Older lenses on film vs. on digital

We often hear how many older lenses designed for film cameras don't cut it on modern digital sensors. And indeed my own experiments have sometimes (though not always) confirmed that.

One such lens that has relatively weaker performance in the corners when shot on digital sensors is the Pentax-A 645 55mm f/2.8 (I used it on the GFX 50R for some time).

Now, that is normally interpreted as implying that such older lenses were never "very good" to begin with, and that the modern high-resolution digital sensors just make it easier to see the flaws.

The reasoning is that, if the lens exhibits poor sharpness even in the corners of a CROPPED 44x33mm sensor (55mm diagonal), it must have been very flawed indeed when deployed to cover its full intended 56x42mm (70mm diagonal) image area.

And yet, I've now used the very same lens on its native Pentax 645 film camera, and guess what? It performs beautifully all the way into the corners.

Here's a recent shot, hand-held at medium apertures (f/8, IIRC), on Kodak Ektar 100, scanned to a 36MP 16-bit TIFF with an Imacon Flextight V, and then converted to 99%-quality JPEG for upload.

Go look at the far corners. I cannot detect any of the "smearing" that ought to have been there, based on its performance on the cropped digital GFX sensor.



Interesting...

Marco

marco1974 said:

We often hear how many older lenses designed for film cameras don't cut it on modern digital sensors. And indeed my own experiments have sometimes (though not always) confirmed that.

Click to expand...

What Ive found that the older Minolta lenses from 40+ year ago have a corner glow/softness from spherical aberration wide open but when stopped down to f/8 and smaller, are at par with most modern digital lenses.

marco1974 said:

One such lens that has relatively weaker performance in the corners when shot on digital sensors is the Pentax-A 645 55mm f/2.8 (I used it on the GFX 50R for some time).

Now, that is normally interpreted as implying that such older lenses were never "very good" to begin with, and that the modern high-resolution digital sensors just make it easier to see the flaws.

Click to expand...

I think the older lenses were better in terms of pure optical performance because they were designed for image fidelity without relying on software correction. I feel that modern digital lenses are designed to be made economically with compromises that are later corrected via software. I know, I know its the end result that matters yada yada, but you get my point.

marco1974 said:

The reasoning is that, if the lens exhibits poor sharpness even in the corners of a CROPPED 44x33mm sensor (55mm diagonal), it must have been very flawed indeed when deployed to cover its full intended 56x42mm (70mm diagonal) image area.

And yet, I've now used the very same lens on its native Pentax 645 film camera, and guess what? It performs beautifully all the way into the corners.

Click to expand...

Maybe the slight curvature in the film vs the flat digital sensor makes a difference.

marco1974 said:

Here's a recent shot, hand-held at medium apertures (f/8, IIRC), on Kodak Ektar 100, scanned to a 36MP 16-bit TIFF with an Imacon Flextight V, and then converted to 99%-quality JPEG for upload.

Go look at the far corners. I cannot detect any of the "smearing" that ought to have been there, based on its performance on the cropped digital GFX sensor.



Interesting...

Click to expand...

I love this pic. Beautiful colors and corner to corner sharpness.

marco1974 said:

Marco

Click to expand...

marco1974 said:

Go look at the far corners. I cannot detect any of the "smearing" that ought to have been there, based on its performance on the cropped digital GFX sensor.

Click to expand...

I don't see any detail at all in the corners. This isn't an image I'd use to test this.

Based on limited experience, I think some older lenses are competitive with the best modern ones, especially if you're stopping down a bit.

State of the art lenses today stand out by performing brilliantly over a wider range of conditions (on and off-axis, wider apertures, a range of magnifications). They're also more likely to have some non-optical advantages. Sometimes they're compact and light (though sometimes very much not). Sometimes they have fast, quiet autofocus. Etc.

My Fuji GF lenses are optically the best I've used, by far ... after years with Nikon, many years with Schneider large format lenses, and a stint with Hasselblad film lenses. And a couple of adapted Pentax 645 lenses. That's far from all the old lenses, but a decent sample.

I think the thickness of the protective glass and the "hot mirror" on digital sensor units refract the light in a way that designed-for-film lenses did not have to account for. That's likely part of the discrepancy.

