Jump to content

Sensor Dimensions - Let's really innovate


Recommended Posts

I would like to see a new innovation from Sony. Abandon 36x24mm sensors in favor of 35x28mm sensors. 20% more area, yet more compatible with all current 35mm format full frame lenses. No need for expensive medium format. If people desire 36x24mm format for video, cropping to 35x24mm still provides a format very close to 36x24mm with less than 3% fewer pixels than 36x24. The 35x28 format (5x4) is a much better portrait format. What do you think, Sony?

Link to post
Share on other sites

20%!? My apologies, math by memory. I confused the value with another recent calculation in my head. 13% is correct. Ok, ignoring my bad memory (I worked as a mathematician, so no intent to obscure the issue with bad math), let's talk about 13% larger sensor. It would work, cynicism about sprocket holes aside. If Hasselblad can do a larger sensor in its new camera no reason Sony could not do a more modestly bigger sensor. !3% larger sensor plus 4x5 format, all improvements. An added benefit is better coverage with current 35mm full frame lenses.

Link to post
Share on other sites

It will likely happen after pigs fly, but ... 13% with

digital is better than 13% back in the film era. It's

a funny world.

 

In the film era, marketing always bragged about

the increase in area, which was bogus. Only the

linear increase mattered, which was much less

than the area increase.

 

The difference between film and digital regarding

increased format size, is that with film, size is only

size. No joke. IOW, a bigger format just involved

MORE film, but it didn't cause the film to become

a superior quality of film.

  

With digital, a larger format still delivers the same

classic advantage of involving MORE film ... but

the film is now a sensor ... but it also provides a

possibility of increasing the actual QUALITY of

the film .... cuz the film is a sensor now.

  

When you use a bigger piece of film the behavior

of the film grains is not affect by the larger size of

the film. But, when you have digital photo sites,

instead of film grains, their behavior is affected by

how closely you crowd them together. So this is

the Big Deal. For any given MP count, the larger

sensor mean less crowding of the photo sites. 

 

Kapische ? So 13% with film was only 13%, but

with sensors, the effect is compounded. There's

no fixed mathematical advantage. It's up to the

engineers how much extra "breathing room" they

will provide to the photo sites. But with a bigger

sensor SOME extra room will be provided, and

that will reduce noise while the larger dimension

of the sensor increases resolution. It's a toofer,

which didn't happen that way with film :-) 

Link to post
Share on other sites

How many of the world's great portraits are displayed at 5:4 I wonder...

  

FWIW, when the main, nearly only, application of photography

was portraits and landscapes, standard plate sizes were not all

of one single ratio. There were 4x5 & 8x10, 3.25x4.25 & 9x12,

5x7 & 11x14, etc.  

  

The was no 2x3. Cine was 3x4 on 35mm perf stock, and when

you use two cine frames together as the 135 still photo format,

you join two 18x24mm cine frames into a single 24x36mm 135

format still photo frame, thus the beginning of the reign of 2x3. 

  

There were formats slightly more elongated than 2x3, not for

panoramic effect, but for contact printing as postcards. This

was before email and txting. Many postcard cameras allowed

you to annotate the picture with a handwritten caption before

winding the film to the next frame. The longer format provided

a bit of extra room for that. This was pre-Adobe, pre-snapchat,

and IIRC it was possibly pre-iPhone, but don't hold me to that.

    

============================================= 

    

I find 2x3 a very useful "capture" format, but by "capture" I am

inferring that I'll generally crop to some other ratio. Action that

is moving on a horizontal plane is likely to result in a "capture"

that needs some recomposing, due to human reflex time etc.

There will be unwanted space ahead of, or behind, the action.

So, 2x3 is handy for that. Rotate 2x3 to 3x2 "portrait", and it's

terrible for action, often awkward for portraits, but can be cool

for architecture. I spoze if rocket launches were still big news

that would be a case where 3x2 serves well for action ;-)  

    

###############################################  

###############################################   

   

Vaaaaaat Digression Ahead ! [but is still about format shape] :

  

FWIW, I prefer square images. Not the square "capture" that

is proclaimed as universally crop-friendly. I prefer square for

finished, hard-copy product. Doesn't matter much on images

intended to be seen on a screen cuz a dynamic environment

like that makes image shape irrelevant. The outline is simply

the outer limit of the image content. IOW the image is not an

"Object" as in "Art Object" etc etc. But when the image IS an

object, I still want the content, the image inside the frame, to

overpower the "object-ness" of the object. And, toward that

end, I find that the more a rectangular shape deviates from a

square, the more it calls attention to the "rectangle-as-object,

containing an image". IOOW the image hasta compete with

its container, and the more the container is "interesting", the

more it competes with its contents ... the image. 

