[BarryE]

I pose this as a question, as I do not know the answer :

Has not the "tech" changed in recent years from the physical attributes of lens and sensor to the software ability of Super Resolution ? If a new FF Pentax of 48MP "requires" a new D-FA STAR+ series of lenses to keep up with sensor, you are then talking of '000 of dollars more for both camera and especially lenses. Who but a few would be interested in that investment when the same results can be achieved with a couple of hundred dollars worth of software upgrade ?

There may be scientific reasons for wanting a lens to have greater resolving power, but it does not feature in my workflow. The maximum image I post online is a little over 3MP. When I print at 300ppi I can do so at 24"x16" with the K1 images without any up-sampling. I regularly print at this size even when using old legacy lenses, like the M series 85mm f2.0. Maybe a "bench test" will tell me that the lens is not up to the task of providing the K1's sensor with enough resolution, but my printed results tell me otherwise.

Neither do I have the answer ... as ever

You're probably correct Super Resolution and its successors are changing the resolution question and rendering the power of a lens to resolve its image on increasing packed sensors somewhat meaningless.

The size of the sensor is the key part. However much software/firmware is thrown at "improving" a small sensor's ability to capture photons, the size of each pixel diode determines much of the image's native data & character.

Sensor size for FF between 24 and 36 mp is where the balance lies between resolution and DR etc. The K-1 prints fine up to and over A1 size and for less demanding output, eg internet, it's overkill. For greater resolution and DR go to MF.

So we have (@ Pentax) a vast range of lenses that have differing character and resolution that are offer us very nearly all we realistically want/need. I appreciate there are extreme use cases not covered, but we're catered for well enough with lenses that match the cameras and that to me is the point. Lenses that match the cameras we have now and for the near/medium future. Who knows where camera and lens technology will go next. Certainly not me. Much of the tech has probably reached a plateau and therefore getting on and using what we have is fine for me and probably a fair number of us here.

[yucatanPentax]

I have to admit I'm a bit baffled (and uneducated on the precise parameters sought), but ... aren't the charts useful mostly for bragging rights and flogging one another with who (which brand, which wallet) has "the best lens"?

Aren't the images created, viewed at full resolution, more indicative of what is possible with a given lens and sensor combination? Where I get lost is: What good do the "LW/PH" numbers do for anyone, if the images created are not pleasing to humans?

Superfine resolution isn't always desirable in photography. If there is a case where that's possible, maybe someone could explain it. I imagine scientific research (wasn't a Pentax DA 300mm just shot into space on a satellite or something?), medicine, dentistry, etc. But are people doing those with a Pentax APSC DSLR? (an endodontist recently showed me remarkably clear images from the inside of my tooth. They were taken with some sort of fiber optic microscope).

Sometimes, the transitions of light and shadow, the rendering of contrast, and other intangibles are what create a fine image. The lines per aren't often a part of it. The charts would be for a small minority of users, it would seem.

[GUB]

I have to admit I'm a bit baffled (and uneducated on the precise parameters sought), but ... aren't the charts useful mostly for bragging rights and flogging one another with who (which brand, which wallet) has "the best lens"?

Aren't the images created, viewed at full resolution, more indicative of what is possible with a given lens and sensor combination? Where I get lost is: What good do the "LW/PH" numbers do for anyone, if the images created are not pleasing to humans?

Superfine resolution isn't always desirable in photography. If there is a case where that's possible, maybe someone could explain it. I imagine scientific research (wasn't a Pentax DA 300mm just shot into space on a satellite or something?), medicine, dentistry, etc. But are people doing those with a Pentax APSC DSLR? (an endodontist recently showed me remarkably clear images from the inside of my tooth. They were taken with some sort of fiber optic microscope).

Sometimes, the transitions of light and shadow, the rendering of contrast, and other intangibles are what create a fine image. The lines per aren't often a part of it. The charts would be for a small minority of users, it would seem.

Not only that, but physics has a way of hindering those in pursuit of the ultimate sharpness - it is called depth of field. And the sharpest of lenses have no effect on it.
This is a file test shot from a newly acquired Ozeck 135mm 2.8 on the K-1. The Ozeck was an ultracheap m42 lens from the 1970s. But surprisingly sharp here wide open.
What I have done is applied a Edge Detect filter to the image. The white areas are where the app detects contrast. And only in the whitest of the white would you see any difference between a sharp and a less sharp lens. And, yes, this sharpness to define the subject is important, but so is the qualities other 95% of the image that a lens sharpness would make absolutely no difference to.
So this brings us to the catch 22 of those obsessed with the sharpness. They would have to stop their lens down to say f8 - f11 to fight this depth of field issue and utilise their special lens in a way that makes it indistinguishable from any other stopped down lens including those ones that are softer wide open.

[Rondec]

There are different types of reviews out there. There are reviews that focus on real world shooting experience (I think PF reviews fall into this category) and then there are a few reviews that focus on performance on test beds.