Best wishes,
Sterling
--
Lens Grit

SterlingBjorndahl said:

I think the thickness of the protective glass and the "hot mirror" on digital sensor units refract the light in a way that designed-for-film lenses did not have to account for. That's likely part of the discrepancy.

Click to expand...

that makes a lot of sense.

SterlingBjorndahl said:

Best wishes,
Sterling
--
Lens Grit

Click to expand...

I'm one of the people who has used the Pentax-A 645 55mm f/2.8 on a GFX 50R and concluded in comments and reviews that it's not terrific at the edges until f/11.

I just pulled up some of the files I would have used in coming to that conclusion.

I have a standard flat target that I try all lenses on. The 55/2.8 is weak in the corners at f/8 and getting usable on a 33mm x 44mm unshifted frame. In comparison, the Pentax 67 55mm f/4 and the Pentax-A 645 45-85mm f/4.5 at 55mm are both excellent across the flat target, right into the corners, at f/8.

I also have a standard outside long distance scene that I use. One thing I noticed is that the lens may have a decent amount of field curvature. When focused at infinity, parts of the scene that are rendered in focus with other lenses at that focal length are a bit soft. However, when I pulled the point of focus forward in the scene, they sharpened up nicely.

But... at the same time as I made this f/8 long distance scene with the 55/2.8, I shot the same scene, focusing in the same places, with the A 45-85mm f/4.5. Once again, the big zoom is a lot better.

So for what it's worth, I remain confident that the A 55/2.8 is OK for a lot of purposes, but if you need a solid, reliable performer in Pentax 645, the zoom is better.

What's missing in this story is the context. I judge lenses not simply on how they do in the central part of the frame. Rather, I'm also very interested in how they do towards the edges, because I use these lenses for movements. So my gripe with the Pentax-A 645 55/2.8 is mostly that it doesn't perform well shifted. If I needed a small, light lens and didn't use movements, and if I wasn't too fussed about edges and corners, it would do the job. But that's not what I need.

Paradoxically, when I used the A 55/2.8 on a Sony A7R with Mirex Tilt-shift adapters, I was perfectly content. This was a long time ago, but I'm pretty sure there's shift involved. The top of the tower, which I think I shifted into, has plenty of detail.


SMC Pentax-A 645 55mm f/2.8 on a Sony A7R using Mirex tilt-shift adapters

I've never used the lens on a film Pentax 645, so can't comment on that. I will say that when we're adapting lenses designed for one purpose and using them for another, we should expect the unexpected.

I'm a glass half full person on this issue. I'm just delighted that I've been able to put together an affordable and high quality set of lenses that allow me to have camera movements with a 100 MP modern digital sensor.

yogi4fitness said:

I think the older lenses were better in terms of pure optical performance because they were designed for image fidelity without relying on software correction...

Click to expand...

Older lenses, just like newer lenses, range from dismal to remarkable. Software correction incorporated into modern lens designs adds an additional digital element which lens designers can use for improving performance beyond what can be accomplished thru optical methods alone.

yogi4fitness said:

I feel that modern digital lenses are designed to be made economically with compromises that are later corrected via software.

Click to expand...

Your understanding of how and why software correction is used in lens designs is too narrowly limited. Software correction makes possible a variety of new lens designs with enhanced performance across a range of price points.

Under correcting in one or more specific areas, which can be well corrected digitally (distortion, vignetting, lateral chromatic aberration), can assist in achieving better correction in specific design areas which need to be controlled optically. Software correction opens doors to designing lenses in ways which would otherwise be impossible or compromise performance and other design goals.

To quote one example from Edmund Optics:

Selective Vignetting to Gain Performance - ADVANCED

Vignetting is often used to maximize the resolution of a lens design across the entire image circle. Since it is more difficult to direct the rays that create the edge of an image to the desired location on a sensor, higher resolution objects are generally more difficult to reproduce at the edge of the image than at the center. Rays that end up on the wrong pixel will degrade the image at that location; one way to manage this is to eliminate these rays from the system. If the undesired rays do not make it to the sensor, they cannot degrade the image. Removing these misdirected rays, however, reduces relative illumination.

yogi4fitness said:

I know, I know its the end result that matters yada yada, but you get my point.

Click to expand...

