[FunkyMonk]

If so how is it affected on the K10D and the K20D?

[Gooshin]

sensor resolution is sensor resolution, it is native and never changes.

crop factor is a term used in the photo industry to compare lenses used on different formats. (medium format = MF, 35mm or full frame = FF, and APS-C = "croped")

[Sailor]

If so how is it affected on the K10D and the K20D?

sensor resolution is sensor resolution, it is native and never changes.

crop factor is a term used in the photo industry to compare lenses used on different formats. (medium format = MF, 35mm or full frame = FF, and APS-C = "croped")

Are you guys talking about the same thing? Funky, are you actually asking about crop factor of lenses, or relative effects of PP cropping of images made with sensors of differing resolution?

Jer

[FunkyMonk]

I am speaking in terms of the number of pixels. Is it reducing the number of pixels? Here is an article that I read that seems to say that it reduces the resolution. Maybe I am reading it wrong though.

The Real Crop Factor explained

[FunkyMonk]

sensor resolution is sensor resolution, it is native and never changes.

crop factor is a term used in the photo industry to compare lenses used on different formats. (medium format = MF, 35mm or full frame = FF, and APS-C = "croped")

That's what I was thinking but I had a guy ask me about that today and I did a search on it and found the article posted above. Maybe I am misreading it but it appears from the article that it is saying that it reduces the number of pixels.

[Marc Sabatella]

I am speaking in terms of the number of pixels. Is it reducing the number of pixels?

I am not sure what you are asking. Are you asking if a camera that is advertised as 10MP actually has less than 10MP because of crop factor? Of course not. It's got 10MP, period. The "crop factor" comes from the those pixels are crammed into a an area smaller than a rectangle of 35mm film - it's as if someone took the piece of film and cropped off the edges. But when they say there are 10MP in a camera, that's how many pixels there are.

Here is an article that I read that seems to say that it reduces the resolution. Maybe I am reading it wrong though.

Could be. You might want to point out the specific paragraph that has you confused. I took a look at the article, and it appears to have not been written by a native English speaker (he's French). As a result, while the technical information is probably correct, it is not expressed as well as it could be.

Anyhow, the actual answer your question is that the "crop factor" *does* affect resolution, but not in the way you seem to be thinking. You have 10MP to work with on a 10MP camera - that much is unquestioned. But because those 10MP are crammed into a smaller area than a piece of film, that means you have to blow the image up more to get a given sized print than you would to create an image from a film camera. The effect is *exactly* like blowing up the image from a 35mm camera to 6x9", then cropping out the edges of the print to yield a 4x6" print. It creates the illusion of having shot with a longer lens, but on that 35mm camera, you'd have gotten better results actually shooting with the longer lens in the first place rather than shooting with the shorter lens and then blowing p the image bigger.

Somehow, I have a feeling the preceding paragraph is going to create more questions than it answers...

[FunkyMonk]

I am not sure what you are asking. Are you asking if a camera that is advertised as 10MP actually has less than 10MP because of crop factor? Of course not. It's got 10MP, period. The "crop factor" comes from the those pixels are crammed into a an area smaller than a rectangle of 35mm film - it's as if someone took the piece of film and cropped off the edges. But when they say there are 10MP in a camera, that's how many pixels there are.



Could be. You might want to point out the specific paragraph that has you confused. I took a look at the article, and it appears to have not been written by a native English speaker (he's French). As a result, while the technical information is probably correct, it is not expressed as well as it could be.

Anyhow, the actual answer your question is that the "crop factor" *does* affect resolution, but not in the way you seem to be thinking. You have 10MP to work with on a 10MP camera - that much is unquestioned. But because those 10MP are crammed into a smaller area than a piece of film, that means you have to blow the image up more to get a given sized print than you would to create an image from a film camera. The effect is *exactly* like blowing up the image from a 35mm camera to 6x9", then cropping out the edges of the print to yield a 4x6" print. It creates the illusion of having shot with a longer lens, but on that 35mm camera, you'd have gotten better results actually shooting with the longer lens in the first place rather than shooting with the shorter lens and then blowing p the image bigger.

Somehow, I have a feeling the preceding paragraph is going to create more questions than it answers...

This is the part of the article that confused me.

Bingo! Certainly we lose in wide angle, but we gain in long focal lengths, when getting closer to the subject becomes problematic. Yes, but slow down ! We considered here that our two sensors had the same number of pixels, now it is not always (if not rarely) the case. . There are a lot of sensors, APS with 6mp (megapixels), 8mp, 10mp, 12mp, and 35FF 11mp, 12.7mp, 16mp. How and why does it affect crop factor ?
If our 35FF sensor has more pixels than our APS sensor, we could manually crop the image of the 35FF, keeping same number of pixels as our APS. So, we could enlarge our FF shot, still at same printing size. The question which arises then is : how much pixels do we need on our 35FF sensor to crop at 1.6x ratio and still the have same number of pixels as in APS, what shall cancel then the crop factor of the APS. To determine this coefficient, it is enough to think in pixel density. The pixel being a surface, if one want to obtain the same number of pixels once cropped at 1.6x, our FF will have to have 1.6 ² = 2.56 times more pixels than the APS.

