Image Scale Confusion

[Northernlight]

Hi All,

I need some help to get my head around image scale. I currently have a 102mm refractor with a 0.8x reducer (approx 571mm F/L) which with my camera gives aound 1.94 Images scale. I'm looking to purchase a MN190 which at nearly the focal length (1000mm) gives me a image scale of approx 1.05 with my ccd.

So essentially, i'm nearly doubling the scope F/L, but at the same time i'm pretty much halving my image scales.  So in my simple mind, does this mean that when i view images on my pc, they will appear approximately the same size, i.e. if i'm imaging a galaxy will it apper to be a larger / zoomed in on the MN190 vs the refractor or will they look about the same ?

NL

[Thalestris24]

Hi

I think it's really your field of view which will determine how your images will look. You can put your scope/camera details into a fov calculator to get a preview of a variety of targets: http://www.12dstring.me.uk/fovcalc.php

Hope that helps

Louise

[Oddsocks]

Once you have calculated the pixel scale for a particular camera/reducer/telescope combo it is not so complicated, you just to need to think about how big the object is that you are imaging in relationship to the arc sec/pixel resolution of your system.

In your first example, the 102mm refractor with reducer gives you an image scale of 1.94 arc sec/pixel, now say you want to image the crab nebula and we know its angular size in the sky is approx 6x4 arc mins or 360x240 arc seconds, in this case dividing 360x240 arc seconds by 1.94 arc sec/pixel gives you the number of pixels that the image will occupy on the camera chip, ~185 pixels by ~123 pixels.

Now go to your second example with the MN1900 and an image scale of 1.05 arc sec/pixel, using the same example of the crab nebula the number of pixels occupied by the crab will be 360x240 arc seconds divided by 1.05 arc seconds per pixel so the image will occupy ~342 pixels by ~228 pixels.

If your monitor is set to display the image at a zoom ratio of 1 to 1, one camera pixel equals one monitor pixel then yes, the crab image from the MN1900 will be twice the size on screen as the image from the 102mm refractor.

But here is the fly in the ointment, remember your monitor display can be zoomed up or down. Say we tell the computer to zoom up the display by a factor of two so that one pixel of the native image is displayed over two pixels of the monitor, you could take the image of the crab from the 102mm refractor only now it will occupy ~370 pixels x ~ 246 on the monitor, in other words it will appear exactly the same size on screen as the image from the MN1900 does when the monitor is set to a zoom ratio of 1 to 1.

Confused?...I think I would be by now!!!

For this reason it is not very useful to think about apparent size on screen since we can fiddle about with the zoom ratio of the monitor and upscale/downscale our images to suit our whim.

It is more useful to think about how the image scale of our system will suit the object we want to image, will the object fit comfortably on our CCD chip, we know the arc sec/pixel of our imaging system and we know how many pixels make up our sensor, finally we know the size of the object from looking it up in a planetarium program or on-line etc so if the object is too large to fit the sensor we can swap to a shorter length telescope or a higher ratio reducer, if the object is tiny only occupying a few pixels then we can remove the reducer or use a longer focal length telescope.

The final part of the jigsaw is to think about resolution, assuming perfect seeing conditions and a perfect mount that is tracking and guiding perfectly, when we image with the MN1900 at 1.05 arc sec/pixel the tiniest detail we can resolve in the image is 1.05 arc seconds wide, imagine two small stars separated by a gap of 1.5 arc seconds, in the MN1900 system you should be able to see the two stars, each occupying a small group of pixels with a dark gap one pixel wide between them, you can say that you have been able to resolve the stars into two individual entities.

Now look at the same system with the 102mm refractor at a pixel scale of 1.94 arc sec/pixel, look in the gap between the two stars, it was only 1.5 arc seconds wide but the single pixel scale of our camera is wider than the gap and is capturing the edges of both stars so now the gap will disappear and in our image we will just see a single elongated, rugby ball shaped star, we can say we could not resolve the stars into individuals.

A bit of a long winded explanation I'm afraid, really should cut down of the caffeine just before bed time.

So don't get hung up thinking about magnification, think about how the object will fit on the camera chip, making the most use of those expensive pixels and think about the resolution of the system, how well can your mount track and guide and what is the seeing like and then choose the telescope focal length and pixel scale to suit those conditions. After that you can zoom the image size on your monitor or upscale/downscale the image in Photoshop to suit your print size etc...

 

[Northernlight]

Oddsocks,

Thankyou for taking the time to write such a helpful explanantion. I've read many different things online about pixel scale, all of which were too technical and the info started to bet lost in transaltion as it were.

Your simple to explanation has help me to finally get a better understanding, and in a simplistic view, i now understand that under ideal conditions a lower pixel scale should in theory yield a higher resolution image, and with a longer focal lenghth having a smaller field of view the target should essentially cover more of the ccd sensor, so at a 1:1 relationship the target in theory should appear larger and when viewing on a screen should allow me to zoom in a bit and see more detail.

thank you very much, ad once again thanks for taking the time & effort to explain.

NL

[Northernlight]

Louise, many thanks for the link :-)

 

NL

[ollypenrice]

Let's try a simplified agreement with Oddsocks.

1) How many pixels do you put on the target? If you double the focal length you'll put 4x as many on the target. (Twice as many wide, twice as many high.)

1b) This will make your full size screen image of the object 4x bigger by area, 2x wider and 2x higher. As Oddsocks says, this means 1 cam pixel for 1 screen pixel, which is the only way a gentleman ever looks at an image! (Joking, but we need to compare like with like.)

2) Is there any real extra detail in the image? (In other words, does the long FL image look better than  the short FL image over-magnified on the screen till it's the same size as the long FL one? If you have the seeing stability and the guiding precision necessary, then it will look better. If your guiding and your seeing don't allow the long FL to strut its stuff then you won't.

I would not hesitate to go after an arcsecond per pixel. To go after less than that is getting very serious.

Olly