[BirdDude007]

I bought the K-5 back in April, fully knowing (from what I was reading in here) that they are coming out with a new camera to replace the K-5. I however am the type of person who likes to buy one of the last models of the last run. This camera is an awesome camera, and it will be for a while to come.
It is not the improvements that create better images, it is the photographer. Just remember that and you will do fine.
Honestly I feel the extra cost that they are charging for the K=5IIs and the K-5II is not really worth it, my K-5 (classic) is way more camera that I will need for a long time to come. I am actually starting to get nice sharp images out of it, especially since I am starting to come to grips with my style and also how to work with Light Room and PS for all the PP.
A camera is a tool, keep that in mind. Some tools are easier than other tools, but in our case, all of our tools will produce the same result, although some tools may require more work than others to achieve the same result. Take for example my K-x, ya I started out in the DSLR world with a K-x and before long I was out growing that little (awesome) camera, I needed more options to make my workflow easier so I got the K-5(mostly out of necessity) I utilize the upper LCD screen, and the front and rear edials are a God send. I needed the bells and whistles in this camera so that I can concentrate on my shooting/composition more instead of the settings in my camera. With the K-x my hands were tied because everytime I wanted to change a setting on the fly, I had to stop and go into some menu and reset the camera to where I wanted it, then recompose the shot, this all took time, but with the K-5 I don't have to worry about my settings and any changes I need to make are right there at my fingertips.
I take just as good of pictures with my K-x as I do with my K-5, but it takes a bit more work with the K-x. I could care less about megapixels, antialising, prettier LCD screens, etc...
A good photographer takes awesome pictures with whatever he has at his disposal. If you can afford the K-5IIs then get it, however in my humble opinion I feel that you could be wasting your money, but I am not one to tell you how to spend your money.
When my K-5 dies, then I will replace it, ........thats it!.....I am not going to chase the pipe dream of having the "biggest and baddest" camera available! To me cameras are tools and they are treated as such by me. If I get a replacement camera then I will just get whatever works best for me, and when that time comes it will probably be when that "unicorn" FF camera comes out, and all these folks in here that just traded up their K-5 classic in for their K-5IIs are dumping them off in the market place on this forum, I will be there to snatch one up at a bargain price. Until then I will be out there shooting my K-5 instead of chasing down the next best thing so I can be "outdoing" the "Jonses"
Just get what works best for you and stop worrying about what others think, ya it is good to do you research and get a good deal, but once you get it then I suggest you stop reading the threads in here and just get out there and shoot your dream, THEN come in here and post your shots.

[Class A]

I probably need to go back and find the thread that touched on this, however in terms of resolution for the APS-c sized sensor, the 16MP is essentially optimal. Any more MPs and you are going past diffraction limit and essentially doing nothing more than oversampling (brings nothing more to the table other than larger files). :

Please find that thread and let the person who made that statement know that 16MP is nowhere near enough. When we have 100MP we can start talking.

[normhead]

Everything I've read about the camera's new AF system seems to indicate that it's only better in low-light. Everything else makes it sound like it's slightly better (but barely even noticeable) in normal shooting conditions.

IN good light the Pentax was less than 2 tenths of a second slower than the fastest Canon. For most of us, it was the low light focussing that was the problem, that's where Pentax really fell behind the others and earned it's "slow focussing" reputation. And really, if someone wants to a really fast focussing camera for birds in flight, the Canon Ring motor is patented every way it can be patented. Pentax isn't going to make a ring motor, just go for it, although I'm guessing when you need that, renting the equipment you need for the day is probably a better option. The focussing test everyone refers to, because it was 3rd party and relatively unbiased, showed Pentax had the highest keeper rate, for normal shooting. No one talks about the Nikon,Canon crap shoot approach where the cameras don't achieve acceptable focus 20% of the time when they are telling you they have achieved focus lock, as compared to about 5% for Pentax. It's a trade off. Pentax obviously went for a slower focus lock with more accuracy than the other brands. As a landscape photographer, I'm cool with that. The only way Pentax's AF is inferior to Canons or Nikons is if your subject is so fast the .15 second the Pentax is slower means you don't actually ever achieve focus lock. There have been guys on tis forum who switched to Canon or Nikon who have reported that. So much of their work was fast moving they got way more keepers with Canon or Nikon equipment. With BiFs and other fast moving objects, that can happen. I'd suggest that photographers who actually need that kind of performance on a daily basis are actually quite few and far between. For most of us, the Pentax system is a better system.

