How about a new lens ?

jgoetz4

Good Afternoon Folks,
I use my Canon 10-22mm about 75% and my Tamron 17-50 2.8 the remaining 25%. What would it take for either Canon, Sigma, Tamron, etc to produce something like a 10-50mm 2.8 or maybe a 4.0 lens ? I'm sure I'm not the only individual who would buy one if it was a pro sumer type lens. Your thoughts ?
Have a good day
Jim...

ziggy53

I am absolutely amazed at the quality that modern 18-200mm lenses produce, although they still have not achieved enough quality for me to want to purchase one. Typically, these lenses are optimized for one end and compromized in between and at the other end.

Still, at small apertures and typical subjects, the flaws are slight enough that many people really don't see them.

A major problem for lens designers is how light must bend, especially at very short focal lengths. The design goals for 10mm on a crop 1.6x/1.5x imager are so very different from 50mm on the same format.

10mm usually requires multiple aspheric elements that simply do not translate well into the 50mm optical formulas. The 50mm end would have to basically "undo" some of the extreme bending caused by the 10mm end, losing a lot of quality in the process.

I have no doubt that there will probably be a lens someday like you describe, in fact similar designs already exist in the camcorder world, but camcorders do not require the same qualities and the same designs would look pretty sad at modern dSLR resolutions.

For now, quality and convenience are still fairly mutually exclusive.

jgoetz4

ziggy53 wrote:

I am absolutely amazed at the quality that modern 18-200mm lenses produce, although they still have not achieved enough quality for me to want to purchase one. Typically, these lenses are optimized for one end and compromized in between and at the other end.

Still, at small apertures and typical subjects, the flaws are slight enough that many people really don't see them.

A major problem for lens designers is how light must bend, especially at very short focal lengths. The design goals for 10mm on a crop 1.6x/1.5x imager are so very different from 50mm on the same format.

10mm usually requires multiple aspheric elements that simply do not translate well into the 50mm optical formulas. The 50mm end would have to basically "undo" some of the extreme bending caused by the 10mm end, losing a lot of quality in the process.

I have no doubt that there will probably be a lens someday like you describe, in fact similar designs already exist in the camcorder world, but camcorders do not require the same qualities and the same designs would look pretty sad at modern dSLR resolutions.

For now, quality and convenience are still fairly mutually exclusive.

Afternoon Ziggy & thanks for the explanation
It's just that Tokina makes a 16-50mm, which I believe is not a bad lens in terms of sharpness, why not start just a bit wider, say 12-50mm ? Even that would help out
Have a good afternoon
Jim...

ziggy53

jgoetz4 wrote:

Afternoon Ziggy & thanks for the explanation
It's just that Tokina makes a 16-50mm, which I believe is not a bad lens in terms of sharpness, why not start just a bit wider, say 12-50mm ? Even that would help out
Have a good afternoon
Jim...

It's just a few millimeters extra, right. The problem is the difference between 16mm and 12mm on a crop 1.6x camera is 16 degrees of angle-of-view. Those 16 degrees are very difficult to achieve in part because the lens has to accomodate the optic relay required for a short focal length lens to even fit on the camera.

If you took a simple 12mm lens and tried to attach it to the camera, where would it fit in front of the image plane?

It would have to be attached 12mm in front of the image plane to achieve infinity focus by definition.

The problem is, the lens mount flange on Canon EF cameras is 44mm ahead of the image plane and you cannot build a lens which protrudes too far into the body lest the rear element hit the mirror. The solution is to design the lens so that convergence takes place at a distance greater than normal for the effective focal length.

The optic relay is just such a solution and an integral part of short focal length lenses. The problem is trying to create an optic relay into the lens formulae that also accomodates multiple focal lengths (zoom) and allows consistant image flatness and converges all wavelengths at the same point and eliminates curvilinear distortion and keeps vignetting to a minimum and ... I'm sure I've forgotten things.

Many of the latest zooms with extreme range have double cam mechanisms, which is partly why they have a different "feel" through the zoom travel. If you could see inside the lens while it moves, you would see a rather complicated series of motions. You can get a sense of that by looking through, for instance, the original Canon "kit" lens, which appears to use the double cam design. Zoom and watch the lens elements move through the range. As you add more range, and especially to include short focal lengths, you might have to go to a triple cam design. At some point either cost or complexity or image quality goals become counter productive and those "dream" lenses stay dreams.