  

Squares, being the closest to a circle that any 4-sided objects

can ever be, are more neutral. It's more "shapeless", thus less

distracting. A circular opening in a square matte is ALL about

the image. A framed panoramic is almost all about the format.

  

I actually LIKE panorama. Format itself can be an element of

making an "art object", and successfully relating the Content

to the Concept is a Big Deal in "making art". But if the content

itself is to be of ultimate importance, then format, concept, etc

hafta take a back seat, waaaay back, behind the rear axle. 

   

Also FWIW, when an image is framed and hung, I find that the

square may need tweaking. Often a 22hx20w looks even more

neutral, more invisible, than a truly square 20x20 square. I'm

thinking that our feeling the effect of gravity between the soles

of our feet and the floor causes this perceptual anomaly ;-) 

  

I know I said stuff like "the object-ness of the object" and also

"a truly square ... square", so if you read this far, thank you :-)  

  

`

Link to post
Share on other sites

I did indeed read that far! Interesting stuff too.

One point I implied was that "portraits" didn't begin with photography... was there a more frequent "format" used by the painters of yesteryear?

I too like a square image although with my main interest being anamorphic 2.35:1 video that's more awkward.

Link to post
Share on other sites

...............  "portraits" didn't begin with photography...

was there a more frequent "format" used by the painters

of yesteryear?

............... 

  

Statues were popular :-)   

  

No rectangular prison for those portraits, which were  

not all full length shots. Busts were quite popular, so

there you have the "head shot". 

  

But you did ask about the proclivities of painters, not

about the sculptors. The answer is soooo accessible

online that I will just invite you to browse for yourself. 

I can offer a trifling bit of guidance: 

 

You will see a lot of approx 4x3 "squat" rectangles.

   

You will see a some 3x2, mostly full length standing

poses, not a pose most photographers have in mind

when discussing portraiture. You may see elongated

rectangles as elements of triptychs.

  

You will see some extreme rectangles, especially of

saints, because such images are not framed canvas

hanging on castle walls. The extreme rectangles are

usually images that fit into the panels of architecture

or furnishings such as altarpieces, sepulchers, etc.

    

Guidance-wise, the main thing I wanna open your

eyes about is the techniques employed to dissociate

the subject person from the outer rectangle. You will 

find lighting is used to vignette the subject. Drapery

and architectural shapes often separate the subject

from the shape of the rectangle. Acoarst, something

you NEVER see is cotton candy background bokeh ! 

  

Despite my tendency toward loooong posts, I'm not

up for posting a whole History-of-Art course here :-)

As I said above, you will very readily find a whole art

education online and I choose to believe you'll have

a lotta fun doing it.  

 

https://www.google.com/search?q=paintings+of+kings&biw=1009&bih=512&tbm=isch&tbo=u&source=univ&sa=X&sqi=2&ved=0ahUKEwjP4oicgNTNAhXRZiYKHca2CyoQ7AkIQA

  

Welcome, dear visitor! As registered member you'd see an image here…

Simply register for free here – We are always happy to welcome new members!

  

Link to post
Share on other sites

FWIW, I prefer square images. Not the square "capture" that

is proclaimed as universally crop-friendly. I prefer square for

finished, hard-copy product. 

 

 

+1

 

If you have to innovate at this point why not a full 36x36mm sensor? Lens coverage is already here (lenses project a circular image, that you can "crop" whatever you like without exceeding the maximum radius).

 

And for people who want a rectangular picture, without the need to choose a final crop at the time of capture, camera makers could even innovate on the typical camera shapes.

 

In the end we might end up with a digital, smaller version of the Hasselblad 500c/m   :D   and assuming prices would still be affordable (I doubt it...) I'd be happy as a hippo in the mud!*

 

 

*here today we're way over 35° C / 95° F with the same level of humidity of a fish tank, so I'm seriously envying hippos...

Link to post
Share on other sites

 

 

If you have to innovate at this point why not a full 36x36mm sensor? Lens coverage is already here (lenses project a circular image, that you can "crop" whatever you like without exceeding the maximum radius).

 

 

 

No, not 36x36mm because the diameter of a circle covering 24x36mm is 43mm and if you want to put a square in that circle, then the side can be 30mm at the most. Figures are somewhat approximative because I have calculated them without access to a calculator or a square root table. But I know my Pythagoras.

Link to post
Share on other sites

It will likely happen after pigs fly, but ... 13% with

digital is better than 13% back in the film era. It's

a funny world.

