Some thoughts on the differences between eyepiece and electronically-assisted observing

[2STAR]

On ‎19‎/‎02‎/‎2016 at 22:22, Martin Meredith said:

Having spent the best part of the last two years observing electronically (EAA), it was interesting to go back to pure eyepiece observing (EO) for a session tonight. It wasn't the first time in that period I've broken out the eyepieces (although nowadays I mainly use binocs) but it was the first time I've sat down and thought about just what it is that makes EO and EAA different. Nothing I have to say about EO is novel, but I found the contrast with EAA illuminating and thought-provoking, and I'd be keen to hear others' experience.

I used my StellarVue Nighthawk, an 80mm f6 achromat, sitting atop a camera tripod, mainly observing the open clusters M41, M46 and M47, and the Orion Nebula M42. M46 and M42 were the first objects I looked at electronically, and with the same scope, using the Lodestar-C a couple of years ago. The images I took that night are still fresh in my mind, having looked at them a few times since. And that's one aspect of EAA that I like: the experience is not as ephemeral as eyepiece viewing!

Some observations on eyepiece versus electronically-assisted observing

One of the main pleasures of EO is the ability to get objects to ping into sharp focus, something that takes longer with EAA, although the use of a Bhatinov mask has speeded up that aspect a lot. And once in focus, seeing a cluster of sharply-defined stars is something where, in my experience, EO beats EAA most of the time. I say most because on occasion I've managed to get sharp views of open clusters with EAA using a Newt (never with the achromat). Double stars, such as the pair just south of the centre of M47, were very clean visually, with none of the fuzziness that I sometimes see in my EAA images. Colour too was subtle but evident, and never garish as it can be in EAA. In EO, colour can be described as stable and well-controlled, unlike the case in EAA where (I find) one can never really be sure whether the colours are 'real', whatever that means, or appropriately saturated.

The eyepiece view also seemed well-controlled in the sense of an absence of gradients unless one is close to a bright star or the moon (which was very bright tonight). In contrast, in EAA there is frequently some kind of gradient or other optical effect such as vignetting that is apparent in the captured image. Likewise, the eye doesn't suffer from hot pixels, dust bunnies and the like, although it is not defect-free...

I noticed no distracting effects of unfocused stars at the periphery of vision, nor other artefacts. They were there (this is by no means a well-corrected scope), but easy to ignore. In EAA, we are served up the entire image all at once, warts and all, and it is easy to spot flared stars due to miscollimation, coma and tracking errors.

In spite of the well-controlledness of the eyepiece view, the scene is alive, with details coming (and going) continuously. I suppose that in particular for open clusters there are nearly always stars at the edge of detectability, and these appear to scintillate as a result of factors such as seeing and dark adaptation. Apart from very short exposures of the kind I use when focusing and framing, this liveness is largely missing in EAA. 

In short, I'd characterise the eyepiece experience as immediate, stable-but-interestingly-live, nicely-subdued, and lacking distracting artefacts. 

But all this comes at a price: depth. As I struggled to make out more than a dozen of the mag 10-13 stars in M46, I couldn't help but recall the EAA view with the same scope, which served up a field of 100s of stars after 15s or so, together with a very clear view of the planetary nebula NGC 2438. As I examined M42 (beautiful and detailed in the eyepiece in spite of the proximity of the moon), I was thinking of how much greater it is in extent than the portion I was seeing. And also thinking of the Horsehead not too far away, an easy catch for EAA. And while scanning the starfields, I remembered all the galaxies that pop up in unexpected places in a typical EAA capture, and the near-futility of looking for any but the brightest on a moonlit night with this scope.

There were a few other differences too. I was hugely-affected visually by the moon in terms of dark adaptation; I was never quite as comfortable in my observing position as I would be for EAA (that is, each new object demanded a slight change in position); having to focus on the atlas and refocus at the eyepiece was an issue that goes away in EAA. On the other hand, any lack of damping of the mount after moving to a new field didn't matter for EO as the field of view was typically much wider and the effects less noticeable than for EAA; magnification changes were easily accomplished in EO.

What this experience brought home to me really forcefully is that EO and EAA are very different beasts, almost not recognisably part of the same pursuit! For any given sized scope, target DSOs for EO and EAA might hardly overlap at all; the physical experience of observing in the two modes is very different (apart from the chilliness, which for me at least is a common factor ;-); the nature of the visual information presented to the eyes is barely comparable. A positive way to put it is that the EAA and EO approaches to astronomy are almost entirely complementary, which has to be a good thing.

Why does the visual system sometimes win?

This post is already quite long, but I wanted to raise one other musing: what it is in our supposedly-impoverished eye-brain combination that nevertheless can deliver so well at the eyepiece in some of the ways I outlined above. 

In EO, supposing focus is good, are we less affected by the poorer quality image off centre because we simply direct attention to one place at a time?

Is the eye mainly responding to the upper parts of the point spread function so that stars appear sharper (we don't see the fuzzy edges as we do in captured images)?

Do the visual pathways put a premium on rather heavy-handed noise reduction, so we are never really aware of photon noise (perhaps one reason why we the limiting magnitude is quite low)?

Does the less-sensitive/noise-suppressing optical chain mean that we are simply less able to detect the gradients that are fairly common in images?

