Zoom...today vs yesterday

[Eric Rose]

One thing that must be kept in mind is that many modern zooms are designed for digital cameras. The distortion control is now done via software in the camera rather than in the glass as it was done prior to digi. If you were to find an adapter to use a modern Sony zoom designed for the A7 series on a film camera I would expect to see a lot of distortion. Where digi cameras can use the older film era zooms and primes with stunning results, the reverse is not always true. Legacy companies like Nikon and Canon probably still try and do most of their distortion control in-lens but companies like Sony don't have that burden. The non-distortion controlled zooms are cheaper to design and produce thus allowing the manufacturers to spend R&D money on producing "faster" lenses. The new holy grail.

Now back to the OP. I used/still use zooms extensively on my 35mm film cameras. However if the situation allowed primes to be used I would always choose them first. My biggest problem with zooms has always been barrel distortion and flare.

 

[Dali]

I have a few zooms for both my Canon and Nikon. Some are fairly wide...not sure exactly, somewhere in the 50-150 range maybe.? I never use them.
I am about to send some Nikon lens to get the AI conversion, one of them was a Zoom, but i decided there was no need to, as i do not use it anyway. So my Zooms sit on a shelf.
Fast forward to today's cameras, and Zooms seem to be Very Popular. On other forums i belong to (non photography related) when a guy asks (in the off-topic section of the forum) for a lens recommendation, the answers are always full of Zoom suggestions.
I guess that makes me wonder a few things:
Are Zooms of today "better" made than a zoom made circa 1975.?
Do you guys use a Zoom very often with your Film SLR.?
Thank You

Answer to your first question: In general, yes, recent zooms are better then 40 years old ones.

Answer to your second question: No, I don't use zooms often for 3 reasons: First, the max aperture is often limited compare to prime lenses. Second, a zoom is somewhat bulky compared to a prime lens. Remember, with a zoom, you carry the whole focal range even if you don't need it! Third, I mostly shoot 28 and 50mm and zoom lens choice is rather limited in this range.

Side note: I started photography decades ago with zoom lenses but I switched to prime lenses when I found out it gives me more freedom. I know, it seems contradictory but not be tempted to change of focal length all the time is a real advantage (not to mention what I answered to your question #2).

 

[benveniste]

If you want to control perspective, a zoom lens is a hindrance not a help.

Why?

"Hindrance" may not be quite the right word, but I've fallen into the "zoom trap" before, will probably do so again, and I doubt I'm alone in this. The "zoom trap" occurs when a photographer relies too much on the zoom and not enough on their feet. The victim ends up taking multiple shots from the same position at different focal lengths instead of choosing the position based on the right perspective for the shot.

It's not the fault of the lens, but rather a poorly calibrated nut behind the finder. One of the several reasons I still occasionally shoot medium and large format is that it forces me to slow down and concentrate on pre-visualization and composition, if only to save my poor wallet!

 

[CMoore]

I have only been a member a short time, but....."We" should compile a list of Mark Twain, literary gems like this one, from benveniste - "It's not the fault of the lens, but rather a poorly calibrated nut behind the finder."
AND.....that IS an interesting observation about Zooms. The non-zoom user deals with perspective change, while the zoom user does not. Obviously, when used for extreme distance changes, a person would not be zooming with their feet. But, something like a change from 35 to 45 would involve what he is talking about.

 

[Nodda Duma]

But, recently optical designers become lazy: they are using editing software to correct distortions and aberrations in a picture rather than trying to correct it in the lens in the first place. That and the fact the newer zooms are electronically controlled doesn't bode well for durability.

I strongly disagree. For one, this sort of mapping is incredibly complex and labor-intensive -- much more so than the comparative effort to correct the aberrations.

For another, the push in zoom design is to combine optical zoom with electronic zoom to reduce the optical zoom range to provide better overall image quality and reduction in mechanical complexity over optical zoom alone. The exception is in thermal zoom objectives where the limited resolution of state-of-the-art thermal imagers does not support electronic zoom. There the push is to reduce overall length...this is the cutting edge market for zoom designs.

