What did you see in the sky tonight?

[Chris Peterson]

It does look more detailed and smoother, Nitpicker. Is there an adjustment for sharpness or deconvolution? That's way too high if so.

To be fair, we should distinguish between images with scientific intent and images with aesthetic intent. In the case of the former, it is frequently worthwhile to use wavelet processing or some other sharpening to enhance real detail, at the expense of creating (often) unaesthetic artifacts.

A technique I use very often is to create a highly sharpened image and composite it with the original, using Photoshop's layer blending (Lighter and Darker) and layer transparency. That often allows for bringing out edge detail without the edge artifacts that most sharpening algorithms produce.

[geckzilla]

Sure, sharpening is great. But this is a good example of too much. Halo artifact.

[Chris Peterson]

Sure, sharpening is great. But this is a good example of too much. Halo artifact.

nitpickermoon.jpg

That is only too much sharpening if your intent is aesthetic. If your intent is to identify subtle structure, however, it could use quite a bit more sharpening.

[geckzilla]

Maybe elsewhere in the image, but not where there is already too much sharpening. The dark rim around the base of the crater is a fake feature. I mask areas of already high contrast off when I process.

[stephen63]

Chris, when you say sharpening, are you referring to wavelet function or some sort of sharpening in Photoshop?

[Chris Peterson]

Maybe elsewhere in the image, but not where there is already too much sharpening. The dark rim around the base of the crater is a fake feature. I mask areas of already high contrast off when I process.

I disagree. Even this area could use more sharpening. The thing with sharpening for scientific intent is that you need to understand the sort of artifacts that different techniques produce. You clearly understand what some of those look like, and have identified them. That means you aren't mistaking an artifact for an actual structure. If you were trying to understand the geology (selenology?) of this area, you would be well advised to apply more sharpening, and different sorts (deconvolution, wavelets with different frequencies, etc), since this will bring out specific kinds of detail and also produce different kinds of artifacts. By careful examination, a skilled observer can separate these things and learn a lot more.

You should see some of the massively oversharpened images of nebulas that appear in the literature. Ugly as sin, but critical to understanding things like shock front geometry.

[Chris Peterson]

Chris, when you say sharpening, are you referring to wavelet function or some sort of sharpening in Photoshop?

I'm using the term to refer to any sharpening function, which could include deconvolution, manipulation in frequency space, unsharp masking, wavelet processing- anything at all.

[stephen63]

In the case of processing for aesthetic images, there really is no set rule of thumb for the amount of sharpening or anything else. What I mean by this is that how you manipulate an image depends a LOT on what object you are processing. I'm sure it's no one's goal to overcook an image, but that's what happens when too much sharpening/color enhancement is used.

[geckzilla]

Alright, I confess I have no idea what scientific sharpening is. I always figured you'd want to avoid introducing fake details, aesthetic or for science.

[Chris Peterson]

Alright, I confess I have no idea what scientific sharpening is. I always figured you'd want to avoid introducing fake details, aesthetic or for science.

All processing can add artifacts. When you're evaluating an image for scientific purposes, the goal is to distinguish what is an artifact from what is real, not necessarily to prevent artifacts, which you can't always do and still achieve the processing goals.

[Ann]

This time of year, the skies over southernmost Sweden are either murky and foggy or clear and bitingly cold. The weather here has been mild, so our skies have been relentlessly overcast. But yesterday I saw Venus (yay!) and the Summer Triangle. I could see that Vega looked very subtly blue and that Altair looked white with no hint of blue whatsoever. Deneb looked so-so, maybe just a little blue and maybe not at all.

When I got up extremely early in the morning, fog was in process of conquering the sky. The full Moon shone brightly, surrounded by a huge halo. I saw the Winter Triangle too, and I could just barely make out Rigel. An hour later, even the Moon had been swallowed by the thick, dense fog.

Ann

[Nitpicker]

Thanks for everyone's feedback. It is very much appreciated. I have a lot of big new words to learn when it comes to image processing. It is the part of astrophotography that is most mysterious and unintuitive to me. At the moment, many of your comments still sound to me like you're all from another planet. I will have a good read of that tutorial from Paul Haese -- thanks muchly. To date, I've only watched a couple of pedestrian YouTube demos on Registax, which offer nothing more than exercises in button clicking. I suppose I'll have to install a copy of Photoshop or Gimp at some stage too.

