[builttospill]

24MP would still be fantastic, but why not catch up to 36MP? Why start out behind?

I believe this is the question and reason this discussion has continued between the group wanting a ridiculously high MP count and the one wanting a low one. I think 36MP is starting behind. What in the world would I need 36MP for anyway?

I don't crop my photos, and I rarely spend more than 30 seconds per photo in post-processing. A larger file is just wasting my time and storage space. I don't shoot sports, so the unlimited buffer in my K10D isn't used but it's there if I need it. I have the Pentax cameras providing the highest-quality 100 ISO images out there. I only have 10MP but it's enough for several of my photos to be used in bulletin billboards along the freeway. I understand these are my preferences and shooting style, just as wanting a 50+ MP sensor may be your style. It would be ideal to have two models to please both crowds.

A 16-18MP FF would be perfect. Yes, technology has improved but use that same technology to produce a new 16MP sensor and it would be unbeatable. Make the next Pentax cameras, FF and APS-C, with the largest optical viewfinders out there and you have another winning element.

[pid]

a FF 36MP would let you use your APS-C lenses in same resolution you had on the K-5's

[clackers]

A 16-18MP FF would be perfect. Yes, technology has improved but use that same technology to produce a new 16MP sensor and it would be unbeatable.

I'm not sure you'll get your wish, Builttospill. Sony don't have such a new sensor to sell to Pentax. Their recently announced 12Mp doesn't fit your bill either.

[builttospill]

I'm not sure you'll get your wish, Builttospill. Sony don't have such a new sensor to sell to Pentax. Their recently announced 12Mp doesn't fit your bill either.

I think you're right, but it doesn't hurt to have a wish.

[carrrlangas]

I´ll have a 36Mp APS-C K-3 successor to use my APS-C Pentax K mount lenses (which are a big investment for me). So... I´ll take a FF with as many Mp as technology allows, to use my legacy 135 format Pentax K mount lenses (and would probably add DFA limiteds when available used, for good prices)

[clackers]

I´ll have a 36Mp APS-C K-3 successor to use my APS-C Pentax K mount lenses (which are a big investment for me).

It's been suggested that for APS-C 16Mp gets the size of each pixel right, and that 24 is probably too much as far as noise in low light goes.

If true, 36Mp would be worse.

[dtmateojr]

It's been suggested that for APS-C 16Mp gets the size of each pixel right, and that 24 is probably too much as far as noise in low light goes.

If true, 36Mp would be worse.

It will be so much worse. People are forgetting that dxomark scores are normalized to a small 8x10 print, giving a false impression that a 36mp image is free of noise. The K5 is already superior to the K3 in IQ. If you buy a 36mp camera so that you can resample it at 8x10 then something isn't right.

I did my calculations and found out that a 16Mp camera with SNR as good as the 7-year old Nikon D700 will beat the D800 at 20x30 prints and larger. People are blinded by dxomark's very unfair 8x10 resampling.

Larger is not better.

[carrrlangas]

It's been suggested that for APS-C 16Mp gets the size of each pixel right, and that 24 is probably too much as far as noise in low light goes.

If true, 36Mp would be worse.

Maybe for a certain sensor pixel size isn´t the only parameter to determine performance). According to the review on this site the K-3 is not noisier than the K-5IIs at base ISO. (I honestly don´t care that much about performance above ISO1600). About low light, I dare to call it "bad light" aswell, and most good pictures are made in good light.
Larger dynamic range at high ISO is appealing though..

[ElJamoquio]

It's been suggested that for APS-C 16Mp gets the size of each pixel right, and that 24 is probably too much as far as noise in low light goes.

If true, 36Mp would be worse.

Making the pixel sizes smaller doesn't adversely affect the noise in the overall picture. There's a tiny edge effect but doesn't change substantially from 16MP to 24MP.

More MP = More noise is a myth that needs to die.

---------- Post added 05-12-14 at 03:31 PM ----------

It will be so much worse. People are forgetting that dxomark scores are normalized to a small 8x10 print, giving a false impression that a 36mp image is free of noise. The K5 is already superior to the K3 in IQ. If you buy a 36mp camera so that you can resample it at 8x10 then something isn't right.

I did my calculations and found out that a 16Mp camera with SNR as good as the 7-year old Nikon D700 will beat the D800 at 20x30 prints and larger. People are blinded by dxomark's very unfair 8x10 resampling.

Larger is not better.

You're misinterpreting the meaning of that picture and the 'resampling'.

[dtmateojr]

The point is, at a 16mp print, the D800 sensor will start showing it's inferiority vs a 7-year old sensor. That's just half it's image size. At print sizes larger than that it will start showing it's true colours -- that after years of RnD it hasn't really improved much. The D4 sensor will easily beat the D800 at 16Mp. An APS-C sensor at 16mp with the SNR of a D700 (very possible) will easily match the D800 at 16mp print sizes.

Now if a 36mp is printed at 16mp you are just wasting it. Print larger and the image quickly degrades. There is no advantage at all.

[ElJamoquio]

Your post is hard to follow.

Why would you 'print at 16MP'? Why would you reframe an image to match a certain pixel size?

FYI, the APS-C 16MP will precisely match the FF D800 cropped to 16MP, for what it's worth.

I agree that the sensor tech hasn't changed at all in the past four years. From years 4-7, though, it changed quite a bit.

[dtmateojr]

In the same way that dxomark "prints" ALL their tests at 8mp (8x10). There is no misinterpretation there. I'm using the same math that dxomark uses. But instead of their measly (read unfair to good sensors) 8mp print I calculated mine at 16mp.

---------- Post added 05-13-14 at 10:06 ----------

Just in case you are unaware, the "print" SNR graphs in dxomark are NOT measured SNR. They are calculated based on a 8mp print from the MEASURED SNR as shown in the screen graphs.

[ElJamoquio]

I am aware of their methodology. The resizing of their data to one arbitrary standard is the same as resizing to any other arbitrary standard, because they are reporting signal / noise, rather than just reporting absolute noise.

If you're getting a different answer by rescaling to 16MP then something is wrong in your math.

[dtmateojr]

It's the same math that can be done by anyone who can use a calculator At 16mp print size the D4 has SNR of 39.7, and the D800 is 39.3. Just for reference the SNR of the D700 at 12mp is 39.4

---------- Post added 05-13-14 at 10:29 ----------

BTW at 36mp print size the D4 can do SNR of 36.178 while the D800 is only 35.8. The D4 even upsized to a whooping 36mp beats the D800 in its own game

[builttospill]

If you're getting a different answer by rescaling to 16MP then something is wrong in your math.

It's the same math that can be done by anyone who can use a calculator At 16mp print size the D4 has SNR of 39.7, and the D800 is 39.3. Just for reference the SNR of the D700 at 12mp is 39.4

dtm's point is valid though and I don't need math or calculators to see the difference. I don't care about tests and numbers, but real world samples from my work and others'. I'll say it again more clear this time: My eight year old sensor at ISO 100 in my K10D blows away all newer sensors, whether it's the bloated 24 or 36MP ones. Take a look at landscape photographs (Web-sized for this Forum) with blue skies in them from the K-3 or D800 compared to one from my sensor made in 2006. Newer technology is not an improvement and the massive MP count magnifies this.

That being said, I understand the sensors compared above are different technologies and that makes a huge difference. I'm also a bit envious of the newer cameras' abilities in low light, but because less than 1% of my digital photos are above ISO 100 I'm sticking with my old camera until technology offers an improvement. Or until Pentax releases a low MP FF.