[bxf]

If you were paddling your canoe across the desert and the wheels fell off, how many pancakes would it take to shingle a doghouse?

None, because ice cream has no bones.- Eric

Hahaha. Are we attempting to create a new debate, to rival or perhaps sidetrack the existing endless one?

Though I am expecting the new K1 to have better noise characteristics and more resolution than the bright yellow one I shoot with now... - Eric

You have a bright yellow K1?

[clackers]

It can be easily shown that the size of the sensor has a direct impact on the noise in an image, If we use your pixel level metrics at DXO you can that the K3 gathers less signal for that portion of the sensor used.


When the screen view is selected what we are in essence doing is comparing a 16mp crop of the K3 sensor to the full 16mp k5 sensor ( viewing at 100 % ) with no scaling and you can see that the noise performance of that K3 crop drops by the factor that we have cropped the k3 sensor, a factor of about 1.224 times.
If we compare the K5 at iso 150 and the K3 at iso 100 they are very close to noise performance demonstrating that the size of the has a direct influence on how much noise would show up in the image ( this is without any scaling).

You are worrying me, Ian.


There is no crop - where did you get that from?


They are both APS-C cameras.


The K-3 performs less well because in order to fit an extra 8 Mp on the sensor, each pixel pitch has been reduced (the electron well capacity is less).


It was a concern K-5 owning Pentaxians had on this forum ... would an upgrade hurt their high ISO performance?


My answer in such threads has consistently been yes, but offset by the extra downsampling possible in post.

[clackers]

No one has ever said that downsampling has anything to do with sensor size (except you).

I have said the opposite of that, Gimbal.


Your absurd claim without evidence was:

"The print graph shows us that when the 810 uses the full frame it will outperform the K5. Not because it has bigger pixels (which it doesn't), but because it has a bigger total sensor area."

Ask yourself: Why would this not be true in the screen graph as well?

You keep denying you have a shaky grasp of the basics, but your posts are repeatedly saying otherwise. ☺

So since we all understand downsampling, what is your problem with it?

Now I can't say being Swedish is an excuse.


My post 218 is unambiguous - don't pretend otherwise, please.

[BrianR]

"The print graph shows us that when the 810 uses the full frame it will outperform the K5. Not because it has bigger pixels (which it doesn't), but because it has a bigger total sensor area."

Ask yourself: Why would this not be true in the screen graph as well?

Take a few minutes and try to see Gimbal and Fogel's point of view. Trying to understand what they're saying will be worth far more than adamantly declaring them to be wrong or somehow unable to understand.

The d810 and the k5 have the same size pixels (or close enough). The d810 has a larger sensor. Therefore, the d810 has more pixels, and due to downsampling will 'win' with regards to noise when it comes time to compare resulting images at the same print size. Bigger sensor => more pixels => less noise for a given print size.

Yes, it is true that a larger sensor does not guarantee more pixels, everyone knows this. A larger sensor does imply an increase of either pixel size or pixel count (or both). One could go down, but then the other one goes up even more. The net result is an improvement in noise when comparing the pictures at the same print size (I'm assuming 'all other things' are equal, so keep those canon sensors away from our sony sensors for the moment, especially those 10 year old ones). With this view, the statement "a larger sensor means less noise in the final picture at the same iso" is a pretty reasonable thing to say (with the caveat that the smaller sensor might be 'good enough' that it's effectively noise-free to human eyeballs at a given print size).

I don't want to put words in your mouth so please stomp all over this if I'm misrepresenting you, but I think you are opposed to this phrasing since the larger sensor area isn't the direct mechanism that's helping out to noise? Since a larger sensor will necessarily impact two things that help with noise for a net improvement, I think this is going to get you into repeated misunderstandings.


This was probably a waste of time, but I'd otherwise be waisting my time eating ice cream sandwiches trying to become a full framed photographer, so I figured it was worth a stab

[Fogel70]

You are worrying me, Ian.


There is no crop - where did you get that from?


They are both APS-C cameras.


The K-3 performs less well because in order to fit an extra 8 Mp on the sensor, each pixel pitch has been reduced (the electron well capacity is less).


