Lens design: why constant apertures?

[vikingar]

Why are high end lenses constant aperture? Isn't the max aperture a limitation of the long end due to the size of the lens entrance pupil? Couldn't these lenses have even bigger apertures at the wide end with no increase in size?

What I'm wondering is: what would the physical difference between a 70-200 f2.8 and a hypothetical 70-200 f2.0-2.8 be? Shouldn't those come out at the same physical size? Or say a 24-105 f4 vs a hypothetical 24-105 f2.8-4?

Perhaps someone in physics and lens design has a good answer.

 

[neuroanatomist]

Size isn’t the problem, per se. Rather, it’s image quality. The wider end has more aberrations, and they’re going to be more prominent if the wide end isn’t stopped down from its theoretical max. So your hypothetical 24-105/2.8-4 would be pretty mushy at 24/2.8 and not worthy of the L designation.

 

[EricN]

Size isn’t the problem, per se. Rather, it’s image quality. The wider end has more aberrations, and they’re going to be more prominent if the wide end isn’t stopped down from its theoretical max. So your hypothetical 24-105/2.8-4 would be pretty mushy at 24/2.8 and not worthy of the L designation.

It would be interesting (if there's a way to modify the aperture blades without destroying the whole lens I doubt it's possible in a modern design, but maybe something like FD 28-85mm F4 could be done) to see just how much the quality is diminished.

 

[EricN]

Why are high end lenses constant aperture? Isn't the max aperture a limitation of the long end due to the size of the lens entrance pupil? Couldn't these lenses have even bigger apertures at the wide end with no increase in size?

What I'm wondering is: what would the physical difference between a 70-200 f2.8 and a hypothetical 70-200 f2.0-2.8 be? Shouldn't those come out at the same physical size? Or say a 24-105 f4 vs a hypothetical 24-105 f2.8-4?

Perhaps someone in physics and lens design has a good answer.

Besides what @neuroanatomist said, it keeps the levels more consistent than if you're using a variable aperture starting wide open and zooming in it's stopped down. some what similarly, if we want to zoom without changing the focus or change the focus without zooming, it's also going to cost more.

Edit: I'm making all kinds of writing mistakes

 

[neuroanatomist]

Besides what @neuroanatomist said, it keeps the levels more consistent than if you're starting wide and zooming in.

If you mean keep the exposure constant while zooming a variable aperture lens, that is possible with a setting (Same expo. for new aperture) on many cameras.

 

[EricN]

If you mean keep the exposure constant while zooming a variable aperture lens, that is possible with a setting (Same expo. for new aperture) on many cameras.

That's certainly true. Even now, we both know some people love to complain and will say "the difference between F/6.3 and F/7.1 is unusable." Until the digital age, the only two options were to change the shutter speed or use push processing. Can you imagine if some of today's trolls and constant complainers lived back then? "OMG Sony got it right with Mavica! Canon and their EOS will never last!"

 

[vikingar]

If you mean keep the exposure constant while zooming a variable aperture lens, that is possible with a setting (Same expo. for new aperture) on many cameras.

With the R5 + 100-500 you can set it to open the aperture back up automatically when zooming out. And if they really wanted they could add a limiter switch on the lens. Not going to happen of course but technically possible.

I was thinking how Canon could / would improve the 70-200 f2.8L for the rumored Mk II. I guess it's going to be an internal zooming video "Z" version, but I think it would be really cool if we got f2.0 even just from say 70-90mm.

I do like my constant aperture lenses a lot for the convenience. But with Canon tickling extra millimeters out of every existing lens design and not being afraid of heavy corrections I was wondering if there's anything preventing them from opening the aperture a bit.

With the WA lenses they are already hitting the limits for "L" lenses (14-35 but also 15-35) so clearly they're not afraid of heavy vignetting and distortion if it can be corrected. Opening the aperture would also result in more vignetting, so maybe adding extra mms was easier/better.

Would certainly be interesting to see just how bad the image quality would be with a wider aperture.

 

[AlanF]

I do like my constant aperture lenses a lot for the convenience. But with Canon tickling extra millimeters out of every existing lens design and not being afraid of heavy corrections I was wondering if there's anything preventing them from opening the aperture a bit.

It would be great if the 200-800 f/9 zoomed out to 400 f/4.5 and 200 f/2.25, which they could in theory do.