[pentax_amateur]

All the lenses you tested are on a high level IQ wise. Even the 18-55 is a very good kit lens. All these lenses have there issues on the short end.

I have also tested 7 lenses at 50mm (yet to "publish" ), and 18-55 was the worst performer there.

Thus you are testing a very good lens (Sigma) on it's sweet spot against a very good lens (16-85) on its weak side. Outcome is predictable, don't you think so?

16mm is why DA16-85 sells so well! This is also why I bought it instead of 20-40 it in the first place.

[pentax_amateur]

1-How did you focus? Using focus peaking in live view, or with the viewfinder? The later will lead to errors, especially if you have not fine-tuned the AF.

I focused via the viewfinder because this is how I focus while taking shots other than "bills on a wall". I am interested to find out what lens I should use when focal distances overlap.

2-You are not always comparing identical apertures. A better way would have been to compare identical apertures, and consider anything wider to be a bonus. for instance, say the 18-55 offers only F4, and the 16-85 offers F3.5, that wider aperture is a bonus, and at F4 you can compare apples to apples.

I was debating this with myself and decided to put all widest apertures into the same pile because when I don't have enough light, I open up the aperture to the widest, whatever it happens to be on that lens. If the aperture is there, it should perform.

There are more elements than just pure sharpness to define if a lens is "well regarded". It's certainly part of it but doesn't tell the whole story.

Agree, but if a lens fails miserably at sharpness, then nothing else matters.

[RICHARD L.]

But FAJ is voted multiple times as the worst Pentax lens ever in another thread

The FA J 75-300 mm is terrible above 135 mm but has excellent contrast. You can make decent images with it from 75 to 135 mm @ f/11 and f/16. Apparently the FA J 28-80 is horrible but I never had one. OTOH, my FA J 18-35 mm f/4-5.6 has given me really good pictures on a low resolution sensor, such as a K10 or a K20, when stopped down to f/11 and f/16. Anyway their "all-plastic" construction is not conducive to building a great trust in them or a large users' following. They are totally unusable on a K3 or a K1.

Regards

(Left picture) FA J 18-35 mm f/4-5.6 AL on a K20 @ 18 mm FL and f/11 .................. (Right picture) FA J 75-300 mm f/4.5-5.8 AL on a K10 @ 75 mm FL and f/11

[Kwong]

DA16-85 is still struggling a lot. FAJ18-35 is the best at f/8! This is not normal. I must have messed up my test.

FUnny you should mention this i bought the FAJ 18-35 pretty cheap not expecting much and it really surpassed my expectations. Great colours and very acceptable performance stopped down

[UncleVanya]

The da 16-85 at 16 was reviewed as being better than the da* 16-50 and other lenses. It is hard to imagine how the kit lens bested it so completely at the wide end.

[Wingincamera]

You never said if you had checked & adjusted the calibration of the lenses, which will make a difference when focusing through the viewfinder. I have two KP's and each is slightly different with the same lens. I checked & rechecked and it is consistent with each camera. An example is the 50-135 lens, on the older KP it takes a -1 and with the newer KP, it is +1.

[Lowell Goudge]

If you photographed at the same focal length and had to move the camera then there is breathing, meaning they are not really the same focal length at short distance.
How much did you have to move them and it would be intetesting to know which is widest at each length.

This is not necessarily true, it depends upon where the nodal point of each lens is, not where the camera is.

---------- Post added 04-27-21 at 05:52 PM ----------

I really should try something like this as I have all 4 of the lenses mentioned by the OP

[swanlefitte]

This is not necessarily true, it depends upon where the nodal point of each lens is

Shouldn't the nodal point distance ot sensor be the same for any lens with a given focal length? How could this be otherwise? Panoramas use nodal point but mean front entrance pupil which can or cannot shift but the lens center can't shift and f to stay the same.
"The lens formula says that the inverse of the distance from the object plus the distance to the image equals the inverse of the focal distance ​f​."

[Lowell Goudge]

Shouldn't the nodal point distance ot sensor be the same for any lens with a given focal length? How could this be otherwise? Panoramas use nodal point but mean front entrance pupil which can or cannot shift but the lens center can't shift and f to stay the same.
"The lens formula says that the inverse of the distance from the object plus the distance to the image equals the inverse of the focal distance ​f​."

That’s only for the example of a single lens.

Perhaps the nodal point is not the correct reference, but given that all these lenses have retrofocus elements, individual lens construction will dictate where the theoretical plane of the lens is, with respect to the subject.

Usually I consider this to be the front element, as a working distance

[bdery]

I focused via the viewfinder because this is how I focus while taking shots other than "bills on a wall". I am interested to find out what lens I should use when focal distances overlap.

I was debating this with myself and decided to put all widest apertures into the same pile because when I don't have enough light, I open up the aperture to the widest, whatever it happens to be on that lens. If the aperture is there, it should perform.

Respectfully, while that might represent how you use lenses, it also ensures that the comparison, in general terms, is unusable. Especially if you didn't Fine Adjust the AF.

Using the viewfinder, the only way to measure sharpness the way you did it would have been to adjust AF, then take a series of 5 or 10 pictures for each lens.

Also, as I said before, if you didn't use the same apertures, you're not comparing apples to apples.

Agree, but if a lens fails miserably at sharpness, then nothing else matters.

"Fails miserably" is an exaggeration.

[RobSpain]

I have the constant f/3.5 version of the Sigma 10-20mm (on a K-70) and get very good results in real estate photography especially on interiors, using bracketed HDR. The focusing is super sharp with little or no noise, and if there is, I've never had a problem removing in PP in LR.