It was a concern K-5 owning Pentaxians had on this forum ... would an upgrade hurt their high ISO performance?


My answer in such threads has consistently been yes, but offset by the extra downsampling possible in post.

That is how the "screen" comparison work.
It compares the same number of pixels from the cameras, and if one camera has smaller pixels it will then of course use a smaller crop when compared. As I said before, this makes an advantage for sensors with bigger pixels as that will make them use larger crop of the sensor.

See below images, this first 4 crops represent the screen comparison where K50, K3 II and K-S2 use different sized crops in the comparison as that is what you get the comparing the same amount of pixels from same sized sensors with different pixel count.
The the next 4 crops image represent the "print" comparison where you compare the whole image. This by using same sized crops from K50, K3 II and K-S2.
I also added D810 to show what happen if comparing a larger sensor with same pixel size as the smaller sensor used in K50.

[Zygonyx]

You can see a sensible, though absolutely not decisive improvement, in the D810 crop...

[Digitalis]

My measurements from cameras I work with through analyzing a 50X50 px [2500pixel] area of a Gretag Macbeth colour checker middle grey swatch:

Exposure levels were dictated through a flash light meter with lumosphere - f/32 1/125th ISO 12800*

Standard deviation K5IIs - 16.65
Standard deviation D810 - 23.92
Standard deviation Nikon D4 - 16.22
Standard deviation Canon 5DMk III - 21.72


* using such a High ISO with flash is ridiculous.

[Fogel70]

Baffling, Fogel.

You just said it yourself: "Same pixel design".

It is always about the pixel size for the image capture and pixel numbers for the post processing.

A large sensor does not have to have more pixels.

No of course not, but it is still the larger area of the sensor that improve the image noise.

A Nikon D750 has no more pixels than a K-3, and a Canon 1D X has less.

If you compare a APS-C crop from D750 and 1DX and compare with K3 you will see that the noise level are not far off for all of them.

The Canon 5DS has so many smaller pixels crammed into it that a lot is expected of downsampling to compensate, and no one is impressed with the results.

By looking at the DxO mark scores for the above mentioned cameras (and some more) and compensate for sensor size you would get the following.
Pentax K3 II has a "Sports score" of ISO 1106 and with a FF version (54 MP) of that sensor it would score ISO 2461.
Nikon D750 has a "Sports score" of ISO 2956 and with a APS-C version (10,7 MP) of that sensor it would score ISO 1313.
Canon 1D X has a "Sports score" of ISO 2786 and with a APS-C (1.5 crop) version (8 MP) of that sensor it would score ISO 1238.
Canon 5DS has a "Sports score" of ISO 2381 and with a APS-C (1.5 crop) version (23 MP) of that sensor it would score ISO 1058.
Nikon D810 has a "Sports score" of ISO 2852 and with a APS-C version (16 MP) of that sensor it would score ISO 1267.
Nikon D7200 has a "Sports score" of ISO 1333 and with a FF version (54 MP) of that sensor it would score ISO 2999.

[Gimbal]

I have said the opposite of that, Gimbal.

Yes, you are fighting something that no one has claimed.
And once and for all, there is NO correlation between sensor size and shoe size of the photographer. None. Claiming that is ridiculous.

Your absurd claim without evidence was:

"The print graph shows us that when the 810 uses the full frame it will outperform the K5. Not because it has bigger pixels (which it doesn't), but because it has a bigger total sensor area."

Ask yourself: Why would this not be true in the screen graph as well?

Because in the screen graph we are comparing single pixels, regardless of how much area they take up on the sensor. Thus a big pixel is always better then a smaller one (if the technology around it is equal). In this case they are roughly of the same size and thus performs pretty much the same. And that is what the graph tells us, that graph does not consider total sensor area, it just looks at individual pixels.

You might ponder this question.
If a large pixel is better then a small pixel, isn't a larger sensor better then a smaller sensor for just the same reasons?

[Rondec]

Full frame cameras have roughly one stop better noise performance as compared to APS-C. Differences in sensor technology don't seem effect this nearly as much as they effect the amount of dynamic range the sensor has at a given iso. At low isos, the differences will be imperceptible, but at higher isos it certainly is clearly seen.

