[mattipuh]

Stupid question, but its unclear for me so thought asking.
Will setting with less resolution provide better dynamic / better high iso performance.?

As you cannot choose resolution when shooting in RAw guessing answer is no..?

[clackers]

No. You guessed why.

[Adam]

As you cannot choose resolution when shooting in RAw guessing answer is no..?

Yup You'd only get less noise with a lower-resolution sensor that has physically larger pixels.

[mattipuh]

Yes, thought so. Meaning resolution setting differences in one camera does not improve / decrease iso performance. It would be nice if it would be possible, as often quite little reso is enough. But megapixels are easier to market than high dynamic range, even though often the latter provides nicer images...

[clackers]

Yes, thought so. Meaning resolution setting differences in one camera does not improve / decrease iso performance. It would be nice if it would be possible, as often quite little reso is enough. But megapixels are easier to market than high dynamic range, even though often the latter provides nicer images...

Yeah, look, DXO claim that more pixels *do* improve dynamic range.

They don't actually measure it, they automatically make the adjustment to the scores, justified or not, in a process they call 'normalization'.

I can see the case for their doing this to noise performance, since downsampling is basically averaging.

But dynamic range refers to the extreme ends of light and dark, which should be reduced by averaging.

Yes, they provide an explanation that blacks can really be black when noise is removed, but I'm not convinced as to the scale of this effect, and they never actually provide measurements.

[Alahmadi]

yes. Basically the smaller the megapixel amount the more big is each pixel which allow for more light to be stored in.

[fredralphfred]

There's a youtube video from Tony Northrup here:

You Tube

You Tube

where he explains that the TOTAL noise is dependent on sensor size (and to a lesser extent sensor "generation"), and the NOISE PER PIXEL is dependent on the number of pixels. So if done properly, "pixel binning" (or reducing megapixels) CAN reduce noise. A quick Google search on "megapixels noise trade-off" will bring up several articles and posts on other forums discussing this question.

This is why noise reduction algorithms reduce detail -- that's the biggest way they can combat noise. They are (usually) aimed at removing the least amount of detail possible to reduce noise, but if you go pretty heavy with them and downsample, you'll receive a picture very similar to one that you "binned" with a decent algorithm. Most cameras' built-in JPEG processors aren't optimized for this. This is a big reason most folks recommend you shoot RAW and do this sort of thing in post.

I know this seems to disagree with what most of the previous posters said, but I think you'll find we agree in the end. I think you're asking about TOTAL NOISE (if you "added up" the total noise in your final picture, or viewed it at a constant size regardless of megapixel count), which WILL NOT CHANGE, but they think you're asking about NOISE PER PIXEL (what you get when you "pixel peep," or the amount of noise at 100% zoom).

[marcusBMG]

I would not regard T. Northrupp as an authority on any of these matters (albeit his presentational skills are good). Check out his videos on crop factor and effective f stop, where in extensive rabbiting on about this he completely fails to make clear that a different size sensor does not affect light intensity and therefore the only signiificance is in fact relating to depth of field.

[Simen1]

Note that comparing 1:1 is when you compare differently zoomed crops. One image might be zoomed in 5x while the other are zoomed in 8x. Thats why "print" on DxO are the only apples to apples comparison.

To the original question: Comparing same sensor sizes, with same lenses and settings, you will note very little difference. Like Tony Northrup says, 1/3 of a stop is a lot in this context. But if we shrink pixel sizes way beyond todays m43, APS-C and FF pixel sizes there will be a steep increase in noise when the pixel densities are approaching the wavelength of light. Now I talk about hundreds of megapixels in those sensor sizes. BSI and Isocell techniques will help a lot (up to 0,5 - 1 stop?) on pixel sizes around 1 micron when using large aperture lenses. Why only large aperture lenses? Its because they make the light hit the sensor from a wide angle (edges of the back of the wide rear glass) in stead of more or less normal to the sensor plane. Due to diffraction and airy disc size only large apertures make sense with densely packed pixels. Thats why mobile phones (using 0,95 - 2,0 micron pixels) always have constant aperture around f/1,9 - f/2,4. Stopping down doesn't make sense resolution wise.

But there is more to consider. Dynamic range. The pixel wells have an upper limit to how many electrons (converted from photons) they can hold. That limit is largely dependent on the storage area per pixel. So if we replace one pixel with four smaller ones, it usually means they will hold 1/4 the number of photons. This isn't causing any problems so far since the sum of those four would be the same as in one large one. The problem occurs in the other end of the scale. Photon noise per pixel will be higher, actually 4 times higher (or a little more if the pixels are approaching 1 micron). Averaging 2x2 pixels will bring down the noise to 2 times more then the previously large sensor. The net effect is that the signal to noise ratio decreases by a factor of 2, when 2x2 pixels are averaged. Smaller pixels give us less DR, and this is not restorable by averaging. If you look through raw samples from a large range of sensor sizes you will see that this is correct. Just start out with the extremes, phone cameras with raw vs CMOS medium format to see how obvious the difference are, and see that its consistent with formats in between.

That was purely about photon noise, but there are other noise sources. Read noise are normally quite small, but when reading 4 pixels in stead of one makes four instances of read noise in stead of one. Averaging that noise component you will get 2x more read noise from those 4 pixels compared to 1 larger pixel.

