[jeanpierre]

I wanted to test my "old" SMC F 28-80/f3.5-4.5 macro zoom lens on the K-3 for DOF.

Well, I'm surprised how good DOF is. Almost as the good old photography time.
However, there is an important factor, that we not must forget. Diffraction!!!

https://www.dropbox.com/s/xw39qba66iu2iko/DOF_SMC_F_28-80_f_5_6_to_32.png

Please look at the picture. From left to right you see the shots from f/5.6 to f/32.
From f/14 the branch is no longer sharp. The strongest focus is at f/11!

Bit of a shame. But, if I am not dependent on focus, then I can stop down to f32! Yes, the good old lenses are still useful! I do not know, if the new digital lenses works equal. Maybe, someone can tell me.

Good to know. I hope to help with it.

[normhead]

I find down to F22 to be useful on most lenses…. but at 28, f11 is probably all you need. Check your DoF guide. That diffraction is there on film as well, you possibly didn't notice because film is a naturally fuzzy process, but it was always there. With film, if you used large grained high speed film, one corner of a grain being hit by a photon would cause un-exposed areas of the grain to be developed. The only way to defeat that was to use a fine grained film, and then you had low ISO. lets not get nostalgic here. There were always trade offs. But then people were a lot less critical, images were often not viewed beyond 4x6 so f32 was acceptable. Reduce your image size to 4x6 on your screen, and it will probably still be acceptable.

The digital lenses should give you less Chromatic Aberrations and purple fringing. But older lenses are still quite useful. Use the forum lens guide. A bad lens from the film days is still bad in digital format. And an FA 31 ltd is still a great lens… you can't generalize…

Here's your lens… SMC Pentax-F 28-80mm F3.5-4.5 Reviews - F Zoom Lenses - Pentax Lens Reviews & Lens Database
As you can see, there are quite a few lenses rated higher…. but people who own it like it.

[lister6520]

There is a bit of misconception about diffraction and sensor resolution. The blurring caused by diffraction is always the same and dependent only on the aperture.

The reason we say diffraction 'sets in' earlier on on a higher resolution sensor is simply because its pixels are smaller than the blurring. With lower resolution the blurring is still there but you cannot see it because there is a worse form of 'blurring' that is pixellisation.

Viewed at the same magnification an F22 picture shot with a low resolution sensor, a film or a high resolution sensor will appear the same. If you magnify more the low resolution picture will break up into individual pixels before you start to notice the softness in the high resolution one. In other words at no point will the low resolution sensor give a sharper image than the high resolution one regardless of the aperture.

[normhead]

There is a bit of misconception about diffraction and sensor resolution. The blurring caused by diffraction is always the same and dependent only on the aperture.

The reason we say diffraction 'sets in' earlier on on a higher resolution sensor is simply because its pixels are smaller than the blurring. With lower resolution the blurring is still there but you cannot see it because there is a worse form of 'blurring' that is pixellisation.

Viewed at the same magnification an F22 picture shot with a low resolution sensor, a film or a high resolution sensor will appear the same. If you magnify more the low resolution picture will break up into individual pixels before you start to notice the softness in the high resolution one. In other words at no point will the low resolution sensor give a sharper image than the high resolution one regardless of the aperture.

Yes, but at many apertures you get just as sharp an image….especially around ƒ5.6. ANd actually, if you're referring to Aperture as test chart resolution, you will get a sharper image within the area of the crop using an APS-c camera, so even that needs to be qualified.

A K-3 image is just as sharp as a 6D image… but that sharpness is concentrated into a much smaller area.

[jeanpierre]

I find down to F22 to be useful on most lenses…. but at 28, f11 is probably all you need. Check your DoF guide. That diffraction is there on film as well, you possibly didn't notice because film is a naturally fuzzy process, but it was always there. With film, if you used large grained high speed film, one corner of a grain being hit by a photon would cause un-exposed areas of the grain to be developed. The only way to defeat that was to use a fine grained film, and then you had low ISO. lets not get nostalgic here. There were always trade offs. But then people were a lot less critical, images were often not viewed beyond 4x6 so f32 was acceptable. Reduce your image size to 4x6 on your screen, and it will probably still be acceptable.

