p.2 #3 · Sony, I command you to release a 8,075 × 8,075 pixel square sensor!
Makten wrote:
Is that why 99% of all painters have used 5:4 and 4:3 over the last 500 years or so? Or why virtually every photographic system except "full frame" uses 5:4 or 4:3 or 1:1?
Edit: Or panoramic formats, of course. But 3:2? Nope, very few have used that. Personally I hate it.
To be fair, APSC uses it, too.
7:5 is the aspect ratio of the gods.
Feb 27, 2026 at 04:02 AM
Steve Spencer Offline • • • • • • • Upload & Sell: On
p.2 #4 · Sony, I command you to release a 8,075 × 8,075 pixel square sensor!
jeffbuzz wrote:
3:2 is closest to the Golden Ratio which makes more naturally appealing compositions. How often does anyone present their finished image in square format?
The 3:2 rectangle captures 96% of the same image circle compared to a square. So there's no real advantage to a square sensor from a material standpoint.
I think 4 X 3 is the Golden Ratio, because a rectangle with a 4 X 3 shape has a diagonal of 5. At least that is what I was taught a billion years ago when I took geometry in high school, but perhaps my teacher had that wrong.
p.2 #5 · Sony, I command you to release a 8,075 × 8,075 pixel square sensor!
One reason behind the avoidance of square sensors may be the effect they have on readout speed, since sensor readout is oriented around rows rather than columns. Stacked and partially-stacked sensors would offset this penalty but perhaps camera makers aren't ready to give up some of those performance gains yet for a more flexible square ratio.
p.2 #6 · Sony, I command you to release a 8,075 × 8,075 pixel square sensor!
Steve Spencer wrote:
I think 4 X 3 is the Golden Ratio, because a rectangle with a 4 X 3 shape has a diagonal of 5. At least that is what I was taught a billion years ago when I took geometry in high school, but perhaps my teacher had that wrong.
Not when I was in the high school; it was 1.62 or 0.62 that are reciprocals of each other. =(SQRT(5)±1)/2
The 3:4 ratio is the simplest example of hypotenuse and two sides being whole numbers. I think it was the ratio in the Golden Age of Television.
EBH
Feb 27, 2026 at 08:13 AM
Steve Spencer Offline • • • • • • • Upload & Sell: On
p.2 #7 · Sony, I command you to release a 8,075 × 8,075 pixel square sensor!
EB-1 wrote:
Not when I was in the high school; it was 1.62 or 0.62 that are reciprocals of each other. =(SQRT(5)±1)/2
The 3:4 ratio is the simplest example of hypotenuse and two sides being whole numbers. I think it was the ratio in the Golden Age of Television.
EBH
As I was taught the golden ratio of 4 X 3 was first made famous by Pythagoras about 500 BC and was central to developing his theorem. So, it predates television a bit. Perhaps there are more than one ratio that has been deemed, "Golden"
p.2 #8 · Sony, I command you to release a 8,075 × 8,075 pixel square sensor!
old-gregg wrote:
Consider this: not a single native digital format is as wide as 2:3. Larger medium format sensors are 4:3. Smartphones are mostly 4:3. But the killer is the iPhone front camera whish uses the "archaic" square and crops on the fly, like I was proposing above. Even cooler, actually, it's an octagon.
When it comes to photographic aspect ratios, 3:2 is the very definition of archaic: it's inefficient, hardly suitable for vertical orientation, too wide for normal photography and not wide enough for panaramas.
Phones are not 4:3 since around the 2000s. The displays of Galaxy S26 series are 19.5:9 as are the iPhone 17 series. Screens went to the 19-20:1 range that way since they went to near borderless. The main cameras are 16:9 in keeping with the global video standards.
p.2 #9 · Sony, I command you to release a 8,075 × 8,075 pixel square sensor!
Steve Spencer wrote:
As I was taught the golden ratio of 4 X 3 was first made famous by Pythagoras about 500 BC and was central to developing his theorem. So, it predates television a bit. Perhaps there are more than more ratio that has been deemed, "Golden"
I thought it was part of Euclidean geometry, but I don't recall the 70s so well anymore. Maybe we need a mathemetician.
Frankly I think it is just a number and does not have any magic meaning to a visual appeal. We could all use e-1 (1.71) and give that a magic name just as well.
Nowadays I crop most all images to best fit the composition, but ideally as little as possible. For wildlife and landscapes square is not so good for percent ultilization.
A practical reason not to make large square sensors probably a higher cost of fabrication.
EBH
Feb 27, 2026 at 08:53 AM
Steve Spencer Offline • • • • • • • Upload & Sell: On
p.2 #10 · Sony, I command you to release a 8,075 × 8,075 pixel square sensor!
EB-1 wrote:
I thought it was part of Euclidean geometry, but I don't recall the 70s so well anymore. Maybe we need a mathemetician.
Frankly I think it is just a number and does not have any magic meaning to a visual appeal. We could all use e-1 (1.71) and give that a magic name just as well.
Nowadays I crop most all images to best fit the composition, but ideally as little as possible. For wildlife and landscapes square is not so good for percent ultilization.
A practical reason not to make large square sensors probably a higher cost of fabrication.
EBH
If I remember correctly, and it was a long time ago, I think what you are discussing as the golden ratio has to do with regular pentagons. It is a weird number, actually an irrational number, but shows up in nature quite a bit.
Ok, you made me look it up. It is 89 X 55, so it doesn't exactly work out to whole numbers well. What is distinctive about the aspect ratio is that you can remove or add a square to it and you still get the same aspect ratio. It is also easy to make from a square with just a compass. Kind of cool mathematically, but not an aspect ratio I like for photography. Of course, YMMV.
