The Road to Better Data

In several articles, I have discussed things we can do to improve the quality of our results. The last article on this site that was co-written with my friend Stephen Dow dissected a lot of things unique to the Celestron Origin, but all of which could be applied to any fixed data collection system, where you conduct experiments with the intent of collecting information to tweak a system.

Let’s talk about signal to noise ratio and what exactly this means and why it matters. Signal to noise ratio is not unique to astronomy. It applies to electronics, in both audio and broadcast equipment and more. If a broadcast engineer wants to optimize his transmission pattern, he will do things like tweak cabling to reduce signal loss and use an antenna design that increases gain in a particular direction. All of that improves things without changing the power output.

The same thing applies to us in astronomy. We cannot really adjust the signal so to speak. A source only emits what it emits. When we speak in terms of signal strength such as the magnitude of an object, it is what it is. To increase the result of what we are trying to extract we can do a few things. We can increase the size of the aperture of our collection system (larger telescope), we can use a collector with higher sensitivity (better camera), we can increase the duration of our collection period by changing the length of subs (single exposure), and we can stack those subs to provide more total signal in our collection (total exposure).

Stay with me here now. We did all of that without increasing the amount of signal being emitted and it improved things. In amateur radio, that means we do things like installing better cabling, putting our antenna up higher or even modifying our antenna. We do all of that because we are detecting faint signals from distant lands, and we want to be able to hear them. In astronomy, we are doing the same thing. I often tell my radio friends that we can use our legal limit of power output which might be 3 kilowatts, but if we can’t hear the other station that makes no sense.

The night sky is filled with sources of various signal strengths, and if we want to see them, we can’t increase the signal, so we work to reduce the noise. Take for example a hypothetical object that delivers an output of a fixed number of, let’s say 80. But we have a background that delivers an output of a fixed number of say, 4. We can compute the signal to noise ratio (SNR) using the equation

SNR= Signal Strength/Noise Strength

which in our case would be SNR=80/4 leaving the SNR to be 20. The higher the SNR, the better your data is.

And this is exactly what happens when you go to darker sites. You reduce the noise. It is that simple. Whether it is dark sky sites away from the city, or outside and away from electrical interference in radio, reducing the noise changes the signal to noise ratio. Let’s say we reduce the noise from 4 to 2

SNR=80/2 leaving the SNR to now be 40. We have doubled the signal to noise ratio from 20 to 40.

So why does that matter? Some say you can just collect more subs if your noise level is high. That is true – to a point. You will run into other limits. In radio that might mean using better notch filtering and digital signal processing to reduce the noise. Amateur astronomers use notch filtering too, like light pollution filters tweaked to a specific wavelength, and though it helps, it also has to somewhat remove some of the signal so it is a not a linear relationship.

In astronomy, it can be said that it is all about collecting better quality data because better quality data means less collection and processing in the end to make things great.

If you have any questions please feel free to reach out to me via the form at the bottom of the Hyperstar Image Gallery page and I will do my best to answer them.