The digital lenses should give you less Chromatic Aberrations and purple fringing. But older lenses are still quite useful. Use the forum lens guide. A bad lens from the film days is still bad in digital format. And an FA 31 ltd is still a great lens… you can't generalize…

Here's your lens… SMC Pentax-F 28-80mm F3.5-4.5 Reviews - F Zoom Lenses - Pentax Lens Reviews & Lens Database
As you can see, there are quite a few lenses rated higher…. but people who own it like it.

Thanks for your pointed out. With my old Pentax I usally shot with f22 and I cannot say that the sharpness was less then using f8 or f4! The film was physically build and not digital as the sensor nowaday! All CMOS-sensor have a diffraction limit!
This limit cannot be exceed with lenses! So, the manufacturer of lenses tries to build lenses for "diffraction limit of a type of sensor"! But the Sensor manufacturer will push the maximum of the sensor out! So, we will have must lenses that no go with some sensor sizes!!!
I owned a Nikon D800 and no Nikkor digital lenses works very good or excellent with it! Only the very new Sigma 35/1.4 mm lens works great. I do think that Nikon will build new lenses for CMOS sensor with more then 24MP!

The K-3 works quite good with the "old" lenses. Very nice bokeh and sharpen enough! I do think that the M lenses from Pentax will works extremly good!

Your right the 31/f1.8 is a great lens indeed. All ltd lenses of pentax are awesome!!!

I do not give much of rating see in the lens database. Because the lenses were tested with different digicam with different sensor-type!!!! So, I do have to test myself with my K-3 to see what I have.
I read the review of the F 28-80. After that I would never use this lens an my K-3! And, I did it. The result is better then the review! So, I wish all to test yourself! And don't worry. Sometime you will have big surprise!

Last, almost rawconverter software can remouve chromatich aberration and purple fringing! That is not a reason to not use a lens!! And, every year all software are updated and works better!!!!

[jeanpierre]

Yes, but at many apertures you get just as sharp an image….especially around ƒ5.6. ANd actually, if you're referring to Aperture as test chart resolution, you will get a sharper image within the area of the crop using an APS-c camera, so even that needs to be qualified.

A K-3 image is just as sharp as a 6D image… but that sharpness is concentrated into a much smaller area.

My test with the K-3 and the 35/2.8 ltd macro lens give my very good sharpness:
Only CENTER: from f2.8 to f11
CENTER and CORNER: from f4.0 to f11
So, I can use 2.8 to f11 for DOF and to have best sharpness too!
After f11 the image seems to blur more! The diffration limit of the 24MP CMOS-Sensor!

[normhead]

I have a couple threads on this issue, 'one here <click here>

[jeanpierre]

I have a couple threads on this issue, 'one here <click here>

Thanks for the link. But, I cannot see which digicam was used! That is important, as I wrote. It depends which sensor is use with the lens.
Those compares are nonsense, if you not can read the EXIF data!

[normhead]

The images were taken with a K-5. I think, you're getting confused on how critical a factor this is. Focus on the issues, not the exact numbers. The bottom images posted by Nestor were taken on a D600. If memory serves me well, the difference in diffraction limit comparing a 24 Mp APS-c sensor and a 16 Mp APS-c sensor is less than a stop. Some point between ƒ5.6 and f8 becomes the point at which diffraction starts to become apparent on many lenses on both systems. Again , working from memory, the diffraction limit for APS-c doesn't go below ƒ5.6 until about 40 Mp. And and ƒ8 is still an acceptably sharp image on both systems. Most charts measuring MTF, see photo zone.de show sharpness increasing until ƒ5.6 although some peak at ƒ4 and some at ƒ8.

And usually, neither ƒ2.8 or ƒ4 are as sharp as ƒ5.6 although it is possible that with a really expensive lens that could be true. Usually lenses get sharper as they are stopped down until they reach the diffraction limit.

[Class A]

A K-3 image is just as sharp as a 6D image… but that sharpness is concentrated into a much smaller area.

Because the same sharpness has to be concentrated in a much smaller area (as you've put it), it is actually quite a bit harder for the K-3 to produce images with the same sharpness of a 6D.

The lenses on the K-3 have to be 1.5x as sharp. The AF on the K-3 has to be 1.5x more precise. If you use a lens on the 6D at f/2.8 -- an aperture were good lenses produce very good sharpness already -- you'll have to use f/1.8 on the K-3 to get the same shallow DOF, a much harder ask if you want the same sharpness.

Full-frame is not just about obtaining shallower DOF, it is also about getting better IQ for less money.