[interested_observer]

Please find that thread and let the person who made that statement know that 16MP is nowhere near enough. When we have 100MP we can start talking.

Morning Class,

I have not tried to go back to find the post, as that would serve no useful function. The desire "that 16MP is nowhere near enough" and that "when we have 100MP ....", need to be based on what is obtainable, practical and useful. Unobtainium is not useful to anyone.

Where I was going was along the lines of diffraction. There are limits (laws of optics) that come in to play as various parameters get jockeyed around, in terms of designing camera systems. I do landscapes, cityscapes and the like. I also dislike hauling around large cameras. The APS-c sensor size as its packaged is pretty optimal for me. I really do not want to go much larger, nor spend a lot more. Nor do I want to dump my glass and start over again.

So back to diffraction. The K5 at 23.7 x 15.7mm yields about 370mm^2 providing a pixel pitch (diameter) of 4.81um. Just for comparison (FF) the D800's pixel pitch is 4.9um. So you start to deal with the Circle of Confusion or Circle of Diffraction.

Using the Cambridge Color site...

• Digital Camera Diffraction – Resolution, Color & Micro-Contrast

Lets look at the Diffraction Limited Aperture
..........................................................................APS-c (16.1MP).......FF (36MP)
Diffraction May Become Visible ..................................f7.2 ........................f7.4
Diffraction Limits Extinction Resolution ........................f9.1 ........................f9.2
Diffraction Limits Standard Grayscale Resolution ........f10.9 ......................f11
OVERALL RANGE OF ONSET .................................f7.2 to f10.9 ...........f7.4 to f11

What this tells me is that at f8 - good for landscapes needing good depth of field, we are already have some diffraction limitations. I believe that its already generally accepted that the APS-c sensors has a hard limit of f16. Another aspect is color. Reds (due to frequency) are more susceptible to diffraction that other colors (i.e., blue). Pentax does have some difficulty with reds.

Playing with the calculator a bit you can see that by increasing the resolution the diffraction just walks down through the aperture ranges. So there are some physical limits that do come into play here.

Bottom line is for me, when considering cost, physical body size and sensor performance (resolution, diffraction, rendering/colors), the K5 holds a lot of value and the K5IIs does add capability. Can you add pixels? - sure you can, however you start to play games with the law of diminishing returns. The question is where is the optimum point in all of this. I don't know, however - I believe that its substantially south of 100MP (in terms of APS-c and FF).

There are lots of additional discussions in this area:

• Camera Lens Quality: MTF, Resolution & Contrast
• Diffraction and Fraud in Digicams

The other aspect to all of this is compromise. Photography is full of compromises. It's a constant game of "whack a mole" - whack it here and it (a problem) pops up over there. Pentax especially has been very successful in terms of increasing the upper limits of ISO, which makes the need for fast lenses less important - other than for shallow depth of field, and diffraction. You can avoid diffraction by using a faster aperture, but depending on what you are shooting your DoF may suffer. That's great, until the size of the glass and it associated expense kicks in. This brings in wide angle. There is a point where you expand the field of view, you start to trade sharpness against the area each pixel needs to represent. The wider the field of view the larger the area each pixel needs to represent, hence the reduction in sharpness Additional resolution will help there - however we are back to the initial point of the post.