The ultimate goal is full electronic zoom provided with fixed focal length for best optical correction. This I've done several times now and the image quality is pristine...much different than the cheap electronic zoom found in the consumer world.


The other push is in computational imaging, which allows less complex (cheaper) optics for greater processing power. This is certainly not laziness...it is very difficult to do correctly and is a hot topic of research in the optical engineering community.

Electrically controlled zooms may be less durable in the consumer market, but that is the only place.

 

[Ricardo Miranda]

I strongly disagree. For one, this sort of mapping is incredibly complex and labor-intensive -- much more so than the comparative effort to correct the aberrations.

For another, the push in zoom design is to combine optical zoom with electronic zoom to reduce the optical zoom range to provide better overall image quality and reduction in mechanical complexity over optical zoom alone. The exception is in thermal zoom objectives where the limited resolution of state-of-the-art thermal imagers does not support electronic zoom. There the push is to reduce overall length...this is the cutting edge market for zoom designs.

The ultimate goal is full electronic zoom provided with fixed focal length for best optical correction. This I've done several times now and the image quality is pristine...much different than the cheap electronic zoom found in the consumer world.


The other push is in computational imaging, which allows less complex (cheaper) optics for greater processing power. This is certainly not laziness...it is very difficult to do correctly and is a hot topic of research in the optical engineering community.

Electrically controlled zooms may be less durable in the consumer market, but that is the only place.

I'm sorry, I'm only a photographer.
I only use film and lenses made for it.
I appreciate the work that goes into creating an excellent lens, not the software that goes with modern ones, less the electronic zooms. They're not needed nor they are desirable in my cameras.
Thanks for the explanation.

 

[Nodda Duma]

No worries. The reality is that modern zooms - like almost any other modern lens - aren't designed for film, and haven't been for the past 10-15 years. Nor are zoom designs solely confined to or even driven by the consumer photography market. So any serious discussion about modern zoom lens design cannot ignore the fact there are no "modern zooms for film".

What has happened is that the constraints that were placed by film use on optical design have been lifted, and the result is a trend towards lens design as a more integrated part of the system. The trend has been obvious in its impact on zoom design requirements, transferring some or all of the zoom range in a system to the FPA in order to simplify the mechanics and improve reliability.

 

[Ricardo Miranda]

What has happened is that the constraints that were placed by film use on optical design have been lifted

How about the constraints posed by devices that need the light path to be as straight and parallel as possible? That was the drive for the new (2002 onwards) lenses.

 

[macfred]

I have 3 zoom lenses for Nikon 35mm :
- Nikkor 25-50mm f/4.0 AiS
- Nikkor AF 35-70mm f/2.8
- Nikon Series E 75-150mm f/3.5
Nothing wrong with those lenses - they all have a great image quality.
The old and clunky 25-50mm f/4 is a superb lens which still can bear comparison with modern lenses.
The AF 35-70mm f/2.8 is not too bad in comparison with the current 24-70mm f/2.8 AFS (and much smaller and less expensive).

 

[Poisson Du Jour]

No worries. The reality is that modern zooms - like almost any other modern lens - aren't designed for film, and haven't been for the past 10-15 years. Nor are zoom designs solely confined to or even driven by the consumer photography market. So any serious discussion about modern zoom lens design cannot ignore the fact there are no "modern zooms for film".

What has happened is that the constraints that were placed by film use on optical design have been lifted, and the result is a trend towards lens design as a more integrated part of the system. The trend has been obvious in its impact on zoom design requirements, transferring some or all of the zoom range in a system to the FPA in order to simplify the mechanics and improve reliability.

