[GUB]

The top profile is exagerated by not scaling to zero. If they start it at 2/3 of full resolution the top profile will show differences that are three times as large as the ones that start at zero. And if 1/2 of full resolution they will be exagerated twofold.

Yes but even if I resize the vertical 50% the difference is still striking.
And note the 50 1.4 8 element above at #33. Wide open corner is down to about 600 lpwh. probably "poor" but not below. I still think of those corners as being useful in the context of using that lens wide open.
I know the graduations don't fit but I think it is reasonable to think that 500 is on the K-3 axis and not cropped off.

Attached Images

 

[dms]

Maybe, although the 50 mm f/1.4 is very low in the corners--probably in part the age of the design, and that it is f/1.4. Aspherical lens design is now common, and I would think the poor rating now would be about 3 times higher resolution.

[clackers]

I know it is pushing it to compare these two ephotozine tests - one with a K-3 and one with a K-1

Those are *not* MTF curves, GUB.

They are just a magazine's presentation of IMATEST data, because they did not have proper test equipment.

This is the MTF curve, from Ricoh, for the FA35. Note that the horizontal axis is the distance in millimetres from the centre across the frame, and it'll be different for APS-C and FF. We know this is full frame, because it's reaching 22mm into the corner.

[GUB]

They are just a magazine's presentation of IMATEST data, because they did not have proper test equipment.

So you feel most of the difference in the tests is inaccuracy?

In the Ricoh test then the apsc would be identical but stop at 14?
(And of course this test is saying nothing about pixel pitch.)
I have included below the Ricoh explanation of the test for those like myself that are having trouble understanding the graphic .

Attached Images

 

[clackers]

I know the graduations don't fit but I think it is reasonable to think that 500 is on the K-3 axis and not cropped off.

Below are the MTF curves for that 70-200.

If you want to see what it's like on the K-3, just follow it to the 15mm mark and ignore everything after that.

[clackers]

So you feel most of the difference in the tests is inaccuracy?

Sure. Mediocre testing is all over the internet.

Why do we think Tony Northrup is not a lens designer at Canon?

You can actually buy that IMATEST software ePhotozine uses, it might even have a trial edition. Give it a go and I'm sure you'll see multiple points in the process where it can be stuffed up.

Superficial online reviewers pontificating to us need to remember that two things are being tested - the product in front of them, and their own skills.

Unless they're someone like our great Bernard Dery ( @bdery ), I think almost every social media influencer needs to stick to doing demos of the product they're shilling, not attempt 'reviews'. As a public, I think we deserve better.

[GUB]

Superficial online reviewers pontificating to us need to remember that two things are being tested - the product in front of them, and their own skills.

Unless they're someone like our great Bernard Dery ( @bdery ), I think almost every social media influencer needs to stick to doing demos of the product they're shilling, not attempt 'reviews'. As a public, I think we deserve better.

MMMM I think you are being a little harsh on people's real world efforts here.
The K-1 test is by John Riley. (And I should have credited him!). I follow his tests a little because he has the passion to expend time on testing vintage lenses. Financial gain is surely not a motive here.
What I find interesting about his vintage tests is (in the ones I have examples of ) the test strongly reflects my own practical experiences.
And maybe the range of tests out there have shortcomings in accuracy but but what other tools do we have to, for instance, (to keep on subject) demonstrate real world diffraction limitation? The Ricoh test doesn't go there.

[clackers]

And maybe the range of tests out there have shortcomings in accuracy

The problem is people don't realise that, but make decisions based on them. Look at the fetishisizing of DxOMark and their 'algorithms'.

There are plenty of people badmouthing particular types of lenses - vintage or new - based on their own copy, not anyone else's. But they have a blog, or Youtube channel, or whatever.

but but what other tools do we have to, for instance, (to keep on subject) demonstrate real world diffraction limitation? The Ricoh test doesn't go there.

Of course the effect can be witnessed in MTF curves.

You can read all about it here. Limiting resolution and MTF

There's no point defining 'real world' as 'what I can do myself with what I have in my house'.

They're not the same thing.

It's like home surgery. We don't have the right gear and most of us the aptitude to not screw it up.

Look at The Cameraville guy's methodologies ... and now he's down to 500 views from 40,000 once upon a time when he covered Pentax.

Camera owners deserve better!

[zkarj]

In terms of post-processing — use DxO PhotoLab or DxO PureRAW to do at least the first step of demosaicking. It certainly beats Lightroom, ON1, Luminar, Aperture, DarkTable for absolute sharpness. Probably others, as well.

[GUB]

There's no point defining 'real world' as 'what I can do myself with what I have in my house'.

They're not the same thing.

Na na I was meaning testing with practical photography relevance.

[BigMackCam]

In terms of post-processing — use DxO PhotoLab or DxO PureRAW to do at least the first step of demosaicking. It certainly beats Lightroom, ON1, Luminar, Aperture, DarkTable for absolute sharpness. Probably others, as well.

RawTherapee's demosaicing is decent (with a lot of algorithm models to choose from, depending on image properties), and its Capture Sharpening tool is excellent. I'm sure DxO is better, but if the OP doesn't own it, RawTherapee is a good and free alternative...

[Carlos33]

You'd not get a significant bump in detail using more recent lens models.

What you could do with what you already have is:
- use pixel shift
- use a wider lens aperture and focus stack
- use pixel shift, with wider aperture and focus stack
- get closer and stitch a few frames (e.g using the 100 macro instead of the 50 macro)
- fine grained sharpening (capture sharpening / de-convolution type sharpening)

Combining pixel-shift, focus stacking and stitching should get you a massive increase in image detail, downside being more processing time.

Hello

Does somebody know, is there is difference in IQ between the DFA 100mm Macro f2.8 (older, 2009) and the newer WR version?
It would be good a comparision or all the info.


Thanks.
Carlos