[photoman1]

Morning Class,

I have not tried to go back to find the post, as that would serve no useful function. The desire "that 16MP is nowhere near enough" and that "when we have 100MP ....", need to be based on what is obtainable, practical and useful. Unobtainium is not useful to anyone.

Where I was going was along the lines of diffraction. There are limits (laws of optics) that come in to play as various parameters get jockeyed around, in terms of designing camera systems. I do landscapes, cityscapes and the like. I also dislike hauling around large cameras. The APS-c sensor size as its packaged is pretty optimal for me. I really do not want to go much larger, nor spend a lot more. Nor do I want to dump my glass and start over again.

So back to diffraction. The K5 at 23.7 x 15.7mm yields about 370mm^2 providing a pixel pitch (diameter) of 4.81um. Just for comparison (FF) the D800's pixel pitch is 4.9um. So you start to deal with the Circle of Confusion or Circle of Diffraction.

Using the Cambridge Color site...

• Digital Camera Diffraction – Resolution, Color & Micro-Contrast

Lets look at the Diffraction Limited Aperture
..........................................................................APS-c (16.1MP).......FF (36MP)
Diffraction May Become Visible ..................................f7.2 ........................f7.4
Diffraction Limits Extinction Resolution ........................f9.1 ........................f9.2
Diffraction Limits Standard Grayscale Resolution ........f10.9 ......................f11
OVERALL RANGE OF ONSET .................................f7.2 to f10.9 ...........f7.4 to f11

What this tells me is that at f8 - good for landscapes needing good depth of field, we are already have some diffraction limitations. I believe that its already generally accepted that the APS-c sensors has a hard limit of f16. Another aspect is color. Reds (due to frequency) are more susceptible to diffraction that other colors (i.e., blue). Pentax does have some difficulty with reds.

Playing with the calculator a bit you can see that by increasing the resolution the diffraction just walks down through the aperture ranges. So there are some physical limits that do come into play here.

Bottom line is for me, when considering cost, physical body size and sensor performance (resolution, diffraction, rendering/colors), the K5 holds a lot of value and the K5IIs does add capability. Can you add pixels? - sure you can, however you start to play games with the law of diminishing returns. The question is where is the optimum point in all of this. I don't know, however - I believe that its substantially south of 100MP (in terms of APS-c and FF).

There are lots of additional discussions in this area:

• Camera Lens Quality: MTF, Resolution & Contrast
• Diffraction and Fraud in Digicams

The other aspect to all of this is compromise. Photography is full of compromises. It's a constant game of "whack a mole" - whack it here and it (a problem) pops up over there. Pentax especially has been very successful in terms of increasing the upper limits of ISO, which makes the need for fast lenses less important - other than for shallow depth of field, and diffraction. You can avoid diffraction by using a faster aperture, but depending on what you are shooting your DoF may suffer. That's great, until the size of the glass and it associated expense kicks in. This brings in wide angle. There is a point where you expand the field of view, you start to trade sharpness against the area each pixel needs to represent. The wider the field of view the larger the area each pixel needs to represent, hence the reduction in sharpness Additional resolution will help there - however we are back to the initial point of the post.

this is interesting from a mathematical and photographer's perspective

this is the kind of post that people want to see

[normhead]

The quote about the 100 MP was from falc and I think he may have been being a little facetious.

Anyway, here's my take on it from checkin out the photos at Image Resources comaparison

The max resolution for the average lens on an APS-c sensor is about 8 Mp ( practical max. not theoretical.) . On an FF it's about 16 Mp. Both the K-5 and the D800 are optimized in that they are over sampling. They are about half the size they could be...so within each resolved pixel, there is a pixel capturing each colour, RGB and one extra. So while the limit of the lens maybe 8 Mp, the oversampling improves upon the image until 16 Mp. At least that's my guess.