 

In the film era, marketing always bragged about

the increase in area, which was bogus. Only the

linear increase mattered, which was much less

than the area increase.

 

The difference between film and digital regarding

increased format size, is that with film, size is only

size. No joke. IOW, a bigger format just involved

MORE film, but it didn't cause the film to become

a superior quality of film.

  

With digital, a larger format still delivers the same

classic advantage of involving MORE film ... but

the film is now a sensor ... but it also provides a

possibility of increasing the actual QUALITY of

the film .... cuz the film is a sensor now.

  

When you use a bigger piece of film the behavior

of the film grains is not affect by the larger size of

the film. But, when you have digital photo sites,

instead of film grains, their behavior is affected by

how closely you crowd them together. So this is

the Big Deal. For any given MP count, the larger

sensor mean less crowding of the photo sites. 

 

Kapische ? So 13% with film was only 13%, but

with sensors, the effect is compounded. There's

no fixed mathematical advantage. It's up to the

engineers how much extra "breathing room" they

will provide to the photo sites. But with a bigger

sensor SOME extra room will be provided, and

that will reduce noise while the larger dimension

of the sensor increases resolution. It's a toofer,

which didn't happen that way with film :-) 

Hello.

First post on this site as I venture my way back into doing some photographer as a hobby (bought an A7RII to play with), however I have over 35 years experience working in motion picture film and motion picture digital so I do have some experience with film emulsions and grain structure, as well as digital image capture and post processing (more for cinema than stills, but not that different).

 

What I have read above is a very strange perception of how film emulsions and image size relate, and also digital sensor size and image size.

Regardless of whether it is for "stills" or "movie" use, or what format you chose, film gave you options of choosing different emulsions with different grain structures  for different purposes for a single camera.  To put it simply, fast film was large grain structure (larger ISO/EI), slow film was finer grain structure (lower ISO/EI) so for any given capture format (bigger film size = better end resolution) you had options of varying the grain structure depending on a number of factors (available light v artificial lighting budget, speed of lenses etc).  How big the grain structure appeared in the end product depended on your final product format... a small printed photo, A3 poster, billboard or giant cinema screen.

Motion picture (and stills) photographic scientists didn't spend a hundred years sitting on one or 2 film types... there was constant development of emulsions right up to the near demise of Kodak as a supplier (and there may still be ongoing research, but it would be minimal).

When it comes to digital,  higher MP count does mean "more resolution" however you buy a camera and in doing so choose your "film stock" and you are stuck with it, ie, the sensor.... with fixed number of pixels and DR... there is no changing that without buying a different camera, apart from software advances that might milk a tiny bit more during internal processing out of the initial capture by the sensor.  The engineers design different camera products, each to act a certain way, you make your choice and that's why you buy... depending on your budget.

For digital, noise is more relative to the sensors ability to capture light above the noise floor of the sensor itself, so a sensor with a higher dynamic range can hide the noise floor better than one with lower dynamic range by being able to capture better highlights as well as better low light, usually meaning you can grade a final image to provide a pleasing image with less "apparent" degradation than a lower DR sensor might give.

Just adding more pixels didn't increase DR, that comes with sensor development, just like film emulsions were improved over time.

I love reading through the varied posts and peoples views on their cameras, lenses, use etc... but sometimes information appears to get posted as fact, which in my experience is obviously incorrect and needs a (hopefully) balanced contradiction.

cheers

 

 

Link to post
Share on other sites

No, not 36x36mm because the diameter of a circle covering 24x36mm is 43mm and if you want to put a square in that circle, then the side can be 30mm at the most. Figures are somewhat approximative because I have calculated them without access to a calculator or a square root table. But I know my Pythagoras.

 

 

Oops, yes obviously you're right... And to think that dear old Pythagoras lived just in the next valley, on the other side of the mountains I can see from my windows! Well, I guess better for me he is not alive anymore, otherwise he might have beaten me with a stick for my error   :D

 

That said, at least many "classic / legacy" lenses (I haven't tried the latest Sony's) cover quite a lot more than 43mm even at infinity. Years ago I used some (I think were Nikon's, at close distance) even on a 4x5" camera will full coverage (but no movements), and a couple of panoramic cameras (swiss made, don't remember the name, but they were semi artisanal and of the slit-rotating tipe) used Nikon interchangeable lenses to shot on 120 film. So the (or some of the) lenses should be able to cover more or less a 40x40mm sensor according to Pythagoras.

 

So I guess it could still be done, shopping carefully or if the manufacturer did the tests beforehand creating a list of suitable lenses (kinda like the semi-bogus* one Canon did for its 5DsR).

