Undersampling - Oversampling

Bang on Beacon and you identify precisely where and how I deviated from the core point. My last but one post did indeed stray into confusing frequency space with what does it take to sample what represents a point in an image.

Your clarification is correct.
There is considerable confusion as to what the criteria really are and why. I for one (and I think Dave too) would rather have a sound explanation on a logical basis rather than a rule of thumb.

Magic numbers like 3 and 3.3 have no explained basis (that I can find). We may now be able to justify the 3.

Again if you base a decision on Anthony's advice you wont go far wrong.

I'm glad your post did spark of a good discussion.

Mark

I asked the CCD-tech community about this and one of the contributors to that forum sent me the link below. I haven't studied it in detail yet but it seems to answer a lot of our questions, clarify some points and correct some of my own errors. One great thing about the piece is that Stan backs it up with data.

www.stanmooreastro.com/pixel_size.htm

Dave

Hi Dave,

Half way through reading that piece I was beginning to regret starting it......but kept at it and in fact saved it now because I know I will read it again - and perhaps understand it better some day. But more to the point - bottom line is indeed the bottom line and a very important piece of information for me. (Thus a pixel size of 0.5 to 1.5 arcsec can be considered ideal for optimizing resolution.)

Anthony indeed had a good handle on this with his present 0.58"/pixel.....hope he is not disappointed at 0.38"/pixel.

Thanks everyone - this has been an 'eye opener'.

Brendan.

Brendan,

Thank you as well for starting this thread since such discussions are always welcome and educational for everyone involved. Personally, I now see where the "3.3x" comes from and thanks to the diagonal of the (square) pixel.

I am sure at 0.38"/pixel I will be fine since I do have the generally good seeing to pursue such an aggressive image scale. At the same time, I will be able to go even lower for solar system work and around the 0.10-0.12"/pixel which is the de facto norm for such work.

Anthony.

Indeed Brendan - I think we all learned from your post! I guess that's what the forum is for

It's been a great thread - thanks to all - thanks Dave for the reference from CDTech communinty - will read that with glee.

Anthony - I'm not as convinced as I was about the importance of the diagonal and the 2.8 rounding to often cited 3 as a result of the diagonal might just be coincidental. More to learn...

I still have questions and thoughts but I think time spent sourcing some good references are in order for me.

Mark