[WPRESTO]

Does everyone remember the "panoramic" one-use film cameras that were all the rage in the 1980's? The color negatives (always, no chrome) came back with a narrow exposed strip between the sprocket holes, so the format was cropping the width, not extending the length, although those cameras used a wider-angle lens than standard Instamatics. Because of the popularity of those super-cheap cameras, some SLR's were equipped with a cropping shield that could be flipped into place, but it was just throwing away part of the available emulsion. Not the same as a Widelux

[Tony Belding]

Why are we stuck with 3:2 ratio anyhow -- just because 35mm film used it? Why did 35mm film use it -- because early motion picture film used it, in the 1930s? It's kind of silly when you think about it. Our default aspect ratio for still photography in 2015+ shouldn't be dictated by 1930s motion picture film.

I think the idea here is not to simply crop down to different aspect ratios -- because most digital cameras do that already. What would make it really good would be sensor over-provisioning: putting in an oversized sensor, larger than your format's image circle. It would still work by cropping, but instead of cropping down to an undersized frame, it would crop down to the full size that your lenses support.

The benefit is, you could switch between aspect ratios without changing the viewing angle of your lenses, and without reducing your resolution. I like this idea so much, I'm surprised more camera makers haven't gone for it. (Panasonic did make a couple of M4/3 bodies that did this.)

Problems? Well, yeah... Problem one, with an OVF, is that you need some sort of mechanical blinders inside the VF to change the aspect ratio you see, or else some annoying guidelines etched into it. With any sort of LCD or EVF it's not a problem.

Problem two, you are paying for a larger sensor in your camera. That will cost.

Problem three, you are (most likely) paying for a non-standard sensor in your camera, instead of one produced in vast quantities. That will cost.

I still hope this will happen someday -- the sooner the better -- if sensors continue to become cheaper.

[normhead]

The only problem I see with 2x3, is the 8x12 print size, lots of custom mat cutting and a lousy fit in the frames.
3x4 or 9x12, or 12x16 is better.

But 4x5 for 8x10 and 11x14 is standard.

Which is pretty bizarre given that hardly anyone shoots 4x5 or 8x10 film any more. If you crop in your viewfinder, the better a job you do, the worse off you are.

All my favourite images, cropped in the viewfinder, require custom frames and matts. That's just not right.

[WPRESTO]

The only problem I see with 2x3, is the 8x12 print size, lots of custom mat cutting and a lousy fit in the frames.
3x4 or 9x12, or 12x16 is better.

But 4x5 for 8x10 and 11x14 is standard.

Which is pretty bizarre given that hardly anyone shoots 4x5 or 8x10 film any more. If you crop in your viewfinder, the better a job you do, the worse off you are.

All my favourite images, cropped in the viewfinder, require custom frames and matts. That's just not right.

The persistence of old standard sizes for print media (4X5, 5X7, 8X10, 11X14, 14X17,16X20) may have less to do with stubborn devotion to tradition than to massive investment in machinery to generate product with those dimensions. The slightly longer, narrower standard 35mm format, actually a double-frame of the movie film dimensions, became standard in the 1930's yet despite being around for 80 years, and dominating photography for more than half that time, it never managed to displace the previous listed aspect ratios that were established before 1900. As all of us who printed B&W know, either you cropped some of your 35mm frame, or you trimmed off some of the enlarging paper after processing. However, most prints are matted for framing, so odd image shapes can be fitted esthetically into pre-made standard dimension frames most of the time, so long as they aren't square, which looks strange in a rectangular frame, or panoramic. I always assumed that custom cutting a matt opening was part of the process, although not always required.

---

Last edited by WPRESTO; 03-08-2015 at 10:48 AM.

[BrianR]

Keep added lines out of the viewfinder please. Avoiding having to make a mirror so large it hits the rear of my lenses would be appreciated. Include customization for crop lines for live view though.

The only problem I see with 2x3, is the 8x12 print size, lots of custom mat cutting and a lousy fit in the frames.

12x16" mats and frames seem pretty standard these days, giving a uniform 2" of breathing room around an 8x12.

[cfraz]

Why are we stuck with 3:2 ratio anyhow -- just because 35mm film used it?

No, not just because. But SLR cameras and their lenses were designed around 35mm film and 3:2 aspect ratio. For that form factor, the registration distance - distance between lens mount and film - is just sufficient to clear a mirror which reflects the 24 X 36 mm image. One way to get an aspect ratio closer to 1:1 is to make the mirror longer, which forces an increase in the registration distance, requiring a new lens mount and all new lenses. The other way is to decrease the image width to < 36 mm, i.e. use a smaller than 24 X 36 mm sensor. The second option is no different than cropping in PP.