 

 

*semi-bogus because a 50Mp sensor is not more demanding than a 24Mp APS-c sensor, even if obviously the lenses will have to perform good enough even at the borders

Link to post
Share on other sites

....

 

So the (or some of the) lenses should be able to cover more or less a 40x40mm sensor according to Pythagoras.

 

....

 

My first digital camera was the Nex 5n with an APS sensor ( and a touch screen !!! ) When I got my A7 and only had one full frame lens,  I used my Sony and Sigma NEX lenses on the A7. 

 

Depending on focus length and distance setting one could get an area covered that was somewhere between APS and FF. If I remember correctly, the maximum was achieved with wide angle lenses used indoors (i.e. at short distances). The light fall off in the image corners could be partly offset in post processing, and local sharpening would help too. To get maximum mileage, I printed many images square. Still, the usable area was smaller than the area that had been covered by the lens.

 

The NEX lenses on FF  sensors was an emergency solution, but better than no picture at all. And I had great fun testing this out. And I learned that there are good reasons why the lens manufacturers only use part of the area covered by a lens.

Link to post
Share on other sites

The coverage will be worse with tele-centric lenses like the ones "optimized" for digital sensors, that have the outgoing image rays as parallel as possible. And even then, like you pointed out, there are exceptions.

 

But with more traditional lens design lens coverage is often wildly underestimated. 

 

That said mine is wishful thinking anyway, because I'm afraid before we'll see a square sensor at a fair price pigs will have learned to fly...and even then it would probably come with an entirely new lens mount & lens line, if not for anything else for marketing/commercial reasons.

Link to post
Share on other sites

Sensor tech continually improves without any

increase in size. The new live view 'Blad is a

perfect example of what happens where size

enters the equation. Why look for trouble ? 

  

If you just wanna play silly games, use an a7

series with APSC E-mount lenses or adapted

APSC A-mount lenses and enjoy discovering

which APSC lenses have extra coverage. Be

appreciative that a 24x36mm sensor already

represents an increase in size and expense,

and is already available from Sony with the

handily ambiguous E-mount.

Link to post
Share on other sites

Guest Jaf-Photo

Welcome and thank you. You just saved me a lot of typing. I would also say that the differences in projection and resolving power of lenses for different film formats create different looks even if you use the same film stock. Even if you shoot a fine-grained film, like Ektar, the results on 6x7 is like night and day to 36mm.

 

But, yeah, you'll find that everone is an expert here, especially if they're not.

 

 

 

Hello.

First post on this site as I venture my way back into doing some photographer as a hobby (bought an A7RII to play with), however I have over 35 years experience working in motion picture film and motion picture digital so I do have some experience with film emulsions and grain structure, as well as digital image capture and post processing (more for cinema than stills, but not that different).

What I have read above is a very strange perception of how film emulsions and image size relate, and also digital sensor size and image size.

[...]

Link to post
Share on other sites

I don't think discussing about a Change from 2:3 to 3:4/4:5 makes much sense when you have >40MPs in your cam.

But changeing to a 1:1 sensor would mak a LOT of sense and gives a real benefit. I always love the ratio switch on my LX3 - now the same thing but with vertical and horizontal crop. You would need a bigger EVF and a high Screen - but a little more height would be nice anyway for a better grip.

 

Could be a revolutionary camera - but seems like no manufacturer has the balls to be the first one.

Link to post
Share on other sites

I don't think discussing about a Change from 2:3 to 3:4/4:5 makes much sense

when you have >40MPs in your cam.

But changeing to a 1:1 sensor would mak a LOT of sense and gives a real benefit.

I always love the ratio switch on my LX3 - now the same thing but with vertical and

horizontal crop. You would need a bigger EVF and a high Screen - but a little more

height would be nice anyway for a better grip.

 

Could be a revolutionary camera - but seems like no manufacturer has the balls to

be the first one.

  

Perfect.

 

Except I don't think a taller screen is needed and so the body won't

hafta get too tall. Just keep the current landscape oriented screen.

  

Initial "quick check" playback of a square capture doesn't hafta fill

the whole screen. Hit a view toggle button and you would next see

the landscape crop fill the screen. Hit again to see the portrait crop

fill the screen. Simple !  

 

Acoarst typical button customizing means you could change the

order in which the toggle happens, or set either of the crops to be

the initial playback so that you hit the toggle to see the square, etc.

  

This is a simple no-cost firmware idea. Making the screen and the

camera body taller is costly.

Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
 Share

×
×
  • Create New...