This is emphatically not correct. "Modern zooms" from around 1998 all the way to the present, Canon being one manufacturer that comes to mind, work correctly and as intended on either digital or analogue bodies. The zooms know what body they are attached to and have algorithms on board to match optics to meter and system metrics.The last thing Canon (and others) want to do is effectively alienate users who chose not to use digital devices but wish to use whatever zooms take their fancy on whatever camera bodies (analogue) they chose, but couldn't because "the zoom isn't optimised for film". It's simply not true. There are separate zooms made for and optimised for digital (different mount), but consumer photographers are not left out of the equation if they want to use a zoom on e.g. an EOS 1V or an EOS 5 or slap it on a 5DMkIII. No bid deal. It will work. So too will all of the L-series zooms. And by referring to those, I am going back two decades-plus. So they were made for film, and nowadays can be seen in use on digital or analogue bodies.

 

[mtjade2007]

A smart phone can produce excellent images too. To me digital cameras produce images almost effortlessly by the photographers. Most the effort needed done by the photographer is replaced by the chips in the camera instead. To me a digitally captured image is more like a computer drawn image. You point the camera to a scene, press a button and a blurry raw image is left on the CCD. The computer then takes the image off from it and applies a sharpening algorithm on it, calculates the approximated color values for each of the pixels then a final image is produced. There are other algorithms that can be applied to the final images as well with no knowledge about them in the photographer's mind. A hobby like that that requires little effort to produce a final result is not for me. What counts in photography is how much creativity there is in the image by the effort of the photographer.

 

[Diapositivo]

I strongly disagree. For one, this sort of mapping is incredibly complex and labor-intensive -- much more so than the comparative effort to correct the aberrations.

For another, the push in zoom design is to combine optical zoom with electronic zoom to reduce the optical zoom range to provide better overall image quality and reduction in mechanical complexity over optical zoom alone.

I don't know about "aberrations" overall, as it is not possible, or very computationally intensive and not necessarily effective, to try to correct a broad range of aberrations. So I agree with you that nowadays "aberrations" cannot be easily corrected through post-capture processing.

In the specific field of distortion correction though, things get pretty easy and that's what I think the original comment might have referred to.

I have a Photoshop Plug-in, ePaperPress PTLens, which I paid if memory serves around €50, several years ago.
It has hundreds of "canned" profiles for a large number of makers and models. For each lens, it has distortion correction for various focal lenghts. It reads the focal lenght from the EXIF profile automatically.
It works very, very well with the little distortion the Zeiss lens in my Sony DSC-R1 (a non-film camera with a very good, fixed zoom lens) shows at every focal lenght.

It allows the user to build his own profile for each lens he uses. If distortion correction is important for your work, you profile your manual-focus film-camera lenses (you makeyour own profiles) and I'm sure it will make a jolly good job of remedying distortions in a hybrid setup (scans, digital negatives).

To give you a context, I do sell pictures through agencies and I do "pixel-peep" my images, at 100% pixel level, so I'm not talking "software toys" here, but serious stuff.

Regarding the second idea, "the push in zoom design is to combine optical zoom with electronic zoom to reduce the optical zoom range to provide better overall image quality and reduction in mechanical complexity over optical zoom alone", if I correctly get what you mean, I very strongly disagree.

"Electronic zoom" is marketing bullstuff for "cropping". "electronic zoom" does not reduce optical zoom range, nor increases it, nor increases or decreases mechanical complexity, or optical complexity, of the zoom. The zoom is what it is in its opto-mechanical nature, and the "electronic zoom" is just an enlargement of the central part of the picture. In this sense, you can have "electronic zooms" also with your turn-of-century Petzval

 

[Nodda Duma]

Distortion correction inherently reduces resolution proportional to the amount of distortion being corrected. In the broader range of design applications where the ability to resolve objects is a typical requirement, this type of post-processing can sometimes be a detriment but many times is helpful. Foveal applications are an example, where you intentionally introduce distortion to achieve a wider field of view in a smaller package.

In reality, distortion is not difficult to correct in the optical design of a zoom.