How important is the slightly-irregular layout of the visual receptors compared to that of a CCD?

Can we learn anything from human visual processing that could be used to make the EAA experience more EO-like -- at the flick of a switch?

If you got as far as here, thanks for reading!

Martin

Nice report Martin, thanks

Eric

[Astrojedi]

6 hours ago, Martin Meredith said:

Yes, I mean being able to capture the full eyepiece experience for the kinds of objects where we all agree the visual experience is best: double stars, open clusters, etc

From the responses so far it seems there is some optimism that it will be possible. I think so too, but not necessarily in the next decade or two. There's quite a bit of research  on developing silicon retinas, though much of the focus is on encoding the time dimension, which perhaps isn't so critical for us. I'd like to see whether the non-uniform grid of photoreceptive cells in the human optical system makes a difference (and whether it could be simulated using some kind of dither-like processing), and what noise-reduction processes the human/mammal system uses to deliver at all times a clean 'image' at the cost of sensitivity. The things we enjoy about looking through an eyepiece are perhaps just the result of this kind of tradeoff: sharp stars, little noise, but very low gain? So perhaps to simulate the eyepiece view it will be necessary to turn on their heads some of the design decisions used in CCDs and subsequent processing? But I imagine the eye/brain system is much more active/dynamic, sparse, and focusing on encoding change. As observers looking through eyepieces we are really just piggy-backing on whatever evolution has delivered for other purposes...

Martin 

Martin,

Very interesting discussion. Actually the temporal aspect is critical to perception (and to reality and existence but that is the topic for another thread)

If you haven't already read it I would highly recommend you read Jeff Hawkins 'On Intelligence'. I think you will really enjoy it.

[Martin Meredith]

1 hour ago, Astrojedi said:

Martin,

Very interesting discussion. Actually the temporal aspect is critical to perception (and to reality and existence but that is the topic for another thread)

If you haven't already read it I would highly recommend you read Jeff Hawkins 'On Intelligence'. I think you will really enjoy it.

BTW In case it wasn't clear about the temporal aspect, by 'us' I meant the EAA community as opposed to humans in general ;-)

Thanks for the Hawkins recommendation. I'll look out for it. A bit old now, but Dennett on Consciousness is also worth a read, particularly concerning the way we construct stable percepts out of sensory data. 

Until we understand more about what goes on in vision past the retina, I doubt we'll be able to properly simulate the eyepiece view, which is why I think it will take some decades yet. A bit of an aside, but just to demonstrate how complex things can get in vision, here's a nice demo of the way hearing sounds affects what we see. The visual information is the same in each case but gets reinterpreted depending on the number of beeps presented.

Martin

[aparker]

3 hours ago, Astrojedi said:

Contrary to some of the thoughts here I think our eyes are very poor detectors for astronomical observations. Our eyes evolved to adapt to our environment on earth and hence the visible spectrum was limited to the Sun's radiation spectrum. So we cannot see infrared, ultraviolet, X-ray or gamma rays or other radiation.

 

Hence we build telescopes!  

To give evolution some credit, there's only so much you can do with amino acids, fats, and water as the engineering materials.  Get too much shorter than the bluest light we can see, and the photons are energetic enough that they ionize tissue instead of just tickling your detectors (fortunately our ozone layer screens these UV rays out).  As light gets longer into the infrared, we quickly get into the domain where individual photons don't have enough energy to knock electrons off of biological pigments (the way our eyes detect "visible" light).  The animals (like pit vipers) that do have sensitive IR detectors use an entirely different mechanism than how the eyes work.

[aparker]

1 hour ago, Martin Meredith said:

BTW In case it wasn't clear about the temporal aspect, by 'us' I meant the EAA community as opposed to humans in general ;-)

Thanks for the Hawkins recommendation. I'll look out for it. A bit old now, but Dennett on Consciousness is also worth a read, particularly concerning the way we construct stable percepts out of sensory data. 

Until we understand more about what goes on in vision past the retina, I doubt we'll be able to properly simulate the eyepiece view, which is why I think it will take some decades yet. A bit of an aside, but just to demonstrate how complex things can get in vision, here's a nice demo of the way hearing sounds affects what we see. The visual information is the same in each case but gets reinterpreted depending on the number of beeps presented.

Martin

+1 for Dennett's book Consciousness Explained.  Also interesting on this (non-astronomy) vein are Daniel Wegner's The Illusion of Conscious Will and Thinking, Fast and Slow by Daniel Kahneman

[aparker]

3 hours ago, Astrojedi said:

And a myriad of other shortcomings which have led to various visual observing requirements like the need to use high magnification.

 

 

The main shortcoming our visual system has for astronomy purposes is the maximum aperture of 7mm (normal adult pupil).  This radically limits light gathering, and also sets a fairly low ceiling on angular resolution.  The answer to both these problems is to

 

build a telescope!

[Astrojedi]

I think that is exactly the problem. You need larger and larger telescopes to overcome the shortcomings of our visual system.

This is proven out by the fact that with my 8" and a highly sensitive CCD at my light polluted location I can easily exceed what you can see visually with a 20" scope at a dark site.

No doubt that larger telescopes will gather more light and go deeper but due to the shortcomings of our visual system we are wasting most of those photons when observing visually. This is what we would fix with an electronic eyepiece. Make the experience significantly more efficient.