Signal Strength vs Signal Quality

[perkunas]

This question has been asked 100 times before, but hear goes.
On a 100 cm, offset dish, on galaxy 19
The best I can get, is 90 percent intensity, and about 73 percent quality.
1st off, is that normal, moving the LNB does nothing. Ive have a better meter on order, so I can try and tweak it some more.
The quality of picture is good, but we always want the better lol.
P.S.
You guys are the best, I could never of got this working without you guys
tnx

 

[Titanium]

Each satellite receiver brand and model displays a different quality reading, so it is impossible to know what the real signal integrity is. Some STBs display 100% quality even on poor signals, while other models do not go above 60 or 70%.

Your receiver's display may max out in the mid 70's and you may be getting the best signal quality possible. Without an installation meter (or the rare STB) with a BER and CN readout you may never know if your system signal quality is maximized.

The signal quality reading on your receiver is likely a post FEC measurment, so error correction is being applied to any signal packet loss. Small LNBF placement or rotation settings and dish movements are not reflected in the quality reading, but it might be making the percentage of good signal vs noise much better. With a good meter, you will observe changes in the CN (Carrier to Noise) and BER (Bit Error Rate) readings with even a fraction of a degree rotation of the LNBF or sliding the LNBF foreward or aft.

What meter did you order?

 

[FaT Air]

SR and FEC can also influence your Q reading, and 'how low' the Q can go before reception is lost.
Low FEC usually stay locked much lower than high. NHK on 58 stays locked here down to around 10Q on the S9 here. 1/2 FEC.

Best way to know, without a spectrum analyzer, that you're 'centered' in the tune up [with some FTA boxes((and meters??))] is to tune up either side of 'peak' for some low(er) value. Then return it to half way between those points. On the S9 that's about the only way to do it. It typically 'tops out' around 70. It takes a LARGE signal increase to get it higher.

 

[perkunas]

Thanks I have a Sat-link ws-6906 on order, I used a standard type, cheep meter before.
This new meter is suppose to have the CN Ber reading (will need to figure out how that works.)
Going for the better meter, as I have friends and family that want a sat. now, so it should make the job easier.
Learning lots from you guys, keep it up lol.
thanks

 

[Lone Gunman]

The way I peak a satellite TP/channel is to bump East until I loose the picture, then go two or three clicks past that. At that point I start bumping the dish back West until the picture just stops pixilating.

At that point I write down the count number on the Vbox then I continue to bump West until I loose picture again then go about two or three clicks past that point. Then I start bumping the dish back East again until I get a picture that's not pixilating then write down that Vbox number. Subtract the small number from the large number then divide that number by two. Then add that number to the smallest count number and that's where you store that satellite position.

Keep in mind that when you bump until the picture goes away, that point is usually a few clicks past where it will normally lock the signal coming from the opposite direction. Do this a few times and you'll see what I mean.

Hope this all makes sense cuz I got confused writing it!

 

[updatelee]

Its really too bad that we have this method of measuring signal quality and signal strength and yet pretty much no one uses them.

db for SNR
dbm for signal strength

UDL

 

[Babadem]

Each satellite receiver brand and model displays a different quality reading, so it is impossible to know what the real signal integrity is. Some STBs display 100% quality even on poor signals, while other models do not go above 60 or 70%.

Your receiver's display may max out in the mid 70's and you may be getting the best signal quality possible. Without an installation meter (or the rare STB) with a BER and CN readout you may never know if your system signal quality is maximized.

The signal quality reading on your receiver is likely a post FEC measurment, so error correction is being applied to any signal packet loss. Small LNBF placement or rotation settings and dish movements are not reflected in the quality reading, but it might be making the percentage of good signal vs noise much better. With a good meter, you will observe changes in the CN (Carrier to Noise) and BER (Bit Error Rate) readings with even a fraction of a degree rotation of the LNBF or sliding the LNBF foreward or aft.

What meter did you order?

So the BER should be "zero" or close to it if you have your settings/dish optimized? Further explanation will be highly appreciated.

 

[updatelee]

BER = bit error rate. errors are bad, so yes. 0 errors is what you want.

every error is a pixelated screen, so, more errors, the more unwatchable it is.

UDL

 

[Titanium]

BER measurements consist of two variables the number of errors and the bits of data sampled. BER readings look like this: 5E7 (5 is the number of errors and 7 is the number of data bit samples)

1E4 = 1 error for every 10,000 bits of data sampled - Good
1E8 = 1 error for every 100,000,000 bits of data sampled - Very Good

On the other end of the scale:
9E8 = 9 errors for every 100,000,000 bits of data sampled - Very Bad
9E2 = 9 errors for every 100 bits of data sampled - Extremely Bad

 

[Babadem]

BER measurements consist of two variables the number of errors and the bits of data sampled. BER readings look like this: 5E7 (5 is the number of errors and 7 is the number of data bit samples)

1E4 = 1 error for every 10,000 bits of data sampled - Good
1E8 = 1 error for every 100,000,000 bits of data sampled - Very Good

On the other end of the scale:
9E8 = 9 errors for every 100,000,000 bits of data sampled - Very Bad
9E2 = 9 errors for every 100 bits of data sampled - Extremely Bad

Thanks, for educating me on this subject.