I've shown to my own satisfaction that taking APS-c to 24 Mp provides no measurable advantage, and that because 24 MP FF doesn't approach the 32 MP threshold for a 2x oversample it really doesn't offer much improvement over a 12 or even a 20 MP image. My guess would be to get to the next meaningful oversample will be 32 Mp APS-c or 64 Mp FF. SO maybe falc is thinking the next after that FF oversample at 124 would provide some real benefit. I don't know how else he got there. However, sharpness also depends on keeping diffraction and CA values to less than one pixel, and that obviously requires a better lens the smaller the pixels are, at some point adding Mp may not add additional sharpness. It's possible we are already as good as it's going to get.

I just look at pictures try and understand what's going on and make these things up...maybe someone else has an explanation as to why if you expand a Pentax k-5IIs image to the size of a D600 FF or a 24 Mp D3200 (APS-c) the K-5 image looks the same or better.

[Class A]

What this tells me is that at f8 - good for landscapes needing good depth of field, we are already have some diffraction limitations. I believe that its already generally accepted that the APS-c sensors has a hard limit of f16.

In order to capture all the detail of blue light on APS-C with a lens that is diffraction limited at f/8, you need 25MP (see Tab. 3). For f/5.6 you need 51MP and excellent lenses peak before f/5.6.

How does that make 16MP "essentially optimal"?

I'm not sure what you mean by f/16 being a "hard limit"? Surely you can shoot at f/22 as well. You don't need 16MP for f/11 already, so there doesn't seem to be anything special about f/16.

Can you add pixels? - sure you can, however you start to play games with the law of diminishing returns. The question is where is the optimum point in all of this. I don't know, however - I believe that its substantially south of 100MP (in terms of APS-c and FF).

The law of diminishing returns always applies. It is certainly not true that anything beyond 16MP does not make sense or is overkill.

Consult table 3 of the article I linked to and you'll see that I didnt make up 100MP for an APS-C camera. For a lens that is diffraction limited at f/1, you'd need 1612MP. However, such a lens does not exist. Below f/5.6 the vast majority of lenses is not diffraction limited anymore (due to aberrations).

---

Last edited by Class A; 11-04-2012 at 01:23 PM.

[kh1234567890]

For a lens that is diffraction limited at f/1, you'd need 1612MP. However, such a lens does not exist. Below f/5.6 the vast majority of lenses is not diffraction limited anymore (due to aberrations).

At last some sense in this thread.

Of course the hypothetical 1612MP sensor would only collect 1/100 of photons per pixel of a 16MP sensor, with correspondingly higher noise levels. Physics is a bitch

[bilybianca]

I have the K5 and don't feel any need to upgrade, it's a wonderful camera. I do some concert photography, and the better low light AF could have been nice. Only that I do most of my concert photos with my 85mm and 200mm manual lenses!
And after 27 years with Pentax, my best photos till today were taken with my Program A (ProgramPLUS) and *istDS bodies. As said above, sharpness isn't what makes a photo good. Even if it's nice to have as an option.

Kjell

[westmill]

Of course the name for 1612mp on an APSC sized sensor is know as an atom lol. Although plenty of light generating ability if you can split it lol. I think 16 mp on APSC is the optimal amount. Unless you dont need or use the higher ISO range then 24 milion would suit better. Constant impovements may see 24 working as well as 16 milion sometime in the future but I wouldnt like to hold my breadth waiting for it. For me at least, 16 milion is above and beyond what I need anyway so its all irelevent for me.

[bossa]

Pentax has a lot of R&D ahead of them, but I think such a camera would be a big hit. Even if it had AF comparable to the 7D or D7000, it would be a much more compelling option. Add to that a swiveling screen at it would be an industry first! Pentax hinted that they would make such a camera (at Photokina), as they basically have to if they expect anyone to buy their 560mm.

As for myself, I'm looking at the D800 with a couple of lenses- nothing outrageous, but certainly very expensive compared to most people's Pentax systems. Can't complain about their top-of-the-line AF, either!

Yes. that was my intention but look what happened to my signature!

