•

# Calculating arc sec per pixel with a OSC

3 replies to this topic

### #1 lampcord

lampcord

Vostok 1

• topic starter
• Posts: 141
• Joined: 26 Jun 2020

Posted 06 August 2020 - 04:04 PM

Looking at the calculation for arc sec / pixel:

(pixel size / focal length) * 206.3

It occured to me that for OSC cameras this may not be correct. Since each virtual pixel is represented by a 2x2 matrix of RGGB pixels, wouldn't that have to be taken into account?

In other words, I have a OSC with pixel size of 2.4u

Is my effective pixel size actually 4.8u because of how an OSC works?

And if so, isn't the actual calculation:

(pixel size * 2 / focal length) * 206.3

I'm sure I'm missing something but I can't seem to find clarification on this.

### #2 Alex McConahay

Alex McConahay

Voyager 1

• Posts: 11,006
• Joined: 11 Aug 2008
• Loc: Moreno Valley, CA

Posted 06 August 2020 - 04:10 PM

Your resolution is precisely the same whether you have a Bayer Matrix in front of the sensor array or not.

What is different is the color values. For any given pixel, your color value is determined by a combination of the particular pixel and every neighbor of that pixel.

So, is this different resolution or not? After you define "resolution," you can make the call as to whether your resolution is any different.

Seems to me that if you have the same number of pixels across and up and down, covering the same target area, you have the same resolution.

You just don't have completely accurate color values for the individual pixels.

One other thing.....it is not a 2x2 matrix. It is the individual pixel comparing its value to all the pixels around it.

Alex

• Dynan and lampcord like this

### #3 jdupton

jdupton

Soyuz

• Posts: 3,959
• Joined: 21 Nov 2010
• Loc: Central Texas, USA

Posted 06 August 2020 - 04:12 PM

lampcord,

I would say, oddly enough, that it depends on how you deBayer the image.

If you deBayer using a SuperPixel method, yes, your final plate scale looks like pixels are equivalently 2x in size.

If you "deBayer" using CFA Drizzle, your final plate scale looks like pixels are 1x original size.

If you use another deBayer method like VNG, Bi-linear, or other, the final plate scale is the same as the underlying sensor (1x) but some color resolution is lost.

John

Edited by jdupton, 06 August 2020 - 04:15 PM.

• lampcord likes this

### #4 lampcord

lampcord

Vostok 1

• topic starter
• Posts: 141
• Joined: 26 Jun 2020

Posted 06 August 2020 - 04:14 PM

Thank you both for the quick replies!

I have been using VNG since that is the default in PixInsight so I guess in that case the original formula holds.

• jdupton likes this

## Recent Topics

 Cloudy Nights LLC Cloudy Nights Sponsor: Astronomics