To me, the second type of review, while interesting, doesn't tell me a whole lot. It is enough to say that a lens has soft borders wide open and sharpen up by f2.8 or f4. As long as there are uncorrected images available to look at, I can see what I need to to make decisions about a lens purchase.

I will say that sharpness is not really a factor in shooting for me, but contrast, micro contrast, and flare resistance are. The DA 15 limited has been mentioned in this thread and it really has poor border performance till you stop down to f9, but it has excellent contrast and flare resistance and as such, it can be an excellent performer in the field, particularly if you can use a tripod.

[Digitalis]

I see lots of lens reviews on the site, but why aren't there any real lens measurements for any lenses.

What we have here on Pentax forums are many reports on individual experience with K mount lenses and cameras out in the field, which in the end is where the gear we buy is intended to be used. Not everyone is a formally trained optical testing technician here*.

im interested in how sharp lenses are, not vague pictures of test charts.

You're interested in numbers: I get that. But MTF charts can only tell you so much about a lens. MTF charts cannot tell you anything about flare handling characteristics, colour reproduction, physical aspects of a lenses operation - all of which are important variables to take account of when making a purchasing decision.

isn't it possible to determine the lines per inch that the full size K-1 and APS-C K-3 sensors can achieve, then measure every lens of some kind of lens bench, and determine the real resolving power at all the different combinations of aperture and position across the image plane.

Optical benches like the ones I use are capable of going far beyond any current cameras resolution capabilities, however doing this is extremely time consuming and demanding on technique, and mostly a waste of time as cameras are going to take decades to capture 1 gigapixel of resolution in a single shot. Most MTF's from manufacturers are synthetic: calculated from predicted values and hardly representative of actual performance. Only a handful of manufacturers publish actual MTF data from their lenses AFIK.

One of the biggest problems with optical testing is that you have to obtain a perfect, flawless copy of the lens in question - depending on manufacturer** this can be nearly impossible without divine intervention. I have been fortunate to be in contact with manufacturers and receive copies of lenses certified by the company itself to be free of defects - but this was an exceptional case, and is far from the norm. Sample variation can be difficult to assess with some lenses as high performance lenses can be very expensive to acquire. With some lenses I can borrow copies from friends this allows me to test multiple copies without the expense of getting copies of my own to test, this lets me obtain an overview of how great the sample variation with a particular lens can be between lens manufacturers.


* I am though. I will point out there are some user reports here that, for reasons of ignorance and ineptitude... and perhaps spite, seem to trash lenses and assign scores of 1 in every category, these reports are equally useless as those that assign perfect 10 in every category. I have tested many lenses over the decades there are only a handful I'd consider to be deserving of a 9 across the board.

** All I'm going to say is that one of the egregious offenders in the lens copy lottery over the years has a name beginning with an S.

[house]

A few people are still talking as if careful repeatable tests and charts are mutually exclusive to viewing samples and discussing difficult to articulate questions of rendering and final image.

They aren't. In fact they greatly benefit from each other. Any review should imho be based on both and many are. It sounds as if some feel that discussing the "flaws" of their favourite lens somehow takes away from the joy of using it. Or would result in having to chase perfection. I have no understanding how one could feel like that.

The very real issue as quite a few has mentioned is that tests are difficult and only applicable to the individual lens being tested. So it's a big job, useful non the less.

[BigMackCam]

A few people are still talking as if careful repeatable tests and charts are mutually exclusive to viewing samples and discussing difficult to articulate questions of rendering and final image.

They aren't. In fact they greatly benefit from each other. Any review should imho be based on both and many are.

I agree. When you can have both test-bench data and real-world opinions and examples, that's ideal. Given a choice of one or the other, though, I'd take the latter. To put it another way - multiple user opinions and examples tell me more than test data, at least given what I personally consider to be important. Others may feel differently, and that's fine.

It sounds as if some feel that discussing the "flaws" of their favourite lens somehow takes away from the joy of using it. Or would result in having to chase perfection. I have no understanding how one could feel like that.

Not me. I embrace what you call "flaws" (I might call them "characteristics"). Often, they're responsible and / or trade-offs for appealing rendering in images viewed normally, as opposed to pixel-peeping...

As I already stated, I've no objection to test-bench data. I can see some value in it, and I'd happily include it in my pre-purchase assessment process if frame-wide resolution or CA control are essential to my intended use-case(s). In its absence, though, user opinions and example photos are more than enough for me to work with. Regardless, it won't be until I buy and use a lens for a while that I'll know whether I like it or not... and, as I've found on numerous occasions, there's every possibility my own conclusions will differ from those of others (and of test data)...

---

Last edited by BigMackCam; 05-20-2022 at 05:19 AM.

[RICHARD L.]