Despite my lack of understanding at this early stage, it feels like I've just taken ownership of a new camera.

[geckzilla]

If you need any help with Photoshop, let me know. I was thinking about doing some tutorials but ehhh that's a lot of work! I will continue to think about it.

[Nitpicker]

This time of year, the skies over southernmost Sweden are either murky and foggy or clear and bitingly cold. The weather here has been mild, so our skies have been relentlessly overcast. But yesterday I saw Venus (yay!) and the Summer Triangle. I could see that Vega looked very subtly blue and that Altair looked white with no hint of blue whatsoever. Deneb looked so-so, maybe just a little blue and maybe not at all.

When I got up extremely early in the morning, fog was in process of conquering the sky. The full Moon shone brightly, surrounded by a huge halo. I saw the Winter Triangle too, and I could just barely make out Rigel. An hour later, even the Moon had been swallowed by the thick, dense fog.

Ann

Very confusing this Summer/Winter business. I've been watching our Summer Triangle within our Summer Circle here, quite different from your Summer Triangle, which I can't see at all at this time of year.

[Nitpicker]

I was recording a series of short 30s HD videos of Jupiter last midnight, in good seeing conditions, to see just how far I can push my current equipment. This small, two-frame animation is easily the best I've ever accomplished of Jupiter, thanks mainly to some basic stacking and sharpening in Registax:

Jupiter, with (L-R) Ganymede, Europa (occulted in frame 1) and Io, 20-Dec-2013 13:39 & 14:02 UT.

Obviously, this is nowhere near the quality we Starship passengers have become accustomed to, but I was pleased with it anyway. It was accomplished with positive lens projection through a 13mm eyepiece, pushing my little 6" f/10 SCT all the way up to about f/65, with a focal length of almost 10m. This allowed me to reach a reasonable pixel scale of about 0.23 arcsec/pixel, thus forming an image of Jupiter across 200 pixels. The pixel size on the 16MP sensor of my DSLR is an impressive 4.7 microns, but sadly, when using it as a video camera, the 2MP HD format still uses almost the whole sensor, meaning that the effective pixel size increases to 12 microns. This is why my focal ratio needs to be pushed so high.

Anyway, I've read that for good to excellent seeing conditions, one should be aiming for a pixel scale of 0.1 to 0.25 arcsec/pixel. I'm wondering if this is still a realistic target range for smaller scopes like mine, where the resolution (Rayleigh Criterion) is a modest 0.92 arcsec? With my 6" scope, could I really expect a much better image of Jupiter, if I used a video camera with smaller pixels, or would I also need to get a bigger scope to accompany such a camera?

I have been looking for a suitable video camera with small pixels, with a removable lens, and which is not too impossible to mount on a telescope without gaffa tape. Such cameras are not easy to find. But maybe the above quality is all I can expect with a 6" scope. Does anyone have any thoughts?

[Chris Peterson]

Anyway, I've read that for good to excellent seeing conditions, one should be aiming for a pixel scale of 0.1 to 0.25 arcsec/pixel. I'm wondering if this is still a realistic target range for smaller scopes like mine, where the resolution (Rayleigh Criterion) is a modest 0.92 arcsec? With my 6" scope, could I really expect a much better image of Jupiter, if I used a video camera with smaller pixels, or would I also need to get a bigger scope to accompany such a camera?

Non-deterministic sharpening methods like deconvolution perform better when the data is significantly oversampled. Eyepiece projection is generally a bad idea, introducing lots of aberrations. In your case, assuming a camera with 5um pixels (pretty typical), you could use a 3X barlow and image at 0.23/pixel, which would be near ideal.

I have been looking for a suitable video camera with small pixels, with a removable lens, and which is not too impossible to mount on a telescope without gaffa tape. Such cameras are not easy to find. But maybe the above quality is all I can expect with a 6" scope. Does anyone have any thoughts?