TechTalk2 said:

yogi4fitness said:

I feel that modern digital lenses are designed to be made economically with compromises that are later corrected via software.

Click to expand...

Your understanding of how and why software correction is used in lens designs is too narrowly limited. Software correction makes possible a variety of new lens designs with enhanced performance across a range of price points.

Under correcting in one or more specific areas, which can be well corrected digitally (distortion, vignetting, lateral chromatic aberration), can assist in achieving better correction in specific design areas which need to be controlled optically. Software correction opens doors to designing lenses in ways which would otherwise be impossible or compromise performance and other design goals.

To quote one example from Edmund Optics:

Selective Vignetting to Gain Performance - ADVANCED

Vignetting is often used to maximize the resolution of a lens design across the entire image circle. Since it is more difficult to direct the rays that create the edge of an image to the desired location on a sensor, higher resolution objects are generally more difficult to reproduce at the edge of the image than at the center. Rays that end up on the wrong pixel will degrade the image at that location; one way to manage this is to eliminate these rays from the system. If the undesired rays do not make it to the sensor, they cannot degrade the image. Removing these misdirected rays, however, reduces relative illumination.

Click to expand...

The 38 V also suffers from a fair amount of mechanical vignetting wide open. I think that's a consequence of making the lens small.

TechTalk2 said:

yogi4fitness said:

I feel that modern digital lenses are designed to be made economically with compromises that are later corrected via software.

Click to expand...

Your understanding of how and why software correction is used in lens designs is too narrowly limited.

Click to expand...

I admire your almost Chatbot like efficiency at gleaning (selective) information from the web and regurgitating it. I suppose a lot of time is spent in that armchair.

TechTalk2 said:

Software correction makes possible a variety of new lens designs with enhanced performance across a range of price points.

Click to expand...

Esp. at lower cost points. Notice how I didn't say "price point" because some of these lenses are expensive and yet utilize software correction to hit a certain weight and size point.

TechTalk2 said:

Under correcting in one or more specific areas, which can be well corrected digitally (distortion, vignetting, lateral chromatic aberration), can assist in achieving better correction in specific design areas which need to be controlled optically. Software correction opens doors to designing lenses in ways which would otherwise be impossible or compromise performance and other design goals.

To quote one example from Edmund Optics:

Selective Vignetting to Gain Performance - ADVANCED

Vignetting is often used to maximize the resolution of a lens design across the entire image circle. Since it is more difficult to direct the rays that create the edge of an image to the desired location on a sensor, higher resolution objects are generally more difficult to reproduce at the edge of the image than at the center. Rays that end up on the wrong pixel will degrade the image at that location; one way to manage this is to eliminate these rays from the system. If the undesired rays do not make it to the sensor, they cannot degrade the image. Removing these misdirected rays, however, reduces relative illumination.

Click to expand...

I call this cherry picking information to fit your narrative. By the above, carefully cherry picked paragraph, you wanted to sound like there is performance gained (over a well engineered lens) by applying vignetting, and this is far from the truth. Software correction simply compensates for a lens design that has been compromised to fit within a certain size and weight.

A few paragraphs down from the same article that you just regurgitated...

"Vignetting can also be purposefully designed into lenses in scenarios in which the effects from manufacturing tolerances adversely affect the control of rays, causing image degradation. The looser the tolerances on the lens, the more adverse these degrading effects can become, and tightening the tolerances is often not practical due to the increase in manufacturing cost. Often, a balance must be struck between reducing manufacturing cost and maintaining image quality. In cases where cost is a primary factor, vignetting must be utilized in an attempt maintain resolution across the field of view. Doing such will have an adverse effect on the illumination profile. Designing vignetting into a lens can be accomplished in a couple different ways: by purposefully designing the clear apertures of the individual lens elements such that they vignette severely off-axis rays, or by introducing mechanical apertures to block aberrated rays,..."

How do you reduce cost of manufacturing a lens? By opening up the tolerances, by limiting the apertures of the lens elements, and by introducing mechanical vignetting (as shown by Jim). And then applying the crutch of software to brighten up the areas affected by vignetting.

If you want to learn more about lens design and how vignetting can be used in lens design to cut costs in imaging systems, here is a good article by the makers of lens design software Zemax. I personally use Code V but the same design principles apply.