The point is, it's not easy and, if it was easy, they would have done it already.

jgoetz4

ziggy53 wrote:

It's just a few millimeters extra, right. The problem is the difference between 16mm and 12mm on a crop 1.6x camera is 16 degrees of angle-of-view. Those 16 degrees are very difficult to achieve in part because the lens has to accomodate the optic relay required for a short focal length lens to even fit on the camera.

If you took a simple 12mm lens and tried to attach it to the camera, where would it fit in front of the image plane?

It would have to be attached 12mm in front of the image plane to achieve infinity focus by definition.

The problem is, the lens mount flange on Canon EF cameras is 44mm ahead of the image plane and you cannot build a lens which protrudes too far into the body lest the rear element hit the mirror. The solution is to design the lens so that convergence takes place at a distance greater than normal for the effective focal length.

The optic relay is just such a solution and an integral part of short focal length lenses. The problem is trying to create an optic relay into the lens formulae that also accomodates multiple focal lengths (zoom) and allows consistant image flatness and converges all wavelengths at the same point and eliminates curvilinear distortion and keeps vignetting to a minimum and ... I'm sure I've forgotten things.

Many of the latest zooms with extreme range have double cam mechanisms, which is partly why they have a different "feel" through the zoom travel. If you could see inside the lens while it moves, you would see a rather complicated series of motions. You can get a sense of that by looking through, for instance, the original Canon "kit" lens, which appears to use the double cam design. Zoom and watch the lens elements move through the range. As you add more range, and especially to include short focal lengths, you might have to go to a triple cam design. At some point either cost or complexity or image quality goals become counter productive and those "dream" lenses stay dreams.

The point is, it's not easy and, if it was easy, they would have done it already.

Well since you put it that way, Oh
Thanks Ziggy for that advanced physics lesson I didn't realize it was quite that complicated. They can design, build and implement a camera (based on IR technology) that can see through solid objects, but yet can't overcome design obstacles of modern lens making.
Have a good afternoon
Jim...

Robinivich

It should be pointed out that it would be significantly easier without the mirror in there, this is part of how the camcorder crowd gets by. A rangefinder with no mirror would be easier to design for, if you could avoid retrofocal design, however "Easier" should not be confused with "Easy." If it weren't a retrofocal design things get way simpler, but as ziggy pointed out, FOV comes into it.

If you have no requirements for price, weight, mirror, etc, then amazing zoom range is always possible, eg http://www.canon.com/bctv/products/digi100xs.html but even then, notice that it's not that wide on the wide end, only 50.6X39.1 degrees. It IS an f/1.7-4.7 though (it also weighs a mere 23 kg)

TylerW

ziggy53 wrote:

For now, quality and convenience are still fairly mutually exclusive.

I think this will always be true, if for no other reason than tomorrow's technology will push our current ideals of quality and convenience into the stratosphere.

Think of what a 18-200 would have been like 40 years ago. :puke

rpcrowe

I use a 12-70mm focal range...
There is absolutely no problem having a 12mm to 70mm focal range available - with a constant f/4 aperture from 12mm to 24mm and a constant f/2.8 aperture from 24-70mm, all with outstanding IQ and great focusing ability! And, it doesn't take the optical equivalent of rocket science to deliver this...

I use this setup all the time.

It simply consists of two bodies, one with the 12-24mm f/4 Tokina and the other with the 24-70mm f/2.8L. Stick these two body/lens combos on an OPTECH Reporter strap and you have the best of both worlds - a wide focal range with great IQ and a decent f/stop with great autofocus ability available.

Actually, with the prices of used lightweight XT bodies really at rock-bottom prices, it is not terribly expensive, nor terribly heavy to use a two body setup.

Another benefit - you won't lose an important photo opportunity if one body goes down.

Addendum: Another lens duo that I absolutely love - especially for travel - is the 17-55mm f/2.8 IS and 70-200mm f/4L IS lenses.

claudermilk

Same here, except I go a slightly cheaper route--a single body & a ThinkTank Lens Changer bag. Lens swaps don't take too long.