• UPDATE



    The forum will be moving to a new domain in the near future (canonrumorsforum.com). I have turned off "read-only", but I will only leave the two forum nodes you see active for the time being.

    I don't know at this time how quickly the change will happen, but that will move at a good pace I am sure.

    ------------------------------------------------------------

Square or cross shaped sensors

Moving this topic from another thread since it was off topic.

Slyham said:
This conversation brings up a question I have had for a while. Why not have a square or a cross shaped sensor?

Let me explain. Say you are camping near a lake. First you take a picture of the lake during a sunrise. You set the camera to landscape at 16:9, or whatever you want, and only the pixels in that ratio is used in taking the shot. Next your child wakes up and is sticking her head out the tent door. So you switch to portrait 4:5 and you can hold the camera in the same comfortable position as you do in landscape and take the picture. The day goes on with different shots with different ratios.

Obviously this would only work with a mirrorless camera and you could not have a lens hood with pedals.

I think some of the advantages would be you get to hold the camera "normally" for portrait shots, you only use the pixels you want (thus keeping the size of files to a minimum), there is less cropping in post, etc.

What do you all think? Stupid idea or does it have some merit?

dak723 said:
The Olympus OM-D EM-1 mirrorless camera does not have a square or cross-shaped sensor, of course, but it does basically what you want (including portrait 3:4 without rotating the camera). If you shoot RAW, it will use all the pixels of its entire 4:3 sensor image, but JPGs are cropped as you specify.

So, yes, I would say your idea has merit!

Marsu42 said:
As far as I remember, there was a rumor inventive Sony were to release a square sensor ff camera. And it makes sense, after all the lens is round so you're essentially throwing away potential good pixels. The only idea I have that prevents this is that current lenses are qc-tuned for max. 3:2 and the glass might not be up for a 1:1 task? But this is ot, maybe you should open a new thread for this.

mrzero said:
Genius. I want to take portrait photos without removing my baseball cap. That's got to be worth a few grand more.
 
Hi Slyham.
It's a stupid idea right up to the point someone decides it has merit and puts it in to production! ::)
I think square format is out because of the rectangle in a circle (the image circle) being able to have a longer major side than a square fitted in the same circle.
I hope your patent is granted, too late to apply now you shared your idea! ;D

Cheers, Graham.

Slyham said:
What do you all think? Stupid idea or does it have some merit?
 
Upvote 0
i've asked this question before. I figured that a square would also provide the most area, but smarter people than me said that wasn't correct. they stated that a 3x2 sensor provided more area. i think someone even posted a mock up graphic of the formats laying onto of each other. i never bothered to do the math, but i still think the square would be larger, or am i just missing something? more importantly to me, i just prefer a 4x5 or so ratio most of the time. for me, 3x2 is usually a bit to wide for most stuff and not nearly wide enough when i want to go wide. sure i just stitch, but sometimes things are moving and that just doesn't work out to well.
almost forgot- cross shaped sensors? ain't nobody got time for that one. how about we settle on a square and i open a business selling cross shaped GOBOs in spin on filters! sweet!!
 
Upvote 0
risc32 said:
i've asked this question before. I figured that a square would also provide the most area, but smarter people than me said that wasn't correct. they stated that a 3x2 sensor provided more area. i think someone even posted a mock up graphic of the formats laying onto of each other. i never bothered to do the math, but i still think the square would be larger, or !!
The largest rectangular area that will fit in a circle is a square. Its a bit difficult to show mathematically, but they do it here.
http://oregonstate.edu/instruct/mth251/cq/Stage8/Lesson/rectangle.html

However, having to workaround by doing time consuming and error prone processes to get a wide angle is not something I'd want to try, and ultra wide lenses bring lots of issues that we don't see in 24mm or longer lenses.

Some have suggested circular sensors. They might be the most practical, because some of us never quite hold our cameras horizontal or vertical, and the composition could be more efficient. I think there are technical issues with doing that, but nothing horrible difficult. It would possibly result in lower yield of sensors per wafer, but sensors are no longer extremely expensive to make like they were 10 years ago.
 
Upvote 0
risc32 said:
i've asked this question before. I figured that a square would also provide the most area, but smarter people than me said that wasn't correct. they stated that a 3x2 sensor provided more area. i think someone even posted a mock up graphic of the formats laying onto of each other. i never bothered to do the math, but i still think the square would be larger, or am i just missing something? more importantly to me, i just prefer a 4x5 or so ratio most of the time. for me, 3x2 is usually a bit to wide for most stuff and not nearly wide enough when i want to go wide. sure i just stitch, but sometimes things are moving and that just doesn't work out to well.
almost forgot- cross shaped sensors? ain't nobody got time for that one. how about we settle on a square and i open a business selling cross shaped GOBOs in spin on filters! sweet!!

That was me.
 

Attachments

  • Untitled-10.png
    Untitled-10.png
    29.7 KB · Views: 232
Upvote 0
The obvious answer would be to use a circular raw detector, and then crop to whatever format you like (either in camera or post). This would make ideal use of the optics, which inherently provide a circular image. I think we will eventually get sensors like this, once the sensors become significantly less expensive than the optics. The pixels may then be hexagonal as well, perfectly sampling an f/1.0 diffraction-limited image and interpolated to produce the desired orientation of square pixels.
 
Upvote 0
A 36mm x 24mm sensor has an area of 864 sq. mm.

A square sensor with the same image circle would have sides that are 30.59 mm and an area of 936 sq. mm.

However, a cross shaped sensor with overlapping 24 x 36 mm rectangles would have an area of 1,152 sq. mm.
 
Upvote 0