Oh, Euclid did provide the first definition of it, but it was studied earlier by Pythagoreans.
p.2 #11 · Sony, I command you to release a 8,075 × 8,075 pixel square sensor!
weezintrumpete wrote:
I would kill for a large square sensor. What I really want is a Hasselblad / Rolleiflex sized medium format 6x6 sensor. The dream.
Would you pay $100000 for such a camera? Realistically it wouldn't be any less expensive.
p.2 #12 · Sony, I command you to release a 8,075 × 8,075 pixel square sensor!
old-gregg wrote:
Even if we put aside personal preferences (although it's soooo tempting to make an argument that outside of reusing cine film, all dedicated photographic formats have been much narrower), you have to agree that the primary reason we have 3:2 is historical compatibility. Leica wanted to reuse cine film, then the early DSLRs were built reusing the existing bodies, and now we've dragged this relict into the world of mirrorless. As a result, anyone who finds 3:2 too wide is forced to crop and not getting the 60MP either.
No. 35 mm film supports any aspect ratio you want. 2:3 is a practical and aesthetic choice. It matches the A1, A2, A3, A4, A5 paper sizes once you leave a small white margin around the image area which is useful to ensure correct framing (if it is printed to the edge of the paper, lots of things can go wrong, including printer damage).
Today many would probably prefer a 16:9 aspect ratio that matches our screens, so as to not waste expensive sensor surface area.
Human vision is mostly horizontal panoramic, which is probably why many prefer a rectangular image which is not square. If the objective is to just always crop to a rectangle anyway, having a square sensor is not optimal and it would be more expensive than needs to be to achieve the desired end result.
Although lenses typically do have theoretical spherical image circles, many lenses actually don't project a circle as there are internal baffles that are rectangular (non-square) to minimize internal flare. So lenses would need to be modified and/or redesigned to achieve a square image that is taller than the currently used rectangular sensor.
So, would you pay double the current cost of your camera to get a square sensor, and buy all new lenses to cover it?
p.2 #13 · Sony, I command you to release a 8,075 × 8,075 pixel square sensor!
old-gregg wrote:
[...]Leica wanted to reuse cine film, then the early DSLRs were built reusing the existing bodies, and now we've dragged this relict into the world of mirrorless.[...]
Here you are lost in history. Barnack wanted to make a small camera so he designed one around the small existing cine film. The aspect of ratio could in theory have been set to anything he liked.
p.2 #14 · Sony, I command you to release a 8,075 × 8,075 pixel square sensor!
ilkka_nissila wrote:
2:3 is a practical and aesthetic choice.
The question is still: Why do almost all larger formats have sensors that are closer to square, if 3:2 is so good? Doesn't make any sense at all.
Of course I understand that some people prefer it for other reasons than that it's what they got and they've grown into it, but I can't imagine that would be a majority.
Just walk into any art gallery. Almost nothing will be presented in 3:2.
--------
Edit: Personally I got into 33x44 mm sensors partly (actually mostly) to get native 4:3 instead of 3:2. My old Sony didn't even support 4:3 as a crop, and all the newer bodies where too focused on filming rather than photography.
I still have it and I always crop, which is ruining the experience when the darn camera only shows 3:2.
p.2 #16 · Sony, I command you to release a 8,075 × 8,075 pixel square sensor!
Yup, most phones have 4:3 sensors, but they also often fool the user into 16:9 as if it was "native". The image gets cropped without people knowing about it.
Edit: BTW, I'm pretty sure I read somewhere that the first Barnack prototype had a 24x32 mm frame, which is 4:3. But since he wanted the largest possible area, it was widened slightly to 24x36 in the end. Which is too bad, because I'm sure we would otherwise have 4:3 "FF" cameras now.
p.2 #17 · Sony, I command you to release a 8,075 × 8,075 pixel square sensor!
Makten wrote:
Yup, most phones have 4:3 sensors, but they also often fool the user into 16:9 as if it was "native". The image gets cropped without people knowing about it.
People don't care at all what the native resolution is, what format the sensor has, or whether it's cropped. The majority simply choose the format that gives them the largest image on their smartphone screen.
Makten wrote:
BTW, I'm pretty sure I read somewhere that the first Barnack prototype had a 24x32 mm frame, which is 4:3. But since he wanted the largest possible area, it was widened slightly to 24x36 in the end. Which is too bad, because I'm sure we would otherwise have 4:3 "FF" cameras now.😁
It would be an enormous improvement if we had a 24x32 sensor instead of a 24x36 sensor today. That's completely undisputed.😁
p.2 #18 · Sony, I command you to release a 8,075 × 8,075 pixel square sensor!
Makten wrote:
The question is still: Why do almost all larger formats have sensors that are closer to square, if 3:2 is so good? Doesn't make any sense at all.
Of course I understand that some people prefer it for other reasons than that it's what they got and they've grown into it, but I can't imagine that would be a majority.
Just walk into any art gallery. Almost nothing will be presented in 3:2.
--------
Edit: Personally I got into 33x44 mm sensors partly (actually mostly) to get native 4:3 instead of 3:2. My old Sony didn't even support 4:3 as a crop, and all the newer bodies where too focused on filming rather than photography.
I still have it and I always crop, which is ruining the experience when the darn camera only shows 3:2. ...Show more →
So if your goal is 4:3, is it more economical to crop from 1:1 or 3:2?
On 1:1, you lose 25%. On 3:2, you crop to 2.67:2 and lose 11%.
Sony, I command you to release a 8,075 × 8,075 pixel square sensor!