If you want to go cheap, select a webcam with good reviews and a CCD (not CMOS) detector. Then all you need is to modify it by removing the lens and attaching a standard 1.25" adapter. If you want better, I'd go with one of the Imaging Source USB cameras, one of the Point Grey cameras, or something similar. In general, I'd recommend a monochrome camera with an external filter if you want color, as that gives you a lot more control. But it adds to the cost.

[Nitpicker]

Thanks Chris. Despite being an expensive answer, it does make perfect sense.

How does one compose a sharp colour image of Jupiter using a monochrome camera and coloured filters? Sounds like the "too hard basket" for this lazy amateur.

[geckzilla]

I'm positive you already understand the concept of combining red, green, and blue light to create a color image so I'm guessing it's the image processing software part you don't feel like learning. Too bad, because once you do learn it's really fun. Even Hubble's pictures of Jupiter or Cassini's pictures of Saturn are a lot harder to put together due to all the moving they like to do, though. You'll end up with a little moon in three different places so there's a red dot, a green dot, and a blue dot for each exposure where they don't line up.

[Chris Peterson]

Thanks Chris. Despite being an expensive answer, it does make perfect sense.

How does one compose a sharp colour image of Jupiter using a monochrome camera and coloured filters? Sounds like the "too hard basket" for this lazy amateur.

There are all sorts of free or cheap image processing tools that let you combine color filter channels and generate color images. The challenge with Jupiter is its rapid rotation rate. You need to acquire all the images you want to stack and combine in just a few minutes, or you'll start getting registration errors with surface features. Luckily, Jupiter is very bright, so that's usually possible.

[Nitpicker]

I'm positive you already understand the concept of combining red, green, and blue light to create a color image so I'm guessing it's the image processing software part you don't feel like learning. Too bad, because once you do learn it's really fun. Even Hubble's pictures of Jupiter or Cassini's pictures of Saturn are a lot harder to put together due to all the moving they like to do, though. You'll end up with a little moon in three different places so there's a red dot, a green dot, and a blue dot for each exposure where they don't line up.

It is certainly true that I am yet to learn how to process images from separate LRGB channels. And thus far, I have only ever attempted colour astrophotography with a colour sensor. I do have a nice set of colour filters, still in very good condition, in the bottom of an old camera bag at the back of the cupboard. At this early stage in my journey into astrophotography, I figure I need all the light I can get.

But in this instance, I was referring more to the fast rotation rate of the gas giants (and the short orbital period of their moons). In order to get a sharp image of the clouds of Jupiter, one needs to stack an array of images all recorded within a very short time period (30-60 seconds, say). Having to break that into separate channels sounds "too hard" with only one scope and one camera. I know there are semi/fully automatic filter wheels available, but I also still want to purchase a piggy-back mount, a better video camera, a 3x-5x barlow lens, an equatorial wedge, a bigger scope, an observatory, a higher mountain, a spaceport, recycled ICBMs, etc, etc.

Even though I'm only just starting out on the image processing side of things, I'm already feeling that I don't enjoy it quite as much as actually being out under the sky at night. This may prove to be the real limiting factor of my attempts at astrophotography. It has to be fun, above all else.

[owlice]

Nitpicker, this is a very nice two-frame shot; thanks for sharing it!

[Chris Peterson]

Even though I'm only just starting out on the image processing side of things, I'm already feeling that I don't enjoy it quite as much as actually being out under the sky at night. This may prove to be the real limiting factor of my attempts at astrophotography. It has to be fun, above all else.

That, of course, is what ultimately determines whether someone is a serious visual astronomer or a serious imager.

[neufer]

---

Even though I'm only just starting out on the image processing side of things, I'm already feeling that I don't enjoy it quite as much as actually being out under the sky at night. This may prove to be the real limiting factor of my attempts at astrophotography. It has to be fun, above all else.

That, of course, is what ultimately determines whether someone is a serious visual astronomer or a serious imager.

[stephen63]

---

Even though I'm only just starting out on the image processing side of things, I'm already feeling that I don't enjoy it quite as much as actually being out under the sky at night. This may prove to be the real limiting factor of my attempts at astrophotography. It has to be fun, above all else.

That, of course, is what ultimately determines whether someone is a serious visual astronomer or a serious imager.

Siriusly, Neuf, was that necessary?

[neufer]

Siriusly, Neuf, was that necessary?

• O, reason not the need.