[Tony Belding]

the registration distance - distance between lens mount and film - is just sufficient to clear a mirror which reflects the 24 X 36 mm image.

Ouch! I hadn't thought of that.

But that problem doesn't arise in mirrorless systems. Hmm. . .

[Boris_Akunin]

No one is going to make a custom-sized sensor for Pentax, not at a viable price point anyway. It's either 23.6*15.7mm or 36*24mm...

I'd love to have a crop-preview in the viewfinder (for different aspect ratios and smaller image circles), I wonder if a zoom-OVF would be possible.
The viewfinder could zoom to match the desired image height and movable VF-blinds or framelines could mark the image width, that way the viewfinder image would always be as large as possible.

The zoom factor and the position of the blinds/framelines would depend on the chosen aspect ration and the size of the (usable) image circle:


IDK if there is enough room for zoom optics in the viewfinder and the viewfinder info display would have to be realised some other way but it should be possible in principle.


image: K-5IIs cross-section

The camera would need a database containing the size of the usable image circle for every non-FF lens (ideally with focal-length-dependent values for zooms and fully customisable).

[Kunzite]

Nice thread. By the way, "35mm full-frame" refers to a very specific format, i.e. approximately 24x36mm.

[Clavius]

Ouch! I hadn't thought of that.

But that problem doesn't arise in mirrorless systems. Hmm. . .

Flapping mirrors, restricting our creativity since the 1930s. With a mirrorless a square sensor of 36x36 could be used and we would STILL be able to use our old FF lenses.

[BrianR]

With a mirrorless a square sensor of 36x36 could be used and we would STILL be able to use our old FF lenses.

It would vignette like crazy if you tried to use the whole square, 24x36mm has a diagonal of 43.27mm, but 36x36mm would have a 50.91mm diagonal. Still, being able to crop to the max square size the image circle can handle would be nice, and the lazy would never have to rotate the camera to portrait orientation ever again.

Being able to output circular images the exact size of the imaging circle would also be awesome, and adds worthiness to the term 'fullframe'. Circle prints! Circle mats! Frame with old bike tires!

[lawsonstone]

Yes, it would be cropping the sensor. But with forethought and intent, instead of random post-hoc hacking away at the capture. Thus the idea of gently glowing frame lines in the viewfinder, indicating the selected capture geometry.

Note too, that the capture areas are (in square mm) 864 (3:2), 884 (4:3), 800 (5:4) and 900 (1:1). So we're not compromising on image quality.

Note too, some of the comments in other threads, bemoaning issues of bloated file size with a 50Mpx sensor. With a selectable aspect ratio, the photographer chooses their preferred geometry, and the camera only need retain those specified data points. So, the rate of reading the sensor is preserved, and file size is contained.

Weird. What you call "post hoc hacking away at the capture" I used to call "Good work in the darkroom" for film. Post-Processing is not post hoc hacking; it's taking a good, solid "negative" and doing the work in software that we used to do in the darkroom: turning a "capture" into a "picture."

[cfraz]

Note too, that the capture areas are (in square mm) 864 (3:2), 884 (4:3), 800 (5:4) and 900 (1:1). So we're not compromising on image quality.

In order to get that 30 X 30 image you need a larger mirror in the vertical dimension. As far as I can tell, the mirror in 24 X 36 SLRs will not reflect a 30 X 30 image because it's too small in the vertical dimension. I don't think a 30 X 30 mirror will clear the lens mount.

It seems to me in addition to a larger sensor and viewfinder optics, you need a bigger mirror, a deeper mirror box, and all new lenses to match the longer registration distance to get that 1:1 900 mm^2 image.

[jonby]

Weird. What you call "post hoc hacking away at the capture" I used to call "Good work in the darkroom" for film. Post-Processing is not post hoc hacking; it's taking a good, solid "negative" and doing the work in software that we used to do in the darkroom: turning a "capture" into a "picture."