[Mattox]

Keep your eyes open wide everyone. The Pentax FF website isn't working, which means they are currently updating the website and have taken it offline I'd say. That's what happened on the 24th of December. Let wait a few hours to see what news there is, if I am right

[beholder3]

Full frame cameras have roughly one stop better noise performance as compared to APS-C.

Not if you compare comparable pictures with the same DoF. As soon as the pictures are technically comparable the noise performance automatically is the same.

There are enough people shooting groups of subjects (where you need a predefined DoF not equal the minimum) wondering why the FF advantage in noise is zero.

If you discount the difference in DoF between FF/APSC as irrelevant then that is fine (though has been argued over by trillions of threads).

Just be aware that if you compare the noise of pictures with different DoF you also have to accept that comparing a mFT 300mm F4 lens to a FF 600mm F4 lens is the only right approach.

Either you apply the equivalence rule of thumbs to your comparisons or you don't. People tend to do cherry picking on this for propaganda reasons.

[Rondec]

Not if you compare comparable pictures with the same DoF. As soon as the pictures are technically comparable the noise performance automatically is the same.

There are enough people shooting groups of subjects (where you need a predefined DoF not equal the minimum) wondering why the FF advantage in noise is zero.

If you discount the difference in DoF between FF/APSC as irrelevant then that is fine (though has been argued over by trillions of threads).

Just be aware that if you compare the noise of pictures with different DoF you also have to accept that comparing a mFT 300mm F4 lens to a FF 600mm F4 lens is the only right approach.

Either you apply the equivalence rule of thumbs to your comparisons or you don't. People tend to do cherry picking on this for propaganda reasons.

I understand equivalence. My point is merely that if I am shooting at iso 1600 on a full frame camera and iso 1600 on a crop frame camera, I will have less noise in the final output.

I guess you could just as well say that the reason for the noise improvement with full frame is that there are faster equivalent lenses available for it. We are limited to the lenses we have. There is no 50mm f1.4 equivalent on APS-C, no 35mm f1.4 equivalent, and no 24-70 f2.8 equivalent. And so we have a situation where full frame performs better, as long as you can deal with a little less depth of field.

There are few times in dark situations -- say a museum -- where I wouldn't be willing to trade a little less depth of field for the ability to shoot an image with a little less noise.

[Kunzite]

Keep your eyes open wide everyone. The Pentax FF website isn't working, which means they are currently updating the website and have taken it offline I'd say. That's what happened on the 24th of December. Let wait a few hours to see what news there is, if I am right

Is working here (no update, perhaps next week).

Down For Everyone Or Just Me -> Check if your website is down or up?

[Fogel70]

Not if you compare comparable pictures with the same DoF. As soon as the pictures are technically comparable the noise performance automatically is the same.

The normal procedure to compare noise is to compare at the same ISO with same light conditions.

There are enough people shooting groups of subjects (where you need a predefined DoF not equal the minimum) wondering why the FF advantage in noise is zero.

I don't think anyone says that you always can use the advantage of larger sensors in every image. but there are situations where larger sensors have advantages. FI when the smaller sensor camera can't give you shallow enough DOF, or when you want to stop down the lens for better IQ and still get limited DOF.

Another reason for large aperture lenses is to use large aperture in low light conditions when no other option can give you acceptable images. As long as you don't have faster lenses that can compensate for cameras using smaller sensors, then larger sensor will have an advantage.
I believe these reason are why most of us use large sensor cameras instead of small sensor cameras.

If you discount the difference in DoF between FF/APSC as irrelevant then that is fine (though has been argued over by trillions of threads).

Just be aware that if you compare the noise of pictures with different DoF you also have to accept that comparing a mFT 300mm F4 lens to a FF 600mm F4 lens is the only right approach.

I guess if you want to use the same FOV and DOF in the images and use equivalent lenses, it then should be mft 300/2.8 compared with a FF 600/5.6.

Either you apply the equivalence rule of thumbs to your comparisons or you don't. People tend to do cherry picking on this for propaganda reasons.