[whojammyflip]

Hi, I have got up and running with MTF Mapper, printed a grid target and am performing some MTF testing. It generates MTF curves, ie contrast at chosen resolutions 10, 20 and 40lpmm. I am running through testing of 31 lenses at the moment, at each marked f stop. My most interesting results have been:

- Pentax M 85mm, sharp wide open, and hits peak average sharpness at f5.6, with contrast across the frame of over 63% at 40lpmm
- Vivitar Series 1 v1 70-210mm, sharpest lens at 210mm at any aperture with average contrast of 50% at 40lpm at f5.6
- Tamron 278D 80-210mm, sharpest lens at 135mm at f11 with average contrast of 64% at 40lpmm

I particularly like the slanted edge MTF testing on a grid, as its a quick way to profile a lens across the image plane.

Here is the result for the M 85mm at f5.6:


Here is a list of lenses I have for testing (cannot mount the Vivitar S1 35-85 at the moment as its rear flange hits the K7 body):
Pentax DA 18-55 AL
Pentax M 75-150
Pentax M 80-200 v1
Vivitar S1 V1 70-210
Vivitar S1 V2 70-210
Vivitar S1 V3 70-210
Tamron SP 70-210 19ah
Tamron SP 70-210 52a
Tamron SP 60-300 23a
Kiron 80-200 f4
Kiron 75-150
Vivitar S1 35-85
Vivitar S1 28-90
Tamron 28-105 179D
Tamron 80-210 278D
Pentax F 80-200
Tamron SP 28-135
Tamron SP 28-80
Tamron SP 24-48
Pentax M 24-50
Pentax FA 28-70
Tamron 28-80 77D
Sigma 21-35
Pentax M 135
Pentax M 85
Pentax M 50 1.4
Pentax M 50 1.7
Pentax M 28 3.5
Pentax M 28 2.8
Tamron SP 90 2.5
Pentax M 200

[Medex]

Usually MTF is tested for 10 lines / mm and 30 lines / mm for distances 5, 10, 15, 20 mm from the sensor center (depends on sensor type - APS-C or FF). Zeiss is testing its lenses at 10, 20 and 40 lines/mm.
You can't compare directly your lenses with other, tested on another equipment. But I think, it is better to test 10 and 30 lines/mm, because there are more MTF data charts of other lenses on the internet for some kind of analysis of data.
Actually f5,6 will be "impressive" for many lenses, even old. As a portrait lens, M85 is more interesting at f2-2,8, IMHO.

[whojammyflip]

Here is a comparison of the Pentax M 28 2.8 at f4 vs the Pentax M 28 3.5 *wide open*

Pentax 28 2.8 at f4:


Pentax 28 3.5 *wide open*


I've literally just shot around 200 shots, so I am not going to post the lot. I also have the Pentax M 50 f4 and the Pentax A 50 f2 which I am going to compare with the M 50 1.4 and M 50 1.7.

This train is leaving the station! I have been meaning to get around to quantifying the performance of the Pentax lenses for a while, using MTF testing. So here it comes, watch this space.

[StiffLegged]

Have fun, but that's too ubergeek for my liking. Images are more interesting – and enjoyable.

[whojammyflip]

Some MTF curves for primes at f4, interesting to see little or no difference between the M 50 1.7 and 1.4. Using MTF Mapper on an A1 grid, as PEF files.

Pentax M 50 f4 at f4


Pentax M 50 f1.7 at f4


Pentax M 50 f1.4 at f4


Pentax A 50 f2 at f4

[rawr]

Nice work!

Must be exhaustingly tedious to perform these tests, but it's worth doing if you have the technology and time to do so. Most of the lenses on your list have probably never been MTF'd before outside of the factory.

[stevebrot]

Some MTF curves for primes at f4, interesting to see little or no difference between the M 50 1.7 and 1.4.

That finding is in agreement with previous testing where the f/1.4 only attains parity at about f/4.


Steve

[stevebrot]

I particularly like the slanted edge MTF testing on a grid, as its a quick way to profile a lens across the image plane.

Have you tried AF fine-tune using MTF Mapper?


Steve

[whojammyflip]

Have you tried AF fine-tune using MTF Mapper?


Steve

Nope, I have just beenevaluating lenses against each other. The next possible test is to mount some enlarger lenses and profile them...could give Ctein a run for his money. One thing is clear, there is a good reason for the "f8 and be there" phrase. Except for the Pentax M 85mm....that lens peaks in certain respects at f4.

The most controversial result I have had is for the Tamron 278D at 135mm f8 to f16. Wide open its a bit of a stinker though and the extremes are not good.

I have just posted this in the lens review section:

Tamron 278D at 135mm f8, MTF25 resolution


Tamron 278D at 135mm f8, contrast at 20, 40, 80lpmm


---------- Post added 24th Feb 2019 at 12:49 ----------

Nice work!

Must be exhaustingly tedious to perform these tests, but it's worth doing if you have the technology and time to do so. Most of the lenses on your list have probably never been MTF'd before outside of the factory.

This is my kind of expertise. I work in analytics and this is fun for me.

The next step would be to just do the MTF calculation on the MTF Mapper targets, and get an aggregate reading for contrast at say 40lpmm, across the frame, in order to be able to rank lenses on a single metric. I think 100lpmm might make more sense as this is the around the limit for the sensor/film resolution. The trouble is that most lenses score zero at this spatial frequency. That Tamron is exceptional.

[stevebrot]

The most controversial result I have had is for the Tamron 278D at 135mm f8 to f16. Wide open its a bit of a stinker though and the extremes are not good.

I suspect that the previous reviewers might be shooting at focal lengths and distances other than those used for your testing. (Average reviewer rating = 7.5 including your rating of 10). For the convenience of inquiring minds...

Tamron 278D AF 80-210mm 1:4.5-5.6 Lens Reviews - Tamron Lenses - Pentax Lens Review Database

I know almost nothing about this lens other than it and its label variants (sold most famously as Promaster) reputation as "plastic fantastic" by some reviewers and "don't bother" by others. With any luck, perhaps @Jonathan_Mac will weigh in. He rated the lens "6" in his review.


Steve

[zapp]

Nice software that visualizes variations. It would be nice to test multiple samples of each lens. Have fun.
With optics calculation software you can model best/worst performance of the lenses you produce.