---------- Post added 02-21-16 at 04:21 PM ----------

I would not regard T. Northrupp as an authority on any of these matters (albeit his presentational skills are good). Check out his videos on crop factor and effective f stop, where in extensive rabbiting on about this he completely fails to make clear that a different size sensor does not affect light intensity and therefore the only signiificance is in fact relating to depth of field.

Thats nonsense. You should read more about that subject before jumping popular myth conclusions.

Here are one of PF own knowledgeable users writings on the subject.

[marcusBMG]

Thats nonsense. You should read more about that subject before jumping popular myth conclusions.

No it isn't (though I should have said more precisely "that a different sized sensor does not affect light intensity and therefore, for practical purposes with cameras whose sensor sizes are not extremely different, the main photographic significance pertains to depth of field for same field of view".
I watched the Northrupp videos on crop factor and they left me well confused. So is my 50mm f2 lens on my K-r really 75mm f2.5? Only by referring to other sources did I get things straight: that an f2.8 lens is always an f2.8 lens, it's a physical property of the optics. I don't pretend to know if Northrupp's simple failure to make that plain is accidental or deliberate, I just don't see him any more as a reliable informant.

Your linked article looks very informative, but requires study. It does say:

In general, the sensor size does not affect image quality with equivalent parameters, all sensor sizes within one so-called equivalence class deliver indistinguishable images.

[Simen1]

If you are confused you should read more and sort out these things. I have.

So is my 50mm f2 lens on my K-r really 75mm f2.5

No of course not. 50mm and f/2 are physical measurable lengths in your lens. Thats why we talk about equivalences, not physical transformations.

Lets use 2x vs 1x crop factor as an example for the mathematical ease of it. Moving a 50mm f/2 lens from a 1x to a 2x crop camera will not change the physical lengths, but the angle of view will equal the use of a 100mm lens on the 1x camera. Hence 100mm equivalent (again not physical). The light intensity per unit area of the sensor also remains the same, but since the 2x sensor has 1/4 the sensor area it will pick up only 1/4 of the photons the larger sensor would with the same lens. Now what does 1/4 the amount of light do to the total captured light in the whole image? Less light is less signal and thus lower signal to noise ratio. Working out the math it gives a 2 stop difference in signal to noise ratio. Hence the same noise as when using f/4 on the 1x camera. A f/4 equivalent (not physical) lens.

In other words these should give you equivalent images (when using otherwise same settings, ideal lenses, same subject and so on. In short, apples to apples comparable)
2x crop camera with 50mm f/2
1x crop camera with 100mm f/4

If you are still confused, please take your time to sort out the math and physics or carefully read Falk Lumo's article until you understand each step he does. Tony Northrup did his math carefully and correct.

[nozoom]

But dynamic range refers to the extreme ends of light and dark, which should be reduced by averaging.

Well, dynamic range (DR) is actually a ratio. It is the ratio between the maximum measurable signal and the lowest measurable signal. The later is set to be equal to the noise floor in photography. A more conceivable way of explaining DR (when measured in EV) is that it tells how many times you can half the maximum possible signal (just before burn out) before it drowns in noise. If you reduce the noise the result will therefore be increased DR since this will result in a lower signal being usable. The signal in this context is the light from the scene that you want to capture (or more accurate the electrons or voltage representing it). This is not only in theory. It is quite easily demonstrated.

I have often used averaging to reduce noise, but mostly I do it by taking several images that I align and average pr pixel in post processing.

Cheers,

[marcusBMG]

We are sidetracking here, my first post was not to start a general discussion of crop factor but to respond to fredralphfred's post, in which he cites a Northrupp video, to say that I have in the past found Northrupps presentations misleading (and I'm not the only one just glance at the comments for the videos I refer to).
Lets not hijack the OP's qq.

[fredralphfred]

We are sidetracking here, my first post was not to start a general discussion of crop factor but to respond to fredralphfred's post, in which he cites a Northrupp video, to say that I have in the past found Northrupps presentations misleading (and I'm not the only one just glance at the comments for the videos I refer to).
Lets not hijack the OP's qq.

I chose that video because I thought Mr. Northrup had a good, clear explanation of why and how noise-per-pixel and resolution are related. He is not a lone voice espousing this position. Do a Google search on "pixel binning" and read the various explanations if you want more convincing. Also consider that some manufacturers, especially Fuji, often include low-noise "pixel binning" modes on their cameras for low-light situations, allowing the camera to save smaller files if there isn't enough light to effectively use all their resolution.

Concerning the Northrup equivalence video: you may disagree with him, but his presentation was not misleading. I watched it and found it agreed with the laws of mathematics and physics as applied to optics. You might need to watch it again. Equivalence is not an easy subject, but again you can find many other explanations that agree with his by entering "photographic equivalence" into a Google search box. This link is probably one of the best written explanations I've found, complete with formulas: Equivalence.

[SteveM]

There's a youtube video from Tony Northrup here:

He makes a few points but aren't these just generalizations? If raw is truly raw, we shouldn't be able to tell the difference in a shot taken by two manufacturers provided the same sensor is used (or possibly the same MP and physically sized sensors.)? In the section that speaks to ISO scores and the stops relative to the average across cameras, the D4s shows better performance but the only explanation is "well that's a $7,000 camera". I would have liked to know why that camera is performing better than the others, as it seems to go against the point of the article.