Regarding different diffraction limits for different sensors: Most of the time this is discussed in very misleading ways. User lister6520 hit the nail on the head; when looking at the full image, pixel count does not matter with respect to the degree of blur contributed by diffraction.

[jeanpierre]

Because the same sharpness has to be concentrated in a much smaller area (as you've put it), it is actually quite a bit harder for the K-3 to produce images with the same sharpness of a 6D.

The lenses on the K-3 have to be 1.5x as sharp. The AF on the K-3 has to be 1.5x more precise. If you use a lens on the 6D at f/2.8 -- an aperture were good lenses produce very good sharpness already -- you'll have to use f/1.8 on the K-3 to get the same shallow DOF, a much harder ask if you want the same sharpness.

Full-frame is not just about obtaining shallower DOF, it is also about getting better IQ for less money.

Regarding different diffraction limits for different sensors: Most of the time this is discussed in very misleading ways. User lister6520 hit the nail on the head; when looking at the full image, pixel count does not matter with respect to the degree of blur contributed by diffraction.

I do not understand. Why the MFT Olympus OM-D EM-1 has a very good image quality ..... when you say FF is getting better IQ!!????
I do not think, that FF or APS-C or MFT is important for IQ.
One example put the DA* 300mm on the Pentax Q7 (1 1/7" sensor) and you will have a very good image quality!!!! Or an other example, put a Pentax 645A or D lens on the Pentax K-3 and you will have an outstanding IQ!!!!
Oh, I tested my Pentax 645A lense with the Nikon D800 and the IQ is awesome. I never had so good IQ will all Nikkor or Sigma lenses!!!

I am a practitioner and beleive only things that I tested my self!

And I repeat: For good image quality, you have to choise the lens, which match for the digicam not vice versa. Why do you think, that Sigma has built new lenses (35, 85 and 18-35) last year? Because they know the problem with the high resolution sensor for 24 Mp and up.

[mecrox]

I do not understand. Why the MFT Olympus OM-D EM-1 has a very good image quality ..... when you say FF is getting better IQ!!????
I do not think, that FF or APS-C or MFT is important for IQ.
One example put the DA* 300mm on the Pentax Q7 (1 1/7" sensor) and you will have a very good image quality!!!! Or an other example, put a Pentax 645A or D lens on the Pentax K-3 and you will have an outstanding IQ!!!!
Oh, I tested my Pentax 645A lense with the Nikon D800 and the IQ is awesome. I never had so good IQ will all Nikkor or Sigma lenses!!!

I am a practitioner and beleive only things that I tested my self!

And I repeat: For good image quality, you have to choise the lens, which match for the digicam not vice versa. Why do you think, that Sigma has built new lenses (35, 85 and 18-35) last year? Because they know the problem with the high resolution sensor for 24 Mp and up.

Canon 6D + Canon 24-70mm f/4 + Canon 70-200mm f/4 + Speedlight 430EX = 3498 pounds sterling; weight ex flash = 2115 grams
Pentax K3 + Pentax 16-50mm f/2.8 + Pentax 50-135mm f/2.8 + Pentax AF360 FGZ II AW = 3336 pounds sterling; weight ex flash = 2050 grams

Two roughly comparable systems for about the same cost and weighing about the same. Chuck in a fast 35/50mm and either may be all anyone needs. Depending on special offers, reductions and the like, one or the other may be more financially attractive at any one time.

So, in either system, what are you really paying for?

Another example: the new Pentax DA 20-40mm f/2.8-4 for 849 pounds sterling. It weighs just 283 grams. On FF, that becomes a 32-64mm f/4-5.6 lens and suddenly not so sexy anymore, at least at that price. So what are you paying for?

Everyone has to make a choice somewhere along the cost/performance/weight curve. However, folks like Falk Lumo have shown pretty well that the squeeze on the smaller mainstream formats is real and that there are solid reasons for it - provided one is aiming for the same high standard of quality on each format. It's really just "no free lunch" all over again.

[normhead]

Because the same sharpness has to be concentrated in a much smaller area (as you've put it), it is actually quite a bit harder for the K-3 to produce images with the same sharpness of a 6D.

The lenses on the K-3 have to be 1.5x as sharp. The AF on the K-3 has to be 1.5x more precise. If you use a lens on the 6D at f/2.8 -- an aperture were good lenses produce very good sharpness already -- you'll have to use f/1.8 on the K-3 to get the same shallow DOF, a much harder ask if you want the same sharpness.