Personally, when I acquire a new lens, I test it in real-life situations to check for the best aperture for landscape purposes. I clean it thoroughly, check everything works properly, mount the camera and lens on a heavy, solid Manfrotto 028 tripod, compose a picture and adjust my framing. I then take a picture at every aperture setting, making sure focus remains well set. For zooms, I repeat this procedure at three to five focal length settings and at every aperture opening. I then check carefully the resulting pictures on my Mac 27" monitor and determine the best aperture and focal length setting for the lens. So for my testing I always combine 1) Solid support 2) Critical focus 3) Good outside lighting 4) Middle apertures between f/5.6 and f/16 (for Medium-format lenses). I started doing this when I got my 645 FA 33-55 mm f/4.5 for my 645Z. All the reviews I had seen said it was "lowly garbage" so I had to make my own opinion about it (my copy performs extremely well BTW). I look for sharpness, contrast and resistance to flaring.

[BarryE]

... Regardless, it won't be until I buy and use the lens for a while that I'll know whether I like it or not...

... and then some more. I've often bought a lens and loved or loathed it soon after, only to change my mind later - often more than once . My needs change or my understanding develops. Or I'm just plain determined to like a lens so I persevere with it - come to the front FA 43 Limited and take a bow ...

This is what happens in reality. First impressions, I find, are sometimes/often wrong. "Exceptional" lenses tend to creep up on me, usually when they've elevated the mundane to something special. This usually gets me saying out loud "now where did that come from?". This is I suppose the same for artwork/photographs. An instant Wow! general subsides rapidly, whereas an underwhelming image has the ability to grow. Music's the same, for me, anyway.

[bdery]

I see lots of lens reviews on the site, but why aren't there any real lens measurements for any lenses. im interested in how sharp lenses are, not vague pictures of test charts.

isn't it possible to determine the lines per inch that the full size K-1 and APS-C K-3 sensors can achieve, then measure every lens of some kind of lens bench, and determine the real resolving power at all the different combinations of aperture and position across the image plane.

As an optical designer with a PhD in optics, and the person who has written most of the recent Pentaxforums reviews, I'll say:

-We have elected to provide comparative results between lenses. Knowing that a lens is 5,6% sharper than another is useless. Knowing that a lens delivers uniform results, or not, is useful. And since our tests are comparable between lenses, directly thanks to our methodology, it's possible to get a comparison between any two lenses.

-Writing reviews is extremely time-consuming. And I don't have full-time access to a calibrated lab bench, nor do I have the space to install one in my basement. It wouldn't be large enough for many tele lenses anyway.

-Reviewers test one single sample of a lens model. So we have no averages, standard deviations, sample validity test. In those conditions, I would not feel comfortable throwing around numbers which would then be dissected way too much.

That being said, what you propose is not impossible. You're welcome to undertake that task and report on your results.

[Rondec]

A few people are still talking as if careful repeatable tests and charts are mutually exclusive to viewing samples and discussing difficult to articulate questions of rendering and final image.

They aren't. In fact they greatly benefit from each other. Any review should imho be based on both and many are. It sounds as if some feel that discussing the "flaws" of their favourite lens somehow takes away from the joy of using it. Or would result in having to chase perfection. I have no understanding how one could feel like that.

The very real issue as quite a few has mentioned is that tests are difficult and only applicable to the individual lens being tested. So it's a big job, useful non the less.

I don't think they are mutually exclusive. I just think tests and charts tend to say less about a lens than real world images. When I get a new lens, I take pictures with it -- and no pictures of walls. I can detect pretty quickly if there is something going on, like decentering, but even if I had an optical bench, I wouldn't be particularly inclined to use it much.

[Digitalis]

I don't have full-time access to a calibrated lab bench, nor do I have the space to install one in my basement. It wouldn't be large enough for many tele lenses anyway.

You should have been there when I was doing my tests of the Pentax 67 800mm f/4 Takumar Vs the SMCP-M*67 800mm f/6.7 ED. To eliminate vibrations caused by domestic traffic I was working at night, but there was a freight train passing by several kilometers away that was causing enough vibrations to mess with the data. I called in a few favors, and ended up using a massive $200,000 air cushioned holography table to get clean data.

[BarryE]

You should have been there when I was doing my tests of the Pentax 67 800mm f/4 Takumar Vs the SMCP-M*67 800mm f/6.7 ED. To eliminate vibrations caused by domestic traffic I was working at night, but there was a freight train passing by several kilometers away that was causing enough vibrations to mess with the data.

Similarly, with em radiation and vibrations in the physics lab I worked in (lasers). At night. All lifts disabled in nearby buildings. Doors wedged open to stop them banging. Etc.. Worse thing was our kitchen area was out of bounds. That hurt the most. It is remarkable how far vibrations travel. Add that into the mix for the specifications for high resolution lenses testing

[Digitalis]

Worse thing was our kitchen area was out of bounds. That hurt the most.

Pressure waves travel far in water pipes. Because of this the building I ended up working in didn't have any, the air conditioning was specially filtered and piped in from another building and they had fire extinguishers of every type placed throughout the building.

[Mikesul]

Although the OPs question was a bit naïve and probably annoying to some it has prompted a very interesting discussion which I hope will continue a while longer.