[interested_observer]

In order to capture all the detail of blue light on APS-C with a lens that is diffraction limited at f/8, you need 25MP (see Tab. 3). For f/5.6 you need 51MP and excellent lenses peak before f/5.6.

How does that make 16MP "essentially optimal"?

I'm not sure what you mean by f/16 being a "hard limit"? Surely you can shoot at f/22 as well. You don't need 16MP for f/11 already, so there doesn't seem to be anything special about f/16.


The law of diminishing returns always applies. It is certainly not true that anything beyond 16MP does not make sense or is overkill.

Consult table 3 of the article I linked to and you'll see that I didnt make up 100MP for an APS-C camera. For a lens that is diffraction limited at f/1, you'd need 1612MP. However, such a lens does not exist. Below f/5.6 the vast majority of lenses is not diffraction limited anymore (due to aberrations).

I am not disputing you, I believe that we are each interpreting the same information in slightly different ways. I've read the referenced article several times since the IIs was announced. As you indicated for the ASP-C sensor with the blue light the sensor size is at 25MP, however looking at green/yellow you are at 13 and at the red color you are at 8MP. The hard limit I was referring to was using the APS-c sensor size at 16.1 MP at f16, the Maximum Circle of Confusion is 21.07um and the Diameter of Airy Disk is 21.3um, so you are diffraction limited (just scroll down to the last calculator - Diffraction Limit Calculator). You certainly can continue taking pictures at f22+, however you are loosing resolution due to diffraction - that was the point I was making.

Yes, I do see the figures in the table, along with the 100MP sizing, and that does assume some rather fine optics. In terms of chasing the 1612 to 527MP resolution, I doubt that I would be in the market for a f1 lens any time soon. My wife would kill me. 101 to 33MP at f4 is much more reasonable, given exceptional glass.

Right now in order to maximize resolution on the Nikon D800e you need the best glass Nikon has available. Reichmann in a number of his articles on AA or lack of AA indicates that. In order to get the best results from the IIs you are also going to need some good glass to maximize its performance.

Technology has advanced to the point, where to make use of significantly higher pixel densities you are going to have to shoot with higher quality, faster glass at larger apertures in order to make effective use of the additional megapixels, to achieve the higher resolution. Right now, I see that for the APS-c sensor size at 16MP, things appear to be pretty optimum for how I apply it (ISO 80, f8, wanting a good DoF, 14 bit A/D, great noise handling, tremendous DR, exceptional shadow handling) using glass with its maximum resolution in the f4 to f8 aperture range. The K5 physical body is a reasonable size (not extremely large) and can be acquired for a tremendous price. Tossing in the K30 and then the K-01 with its recent price cut, its time to go take pictures.

The K5 has a wonderful sensor for me. The IIs has some capabilities, that I find interesting, and could certainly make use of in the future.

[Class A]

Anyway, here's my take on it from checkin out the photos at Image Resources comaparison

Even Image Resource themselves say that their images are not suitable to make sharpness comparisons (only noise, colour, etc.). They just don't put in the effort to make sure that the focus point is consistent for each camera tested.

...maybe someone else has an explanation as to why if you expand a Pentax k-5IIs image to the size of a D600 FF or a 24 Mp D3200 (APS-c) the K-5 image looks the same or better.

Stop using sources that are not suitable for comparisons and some of the confusion may go away.

[sb in ak]

I really wasn't impressed with the K5 II/S on paper, though the sample images from the K5IIs aren't too shabby. Still seems like a half-baked upgrade to me, and I think it's just a stopgap measure as the true successor got delayed a bit. I won't upgrade from the K5 (just got mine this year) but if I were upgrading from the K7 or earlier, I'd definitely go K5IIs.

[Anvh]

Who says that increase in pixels numbers mean increase in resolution?

What would happen if they can fit 3 pixels (red, green and blue) on the place there is only 1 pixel. It would mean better accuracy when it comes to colours.
They might even add more pixels colours besides RGB just because they don't need the increase in resolution.