 

[updatelee]

Just to add more background, 9e-8 is scientific notation

9e-8 = 0.00000009

0.00000009 = 9/100000000

so easy way to remember, 9e-8 = 9/1 with (8 zeros)

so 9e-8 = 9 errors in 100,000,000 bits

UDL

 

[Babadem]

BER = bit error rate. errors are bad, so yes. 0 errors is what you want.

every error is a pixelated screen, so, more errors, the more unwatchable it is.

UDL

Thanks, for your insight.

 

[climy]

The Reel Multimedia, Reelbox has Signal strength (STR), S/N, BER and UNC. On Ku it will go to near 100 STR and SNR. On the 1.8m (6ft) usually won't measure above 78 STR and 52 SNR. as long as BER is (0) zero, I get a good picture. BER is attempting to "correct" the image as long as there is enough information to do so. any UNC, and you can see it in the image.

 

[updatelee]

Thats defn not SNR, SNR is measured in db. Thats clearly just an arbitrary %

UDL

 

[climy]

Indeed. There is a data sheet for my tuner. STR (Strength): Gives signal strength as a percentage. A minimum level of 60% is required to ensure a quality signal.


Signal Values Explained:
STR (Strength): Gives signal strength as a percentage. A minimum level of 60% is required to ensure a quality signal.
SNR (Signal-noise-ratio): Gives signal noise ratio as a percentage. Providing the signal strength is at least 60% (STR, upper LED) the
signal noise ratio should not cause a problem.
BER (Bit error rate): Indicates whether there is a problem with the chosen data stream. Optimum performance is indicated by “0000“,
and the system can cope with a data stream issue with a BER of up to three. If the display shows a higher BER than three, the reception
can be improved by changing the satellite position or by checking the cable connections.
UNC (Uncorrected blocks): Indicates if there is a data stream issue which cannot be repaired. Optimum performance is indicated by
“0000”, and the reception can be improved by changing the satellite position or by checking the cable connections.
Video: Indicates the reception bit rate for the chosen video data stream (in Megabit/s). This information is provided by the available
channels and cannot be changed by the ReelBox user.
Audio: Indicates the reception bit rate for the chosen audio data stream (in Megabit/s). This information is provided by the available
channels and cannot be changed by the ReelBox user.
LOCK: If highlighted in yellow the reception is within acceptable parameters. If highlighted in red no reception is available.
SIGNAL: If highlighted in yellow, a signal strength of more than 50% is available. If highlighted in red a signal strength of less than
50% is available.
CARRIER: If highlighted in yellow a DVB carrier signal is available. If highlighted in red no DVB carrier signal is available.
VITERBI: If highlighted in yellow the repair function is active and working correctly. If highlighted in red the repair function is not
active.
SYNC: If highlighted in yellow signal synchronisation bytes have been identi? ed by the system. If highlighted in red no signal synchro-
nisation bytes have ben identi? ed by the system.
The symbols LOCK, SIGNAL, CARRIER, VITERBI and SYNC must be highlighted in yellow, to ensure there is no problem
with reception.

 

[updatelee]

How do you convert db to % ?

DVB-S QPSK fec 1/2 min db is 2.7db
DVB-S2 8PSK fec 9/10 min db is 11db
DVB-S2 32APSK fec 9/10 min db is 18.10db

how do you convert that to % ? you have a lookup table for every modulation/system/fec and compare it to min db ? There is no max db, so where does the scale end to determine % ?

% is retarded, there is no other way around it. % gives no usable information.

Same thing can be said for signal strength, every tuner will obv have internal noise, so a min dbm can be determined, and every tuner frontend will get overloaded at a specific point. You think they actually measured those values to determine the failure point to make a scale for % ?

doubt it

UDL

 

[Titanium]

Fun looking display, but not providing the calibrated measurements that we are discussing. These are not actual values or displayed in correct expressions. This banner provides additional information about the signal, but once again, it is the manufacture's own scale and definition, not providing useful data that could be meaningfully compared with instruments displaying calibrated data or with other brands or models.


Brian Gohl
Titanium Satellite

 

[climy]

The original poster never mentioned signal to noise in db and in that spirit, we are talking signal quality in percent which assumes full scale is unity or 100%. I never claimed any receiver measured db. updatelee just lamented the fact saying that db is the only way to measure s/n. true but no FTA receiver on the market would bother with the cost of calibrating this.
Climy.

 

[updatelee]

When I talk about measuring things the first thing that comes to mind is quantifiable measurements. I measure my height in feet and my weight in pounds... neither is %

As I pointed out before, what is 100% SNR ? there is no 100%, 0% is failure I presume, but lots of receivers fail long before their scale hits 0%

Pretty much all tuner/demods are calibrated from the IC's factory. Every tuner/demod Ive written drivers for are already calibrated. If a developer chooses to ignore those values and make their own, thats their prerogative, but the info is already there. They just chose not to use it.

Id love to know what tuner/demod your using that actually outputs the values as a % though, it would be a first.

UDL

 

[climy]

The STB supplier calculates percent. it's a unit less quantity. and so is db. however you can convert db to percent all day.
http://www.lightmachinery.com/percent-to-dB-conversion-calculator.php

from the manufacturer data on the tuner:
STR (Strength): Gives signal strength as a percentage. A minimum level of 60% is required to ensure a quality signal.
SNR (Signal-noise-ratio): Gives signal noise ratio as a percentage. Providing the signal strength is at least 60% (STR, upper LED) the
signal noise ratio should not cause a problem.