I am going to take a page from your playbook and regurgitate a snippet from this article:

"Vignetting factors are critical for avoiding an overdesigned system. Without vignetting factors, we would design a system that corrects the aberrations for the full-size beam over the entire field of view of the system. The result would be a lens system that is overcorrected; when the real aperture sizes are applied to the lenses, the vignetting from the real apertures would result in performance at the corner of the fields that is better than what is strictly necessary."

Over-designing is how you attain superior performance corner to corner and not via software correction. It costs more and it also creates bulkier and heavier lenses. I believe some of the older lenses from the film era were over-engineered because they couldn't rely on the crutch of software correction. This is why some of the old Minolta Rokkor lenses vignette less than the new XCD lenses on a 44x33 sensor even tho they were designed for FF.

Again stealing a page from your playbook and regurgitating the summary of this article:

"In this new world of onboard image processing, optical designers must understand vignetting and use vignetting factors effectively to create smaller and lighter optical systems."

This is what Jim and I, and many others here believe, that the new XCD lenses are using vignetting for size and weight reductions (and cost reductions).

Peace out.

yogi4fitness said:

I admire your almost Chatbot like efficiency at gleaning (selective) information from the web and regurgitating it. I suppose a lot of time is spent in that armchair.

Click to expand...

You can continue to insult me all you like. No one will object to you doing so and I'm immune.

It also won't change the fact that every lens begins with setting parameters for the design specifications. Those will include size, weight, and cost in order to fit the intended use and budget. Making lenses which are large, heavy, and unaffordable by the intended market is not what manufacturers or their customers generally desire.

Software correction provides an additional element enabling better lens performance in meeting all of the designers intended goals. It's a tool not a "crutch". To stubbornly ignore the benefits available would be of zero benefit to anyone including the end user.

Fortunately, lens manufacturers and designers use all of the tools available to them today, including software correction, and we get better lenses as a result.

TechTalk2 said:

yogi4fitness said:

I admire your almost Chatbot like efficiency at gleaning (selective) information from the web and regurgitating it. I suppose a lot of time is spent in that armchair.

Click to expand...

You can continue to insult me all you like. No one will object to you doing so and I'm immune.

It also won't change the fact that every lens begins with setting parameters for the design specifications. Those will include size, weight, and cost in order to fit the intended use and budget. Making lenses which are large, heavy, and unaffordable by the intended market is not what manufacturers or their customers generally desire.

Software correction provides an additional element enabling better lens performance in meeting all of the designers intended goals. It's a tool not a "crutch". To stubbornly ignore the benefits available would be of zero benefit to anyone including the end user.

Fortunately, lens manufacturers and designers use all of the tools available to them today, including software correction, and we get better lenses as a result.

Click to expand...

You are preaching to the choir.

yogi4fitness said:

Again stealing a page from your playbook and regurgitating the summary of this article:

"In this new world of onboard image processing, optical designers must understand vignetting and use vignetting factors effectively to create smaller and lighter optical systems."

Click to expand...

That looks like sound advice.

yogi4fitness said:

This is what Jim and I, and many others here believe, that the new XCD lenses are using vignetting for size and weight reductions

Click to expand...

They have and software correction combined with internal optical components using modern aspheric and ED (Extra-low Dispersion) glass elements enable the production of a compact lightweight lens series exhibiting high performance.

yogi4fitness said:

(and cost reductions).

Click to expand...

No one wants them to cost more than they do already.

A good exercise would be to compare a good Hasselblad lens from the film era (maybe the standard 80mm) with the current equivalent. Do all the corrections / enhancements you can to both. And see which one ends up better.

My money's on the new one. From what I've seen, designers compromise on qualities that are easily corrected in post (distortion, lateral CA, vignetting) to better correct all the ones that aren't. Makes sense to me.

paulraphael said:

A good exercise would be to compare a good Hasselblad lens from the film era (maybe the standard 80mm) with the current equivalent. Do all the corrections / enhancements you can to both. And see which one ends up better.

My money's on the new one. From what I've seen, designers compromise on qualities that are easily corrected in post (distortion, lateral CA, vignetting) to better correct all the ones that aren't. Makes sense to me.

Click to expand...

Yes please go out and shoot and post more pics. The theoretical armchair analysis is pointless and a waste of time.