I don't think Thingo is intending to have a go at anybody who prefers to crop post capture. S/he is suggesting that this approach may not be the best way of doing things for some photographers and in some situations, and is suggesting that it would be a good idea to give photographers the option of working another way, by giving them the tools (temporary viewfinder marks/masks) to frame precisely within aspect ratios other than 3:2 and to only record the data captured within that specific area. In my view this is potentially a very useful feature. Definitive frame lines visible in the viewfinder will allow you to compose carefully and precisely and allow you to visualize the final composition much more precisely than by trying to imagine where the edges will be. And if you're working in this deliberate way then capturing the data outside of this frame is simply a waste of processing time and storage space. Given that many are against a 36mp+ sensor due to the storage and processing requirements, redundant data is clearly an issue for some. Why should professionals working to a particular output format be forced to guess where the edges of the frame will be, to give one example? I believe this is a way more useful feature for serious photographers than many which now seem to have become the accepted norm on cameras, yet people seem to be very quick to dismiss it.

One of the most common arguments given against implementing this feature is, 'I don't want lots of frame lines cluttering up my view'. Absolutely - neither do I, but this is not an argument against the feature itself, it is an argument against what is assumed to be one of the necessary results of its implementation. Yet it's quite possible that this could be implemented in a way which only shows crop indicators when they are in use - you just need a technology which allows for the temporary display of lines/masks in the viewfinder. Nikon do this already with an LCD overlay in the D810 and other cameras. I believe this results in a small loss of transmitted light, though clearly not enough to put people off buying them. There may be other technologies which allow such display without loss of light, such as the current technology used by Pentax to show autofocus points, or something else we haven't seen yet. Even physical dark masks may be a possibility, as implemented in some MF film cameras (and indeed in some Pentax 35mm cameras).

Pentax is going to have to implement this one way or another for the APS-C crop. They are either going to have to provide permanent marks on the focusing screen to show the crop or use another method to show marks/lines when the crop in use. In this case, extending the feature to include a number of other crops/aspect ratios will be trivial, so I really think this is worth considering.

[wgosma]

I don't think Thingo is intending to have a go at anybody who prefers to crop post capture. S/he is suggesting that this approach may not be the best way of doing things for some photographers and in some situations, and is suggesting that it would be a good idea to give photographers the option of working another way, by giving them the tools (temporary viewfinder marks/masks) to frame precisely within aspect ratios other than 3:2 and to only record the data captured within that specific area. In my view this is potentially a very useful feature. Definitive frame lines visible in the viewfinder will allow you to compose carefully and precisely and allow you to visualize the final composition much more precisely than by trying to imagine where the edges will be. And if you're working in this deliberate way then capturing the data outside of this frame is simply a waste of processing time and storage space. Given that many are against a 36mp+ sensor due to the storage and processing requirements, redundant data is clearly an issue for some. Why should professionals working to a particular output format be forced to guess where the edges of the frame will be, to give one example? I believe this is a way more useful feature for serious photographers than many which now seem to have become the accepted norm on cameras, yet people seem to be very quick to dismiss it.

One of the most common arguments given against implementing this feature is, 'I don't want lots of frame lines cluttering up my view'. Absolutely - neither do I, but this is not an argument against the feature itself, it is an argument against what is assumed to be one of the necessary results of its implementation. Yet it's quite possible that this could be implemented in a way which only shows crop indicators when they are in use - you just need a technology which allows for the temporary display of lines/masks in the viewfinder. Nikon do this already with an LCD overlay in the D810 and other cameras. I believe this results in a small loss of transmitted light, though clearly not enough to put people off buying them. There may be other technologies which allow such display without loss of light, such as the current technology used by Pentax to show autofocus points, or something else we haven't seen yet. Even physical dark masks may be a possibility, as implemented in some MF film cameras (and indeed in some Pentax 35mm cameras).

Pentax is going to have to implement this one way or another for the APS-C crop. They are either going to have to provide permanent marks on the focusing screen to show the crop or use another method to show marks/lines when the crop in use. In this case, extending the feature to include a number of other crops/aspect ratios will be trivial, so I really think this is worth considering.

Well now...this is a pretty old thread but since I've been wrestling with my 'aspect ratio anxiety' of late it's of interest.

My 'issue' is I tend toward composing the image in the viewfinder to what I feel is the 'strongest' composition; one can't always do that but often I think (I'm talking landscape pics mainly). So then my 'problem' comes to post-processing and printing: here I sit at my computer and if my field capture is as I (often) expect then I don't need to do any, or maybe VERY LITTLE cropping - fine, but then I want to print and frame, mount, etc. that image that is likely right at 3:2 aspect ratio. There are more frames and precut mats available in 'digital' sizes but still not nearly the abundance of the 8x10, 11x14, etc. stuff. How to deal with this? Go to borderless, such as canvas or gatorboard mounting?

What are your thoughts on my so-called 'dilemma'?
Thanks/Bill