Full-frame is not just about obtaining shallower DOF, it is also about getting better IQ for less money.

Regarding different diffraction limits for different sensors: Most of the time this is discussed in very misleading ways. User lister6520 hit the nail on the head; when looking at the full image, pixel count does not matter with respect to the degree of blur contributed by diffraction.

Because the same sharpness has to be concentrated in a much smaller area (as you've put it), it is actually quite a bit harder for the K-3 to produce images with the same sharpness of a 6D.

You point the camera, you press the shutter… what could be easier?

For both APS-c and FF ƒ 5.6 tends to be the sharpest. IN only the most expensive glass, if you're considering sharpness, are you going to be shooting anything but ƒ5.6 or ƒ8. With 99.9% of the lenses ever made, you are going to sacrifice sharpness, for narrow DoF.

The lenses on the K-3 have to be 1.5x as sharp. The AF on the K-3 has to be 1.5x more precise. If you use a lens on the 6D at f/2.8 -- an aperture were good lenses produce very good sharpness already -- you'll have to use f/1.8 on the K-3 to get the same shallow DOF, a much harder ask if you want the same sharpness.

If you follow MTF scores, you know every lens made to date produces sharper images for the most part based on pixel density. There are no lenses that I know of that are topped out on FF and don't have 1.5 more sharpness than FF requires. IN other words the lens will produce similar results on APS-c and FF. IN fact, just a general observation would be that based on the 3000 lw/ph, the FF advantage is less than 120 lw/ph. That's less than 3% and probably not discernable by the human eye. The notion that you have to have a better lens on APS-c conveniently ignores the fact that 95% of lenses made in the last 50 years are more than adequate on both systems. And that if you want to max out either system, you are going to have to buy expensive glass. They both benefit from better glass, it's not an attribute confined to one or the other. And they both function adequately with old in-expensive glass.

I personally find it hard to comprehend the effort that's been expended trying to create confusion on this issue.

The thing I suspect is happening, is that we are discussing imperceptible differences that couldn't be seen in comparing images at anything other than pixel peeping type resolutions. I've seen many articles by those who work with prints etc. that have suggested as much. I'm still waiting for that definitive set of images, that show beyond a question of doubt, there is an FF advantage. The theory is dramatic. The real world examples, most of the time you can't tell the difference. Especially if pixel peeking isn't part of your real world.

[Rondec]

Because the same sharpness has to be concentrated in a much smaller area (as you've put it), it is actually quite a bit harder for the K-3 to produce images with the same sharpness of a 6D.

The lenses on the K-3 have to be 1.5x as sharp. The AF on the K-3 has to be 1.5x more precise. If you use a lens on the 6D at f/2.8 -- an aperture were good lenses produce very good sharpness already -- you'll have to use f/1.8 on the K-3 to get the same shallow DOF, a much harder ask if you want the same sharpness.

Full-frame is not just about obtaining shallower DOF, it is also about getting better IQ for less money.

Regarding different diffraction limits for different sensors: Most of the time this is discussed in very misleading ways. User lister6520 hit the nail on the head; when looking at the full image, pixel count does not matter with respect to the degree of blur contributed by diffraction.

You are correct, but for many applications, it is actually not that hard to get pixel level sharpness, at least from a lens standpoint. When I shoot landscape photos, I mostly shoot at f8 and photos are plenty sharp. I sort of hate these discussions, because they magnify the small number of applications where full frame is "better" and the reality is different.

If Ricoh releases a full frame camera, I will likely get it. But, I don't expect it to save me money or, that my photos would suddenly take a huge bump in quality.

[normhead]

You are correct, but for many applications, it is actually not that hard to get pixel level sharpness, at least from a lens standpoint. When I shoot landscape photos, I mostly shoot at f8 and photos are plenty sharp. I sort of hate these discussions, because they magnify the small number of applications where full frame is "better" and the reality is different.

If Ricoh releases a full frame camera, I will likely get it. But, I don't expect it to save me money or, that my photos would suddenly take a huge bump in quality.

With the release of the K-3, I'm completely un-impressed with what the extra-resolution is giving me. More room for cropping is nice, better AF, very nice… there are a lot of things I like about the K-3… but I'm seriously questioning needing more resolution than 24 Mp gives me. I was already at that point with my k-5. At 24 Mp, I'm not feeling the urge to go to 36. I seem to have hit my magic number. Of course, other people may have different magic numbers.