TechTalk2 said:

yogi4fitness said:

I admire your almost Chatbot like efficiency at gleaning (selective) information from the web and regurgitating it. I suppose a lot of time is spent in that armchair.

Click to expand...

You can continue to insult me all you like. No one will object to you doing so and I'm immune.

It also won't change the fact that every lens begins with setting parameters for the design specifications. Those will include size, weight, and cost in order to fit the intended use and budget. Making lenses which are large, heavy, and unaffordable by the intended market is not what manufacturers or their customers generally desire.

Software correction provides an additional element enabling better lens performance in meeting all of the designers intended goals. It's a tool not a "crutch". To stubbornly ignore the benefits available would be of zero benefit to anyone including the end user.

Fortunately, lens manufacturers and designers use all of the tools available to them today, including software correction, and we get better lenses as a result.

Click to expand...

Cat's-eye bokeh from mechanical vignetting can't be corrected in post.

yogi4fitness said:

Yes please go out and shoot and post more pics. The theoretical armchair analysis is pointless and a waste of time.

Click to expand...

No one is forcing you to read it or to continue with the same repetitive insults.

You're free to use your time and voice as you see fit — as am I and everyone else.

yogi4fitness said:

paulraphael said:

A good exercise would be to compare a good Hasselblad lens from the film era (maybe the standard 80mm) with the current equivalent. Do all the corrections / enhancements you can to both. And see which one ends up better.

My money's on the new one. From what I've seen, designers compromise on qualities that are easily corrected in post (distortion, lateral CA, vignetting) to better correct all the ones that aren't. Makes sense to me.

Click to expand...

Yes please go out and shoot and post more pics. The theoretical armchair analysis is pointless and a waste of time.

Click to expand...

Are you talking to me or to yourself? I'm not the one neck deep in any armchairs here.

Personally, I stopped using a film Hasselblad years ago. Using Fuji with current glass now. It's dead obvious which optics are better. No armchair needed.

paulraphael said:

yogi4fitness said:

paulraphael said:

A good exercise would be to compare a good Hasselblad lens from the film era (maybe the standard 80mm) with the current equivalent. Do all the corrections / enhancements you can to both. And see which one ends up better.

My money's on the new one. From what I've seen, designers compromise on qualities that are easily corrected in post (distortion, lateral CA, vignetting) to better correct all the ones that aren't. Makes sense to me.

Click to expand...

Yes please go out and shoot and post more pics. The theoretical armchair analysis is pointless and a waste of time.

Click to expand...

Are you talking to me or to yourself? I'm not the one neck deep in any armchairs here.

Click to expand...

That was a general statement for all to go out and shoot more pics (including me). Sorry if it sounded like it was directed at you.

paulraphael said:

Personally, I stopped using a film Hasselblad years ago. Using Fuji with current glass now. It's dead obvious which optics are better. No armchair needed.

Click to expand...

I am also a Fuji user and mainly in the system for its stellar glass. It may be big and bulky and ugly with over-engineered optics but boy does it produce glorious images.

JimKasson said:

TechTalk2 said:

yogi4fitness said:

I admire your almost Chatbot like efficiency at gleaning (selective) information from the web and regurgitating it. I suppose a lot of time is spent in that armchair.

Click to expand...

You can continue to insult me all you like. No one will object to you doing so and I'm immune.

It also won't change the fact that every lens begins with setting parameters for the design specifications. Those will include size, weight, and cost in order to fit the intended use and budget. Making lenses which are large, heavy, and unaffordable by the intended market is not what manufacturers or their customers generally desire.

Software correction provides an additional element enabling better lens performance in meeting all of the designers intended goals. It's a tool not a "crutch". To stubbornly ignore the benefits available would be of zero benefit to anyone including the end user.

Fortunately, lens manufacturers and designers use all of the tools available to them today, including software correction, and we get better lenses as a result.

Click to expand...

Cat's-eye bokeh from mechanical vignetting can't be corrected in post.

Click to expand...

True. I don't think it was suggested that you could do that, but it's true nonetheless.

For a good overview of the various types of vignetting, including the cat's eye effect, the link below is useful.

https://web.archive.org//http://toothwalker.org/optics/vignetting.html