Regarding electronic zoom (and I apologize for the tangent but you guys have me talking shop): As implemented in consumer products to get away with using the cheapest array possible..yes, it is just marketing bullshit. However, outside this "cheap" realm, there are very real advantages to electronic zoom and it is not considered bullshit. In particular is the significant reliability and cost advantages of replacing moving parts with fixed parts and simpler optics.

The key is to use very large arrays to ensure sufficient resolution at the smallest region of interest (or narrowest field of view setting). For example, we developed a 2X electronic zoom camera which maps to a 1280x1024 display. We selected a focal plane array with a native resolution of 2560 X 2048, so for the narrow FOV we map the central 1280x1024 directly to the array with no resolution loss to interpolation. The wide field setting bins nearest neighbor pixels to map the full 2560x2048 to the 1280x1024 display. Very simple concept and the approach does not sacrifice image quality in any way. So not bullshit. The lens was then designed as a fixed focal length optic and we saved significant cost and complexity...and the camera can still resolve the warts on a gnat's ass two football fields away. We met the performance requirement and did it at a cost, size, and weight that far exceeded expectations, so the customer was very happy. This is the correct way to implement electronic zoom. Note in this example the design is "good enough" to meet the requirement without providing unnecessary performance: Over-designing can be as disastrous as under-designing. That principle is at the core of good design practices.

 

[mr rusty]

My view is that you use what you want to use on the day, and pretty much anything is good enough to be capable of taking a good photo. Whether the photographer or the printing is any good is another matter, but seems to me that most lenses can get more than enough detail onto the negative. Sometimes I use prime, sometimes zoom, but I don't really worry about any quality difference.

I've posted this before, but it shows what you can get from a bog-standard cheap zoom from 80's - in this case the 35-70 F4 zuiko - hardly the last word in expensive glass. The extract from the image shows the detail in the signpost, which you can't even see with the naked eye at normal viewing of the full image. What more do you need?

(there was a url link here which no longer exists)

 

[AgX]

Distortion correction inherently reduces resolution proportional to the amount of distortion being corrected.

Are you referring here with distortion correction to computional or optical means?

 

[mklw1954]

Minolta started making excellent zoom lenses in 1977: the 35-70mm f3.5 and 75-200mm f4.5. In 1981 they started making the 70-210mm f4 zoom. I use the 35-70mm and 70-210mm zooms and image quality is at least as good as with prime lenses. Leica used the Minolta designs in their Leitz 35-70mm and 75-200mm zooms. Minolta made their own lenses, even down to making the glass used.

 

[benjiboy]

Zoom lenses! Why would you ever wish to own one, let alone use it. They may make good door stops.

There's nothing like having an open mind, before you advocate their use as doorstops have you ever actually owned a zoom lens Clive ?

 

[Paul Howell]

Minolta started making excellent zoom lenses in 1977: the 35-70mm f3.5 and 75-200mm f4.5. In 1981 they started making the 70-210mm f4 zoom. I use the 35-70mm and 70-210mm zooms and image quality is at least as good as with prime lenses. Leica used the Minolta designs in their Leitz 35-70mm and 75-200mm zooms. Minolta made their own lenses, even down to making the glass used.

I have both the 35 to 70 3.5 and 70 to 210 F4 and while some very good lens I don't think are quite as sharp as Minolta 50 1.7 35 2.8 or 28 2.8 AF lens. I don't have any Minolta long primes in AF so cant say that the 70 to 210 is not as sharp as the 85, 100, or 135 primes.

 

[Nodda Duma]

Are you referring here with distortion correction to computional or optical means?

Sorry, Computational.

 

[Diapositivo]

For example, we developed a 2X electronic zoom camera which maps to a 1280x1024 display. We selected a focal plane array with a native resolution of 2560 X 2048, so for the narrow FOV we map the central 1280x1024 directly to the array with no resolution loss to interpolation. The wide field setting bins nearest neighbor pixels to map the full 2560x2048 to the 1280x1024 display. Very simple concept and the approach does not sacrifice image quality in any way. So not bullshit. The lens was then designed as a fixed focal length optic and we saved significant cost and complexity...

