MUX B). If you use strength*quality as the measure, you would choose MUX A (value=2400) instead of MUX B (value=1400), even though MUX A is likely to give worse reception than MUX B.

In your example, I'm afraid you will see no video nor audio.
What I found (approximatively in DVB-T) for it works is strength not less than 35 and quality not less than 50.

In fact, if I understand correctly, you would like something like that ?

 

[vapourEyes]

In your example, I'm afraid you will see no video nor audio.
What I found (approximatively in DVB-T) for it works is strength not less than 35 and quality not less than 50.

In fact, if I understand correctly, you would like something like that ?

View attachment 204812

Awesome !!!!

 

[vapourEyes]

So far in this discussion we have been talking about "signal strength", but now that you mention it, I think that "signal quality" is the value that we should be using to compare channels. Both strength and quality are arbitrary measures that have no absolute meaning, and TV Server scales each to fit in the range 0-100. Suppose during a scan you encountered the following:

• Channel "BBC2 SD" in MUX A: strength=80, quality=30
• Channel "BBC2 SD" in MUX B: strength=20, quality=70

Which channel should be retained? Surely it should be the one with the higher quality rating (MUX B). If you use strength*quality as the measure, you would choose MUX A (value=2400) instead of MUX B (value=1400), even though MUX A is likely to give worse reception than MUX B.

-- from CyberSimian in the UK

Absolutely noted !!!

In that instance the code is simplified along with the UI. We only need to worry about 'Quality' post-normalisation ?

 

[CyberSimian]

In your example, I'm afraid you will see no video nor audio.

Understood .
I invented the numbers to illustrate the principle; I was not suggesing that they were reasonable!

Your example panel for these new features looks very good, but I would suggest:

(1) Rename the first section from "Tune to strongest channel" to "Duplicate channels". This makes it clearer that we are specifically addressing channels where two identical versions are found during the scan.

(2) Is "Enable this alternative tune" needed? If "Retain strongest channel" is not selected, isn't the eliminate-duplicates processing disabled? Or is there still some other processing occurring that should be disabled?

-- from CyberSimian in the UK