Why dont we...

[Alien 13]

It might be a stupid question but why dont the manufacturers make astro CCDs or CMOS cameras with a round sensor or for that matter telescopes with square lenses? The round sensor would make a lot of sense and enable you to choose which part of the image you display on screen. I suppose another option would be circular displays too.

Alan

[Ouroboros]

 it's a lovely question. 

[Thalestris24]

I don't think anyone makes image sensors for the amateur astro market. They are all second use to the non astro camera market. In any case I don't think round sensors would be practical to make. We'll have to make do!

Louise 

[Alien 13]

3 minutes ago, Thalestris24 said:

I don't think anyone makes image sensors for the amateur astro market. They are all second use to the non astro camera market. In any case I don't think round sensors would be practical to make. We'll have to make do!

Louise 

I wonder if this is true, most imaging systems using lenses be it astro/industrial/microscopy produce circular images.

Alan

[steppenwolf]

I suspect that it is all about the method of 'reading' the data on the sensor - this is (essentially) a column and row process - the columns and rows of a 'circular' sensor will not be continuous.

[andrew s]

I think it is for two reasons firstly it is much simpler to make round optics due to the symmetry involved. Secondly while most crystals grown to make silicon chips are circular in cross section most displays are square so they make the sensors two match. In addition it is simpler to cut rectangular silicon chips rather than circular.

Regards Andrew

[Alien 13]

Thats a good point Steve but it should be possible to map a sensor from the center to the angular circumference. I wonder if it a historical thing in that paintings represented the world as a rectangle although we do see it in circles.

Alan 

[Thalestris24]

Plus the way chips are made is really limited to rectangular pixels. It's the most efficient way of manufacture, I should think.

[barkis]

Are there any Space borne cameras doing deep space imaging that have circular sensors?  If it was of a practical benefit,  then the Idea might be already in use now.   Never read of any,  but asking the question of the right organization  may yield an answer. Whatever that might be.   Nasa  are always willing to give out Info. 

[Ouroboros]

It's worth pointing out that a square optic does not give a square image. 

[Stub Mandrel]

I don't know about round sensors but hexagonal pixels would make a lot of sense

[RayD]

We'd have to find round computer monitors then to look at our images in full screen 

[ronin]

As mentioned it is the row and columns that the data is pulled off in, possibly more relevant is that the software will be a nested loop

For (Col = 0 to Col =25) Do

For (Row = 0 to Row = 25) Do

Read C(n), R(n)

Sort of thing, the Read is a Shift process. Writing the software for a circular array would be somewhat more complicated, also you cannot shift then some of the pixels, they would have nothing to shift into. The sensor on the wafer would still be cut square also, so a circular sensor on a square chip is wasting silicon

 

[michael8554]

Remember Cathode Ray Tubes? They weren't always rectangular.

They started out having round faceplates, and I believe the Baird mechanical TV system had a circular image.

The electronic system that was eventually adopted had a rectangular display on the circular faceplate, it was easier to generate repetitive line and field scans. 

Michael

[Alien 13]

I do feel it might be conditioning that our brains like a rectangular TV or monitor screen but as said it wasn't always like that, we had round CRTs (developed from RADAR tech) and even the early Magic Lantern projectors produced round images. I suppose my point was that we have expensive optics and guided mounts that produce a nice round image circle then use a rectangular chip that effectively throws 50% of the image away.

Alan

[Filroden]

Remember that the field of view for a pair of eyes equate closer to a rectangle than a circle though the binocular element within it would be closer to a circle. We would be wasting a significant amount of vision by using circular screens.

[LukeSkywatcher]

10 hours ago, ronin said:

As mentioned it is the row and columns that the data is pulled off in, possibly more relevant is that the software will be a nested loop

For (Col = 0 to Col =25) Do

For (Row = 0 to Row = 25) Do

Read C(n), R(n)

Sort of thing, the Read is a Shift process. Writing the software for a circular array would be somewhat more complicated, also you cannot shift then some of the pixels, they would have nothing to shift into. The sensor on the wafer would still be cut square also, so a circular sensor on a square chip is wasting silicon

 

*shudders at the flashback of studying VB in college*.

(VB = Visual Basic..........not Victoria Bitter)

[ollypenrice]

While film and digital plates have always been recitlinear in format, I'm not sure that they have always been flat. Didn't the first RCs use curved plates?

Using this method you could make a widefiled all-reflecting system since the flatttening would be physical rather than lens-based. Hmmm... more expense!

Olly

[Davey-T]

The Sloan Digital Sky Survey turns out round plates, don't know how they capture them.

Dave

[Nova2000]

Good question. I was thinking of this. I think one cannot make curved pixels!? 

[andrew s]

14 hours ago, Stub Mandrel said:

I don't know about round sensors but hexagonal pixels would make a lot of sense

[Steve Ward]

37 minutes ago, Davey-T said:

The Sloan Digital Sky Survey turns out round plates, don't know how they capture them.

Dave

http://www.sdss.org/instruments/camera/

[andrew s]

Oops sorry about that seem to have got in a mess with the editor.

What I intended to say was that you can make whatever shape pixels you like using lithography. The issue is how to add the voltages to create the potential well to trap and then move the photo electrons this is most simply done on a rectangular array.

Regards Andrew

[Davey-T]

5 minutes ago, Steve Ward said:

http://www.sdss.org/instruments/camera/

Blimey  and I have problems with one camera,

Dave

[Stub Mandrel]

I suspect the truth is that the practical and cost effective way to make a round sensor that works with existing technology is to put a round mask over a rectangular sensor...