Although I understand the logic behind this (a fixed focal lenght lens is much easier an endeavour than a bi-focal lens, so it greatly pays to avoid the zoom logic altoghether and go for a much simpler optical scheme) as a general logic it seems to me that it always is a cropping solution.

The sensor where the image is projected is capable of 2560 x 2048 pixels.
The "tele" position crops the the central part and only uses the central 1280 x 1024 pixels;
The "wide" position uses the entire 2560 x 2048 but, having a final display resolution constraint of 1280 x 1024 pixels, downscales the image to that resolution.

I certainly understand the logic of this specific application given the particular constraints: if only 2 focal lenghts are needed, and if the output resolution is costrained in any way to a specific resolution (1280 x 1024), the result gives two "virtual focal lenghts" of equal resolution and equal final quality (abstracting here from other optical problems such as diffraction etc.) with a simple fixed focal lenght optical scheme.

Yet, I don't see an easy extension of this logic to generic photographic purposes, where the output resolution is unconstrained and the focal lenght span is generally speaking larger. Ultimately, it seems to me a cropping solution, even though in this case the "maximum crop required" ends up with a design-defined satisfying resolution.

 

[Diapositivo]

Minolta started making excellent zoom lenses in 1977: the 35-70mm f3.5 and 75-200mm f4.5. In 1981 they started making the 70-210mm f4 zoom. I use the 35-70mm and 70-210mm zooms and image quality is at least as good as with prime lenses. Leica used the Minolta designs in their Leitz 35-70mm and 75-200mm zooms. Minolta made their own lenses, even down to making the glass used.

I have the 100-200 f/5.6 Minolta Rokkor. Quite a strange lens with an unusual optical scheme that was used in this lens, the Canon 100-200 f/5.6, the Nikkor 80 - 250 and not many other modern photographic zoom lenses.

The scheme is called "optical compensation zoom" in an article in Italian (that is the literal translation).
There is a fixed group near the film plane, where the diaphragm is placed, so no problems with aperture transmission; then there is a fixed movement of some lenses relative to the fixed group, the mobile groups are moved all together, and not independently as it happens in more modern, and mechanically more complicated zoom schemes.

The defects are obvious: the extension of the lens at 200mm is such that, nowadays, I would be sure to talk to the Police within 2 minutes of extending it in tele position;
The maximum aperture is evidently quite limited;
The distortion is quite embarrassing.

That said, that scheme allows to reach an overall high optical quality (distortion aside) for what I can see.

I use it rarely now because I often deal with architectural subjects so distortion for me is a bit of a problem.

That said, I think it's a zoom lens that, as far as optical quality is concerned, is still very valid nowadays. The number of group/lenses is quite small, there's very good contrast. Contemporary zooms with independent lens movements show a higher lens count, higher flare.

Optical scheme here: http://www.artaphot.ch/minolta-sr/objektive/198-minolta-100-200mm-f56

An early and famous realization of this optical scheme, quoted by the article, is the Pan-Cinor designed by Cuvillier in 1949 and produced by French maker SOM-Berthiot.

 

[cliveh]

There's nothing like having an open mind, before you advocate their use as doorstops have you ever actually owned a zoom lens Clive ?

No and never intend to. I'm a prime man.

 

[benjiboy]

No and never intend to. I'm a prime man.

Your "door stop" remark then wasn't made from any personal experience or deep knowledge of using zoom lenses, but purely out of prejudice and your ignorance of their capabilitys

 

[cliveh]

Your "door stop" remark then wasn't made from any personal experience or deep knowledge of using zoom lenses, but purely out of prejudice and your ignorance of their capabilitys

Correct.

 

[fstop]

How about the constraints posed by devices that need the light path to be as straight and parallel as possible? That was the drive for the new (2002 onwards) lenses.

Why don't you just shoot with a pinhole camera?