Let's take some concrete cases. A 10D is an APS, with 1.6 cropfactor, and 6.3mp. A 5D is 35FF (no or equal to 1 cropfactor), and 12.7mp.
If we crop the 5D like the 10D, we get 12.7 / 1.6 ² = 4.96mp. Our 10D has indeed an advantage, but smaller than its cropfactor of 1.6.
Let's take now an 1DsII of 16.6mp against our 10D. Cropping 1DsII shot like 10D gives us 16.6 / 1.6 ² = 6.5mp ! Same resolution, the advantage in long focal lenghts of the 10D APS on the 1DsII 35FF is thus non-existent !

It appears that the cropfactor alone is not a good indicator, and that it must be moved closer to the pixel density. I thus take the liberty to advance another way of seeing things, by defining a "real magnification factor ".
So be pr the pixel ratio between the number of pixels of the big sensor on that of the small sensor, and cf the cropfactor of the small sensor in regard to the big sensor (here one 35mm). I put down :

rmf = cf / root (pr)

Let us verify it with both quoted cases above : between 10D and 5D, pr = 12.7 / 6.3 = 2.015
10D rmf compared to 5D = cf / root (pr) = 1.6 / root (2.015) = 1.12
10D real magnification factor compared to 5D is so 1.12 !
Between 10D and 1DsII, pr = 16.6 / 6.3 = 2.63. rmf = 1.6 root (2.63) = 0.98 !
Indeed, our 10D, in spite of its 1.6 cropfactor, got a real magnification factor less interesting than 1DsII !

[stevebrot]

That's what I was thinking but I had a guy ask me about that today and I did a search on it and found the article posted above. Maybe I am misreading it but it appears from the article that it is saying that it reduces the number of pixels.

I think you are reading the article wrong.

A pixel is a pixel is a pixel. 10 million of them is still 10 megapixels regardless of sensor size.

The image on an APS-C is only "cropped" in the sense that it has a reduced field of view relative to a larger sensor with the same focal length lens.

Mark's explanation above of the how sensor size affects image resolution is excellent. I can't do a better job of explaining.

Steve

[*isteve]

Anyhow, the actual answer your question is that the "crop factor" *does* affect resolution, but not in the way you seem to be thinking. You have 10MP to work with on a 10MP camera - that much is unquestioned. But because those 10MP are crammed into a smaller area than a piece of film, that means you have to blow the image up more to get a given sized print than you would to create an image from a film camera. The effect is *exactly* like blowing up the image from a 35mm camera to 6x9", then cropping out the edges of the print to yield a 4x6" print. It creates the illusion of having shot with a longer lens, but on that 35mm camera, you'd have gotten better results actually shooting with the longer lens in the first place rather than shooting with the shorter lens and then blowing p the image bigger.

Thats NOT true. The sensor size is irrelevant for digital cameras only the number of MP matters.

If you have 2 cameras which are 10MP (3872X2592) and you want to print at 250 DPI without interpolation then both cameras will give you a print that is 3872/250 = 15.5" wide.

The only issue is that the smaller sensor has more photosites per inch, which means the lens has to be able to resolve more LPI as well. Dedicated APS lenses usually have higher resolution because they have a smaller image circle to cover. Digicam lenses have much higher resolution than 35mm lenses, but much smaller image circles again.

At ISO 100, there is no practical resolution difference between a Nikon D300 and D700.

[FunkyMonk]

Thats NOT true. The sensor size is irrelevant for digital cameras only the number of MP matters.

If you have 2 cameras which are 10MP (3872X2592) and you want to print at 250 DPI without interpolation then both cameras will give you a print that is 3872/250 = 15.5" wide.

The only issue is that the smaller sensor has more photosites per inch, which means the lens has to be able to resolve more LPI as well. Dedicated APS lenses usually have higher resolution because they have a smaller image circle to cover. Digicam lenses have much higher resolution than 35mm lenses, but much smaller image circles again.

At ISO 100, there is no practical resolution difference between a Nikon D300 and D700.

So from what I am gathering from the article the guy is saying that we lose resolution by using short focal lengths and gain resolution when using longer focal lengths. If that is the case then it must have something to do with the focal point of the 35mm lens onto the APS sensor. So then the question is at what length do we actually start gaining resolution? Am I understanding that correctly?

[Marc Sabatella]

Thats NOT true. The sensor size is irrelevant for digital cameras only the number of MP matters.
...
The only issue is that the smaller sensor has more photosites per inch, which means the lens has to be able to resolve more LPI as well.

And that was precisely my point, although worded differently. Lens resolution is definitely what I'm talking about. The smaller image means more magnification of the image, which pushes the lens' own resolution harder. If you have the same lens on 35mm and APS-C and then take advantage of the "crop factor" and print both images at 4x6", you'll have fewer lines of resolution in the APS-C image. In particular, you'll have the same number of lines of resolution you have in the central portion of the 35mm image, but stretched out to fill a larger area - meaning lower effective resolution.

[Marc Sabatella]

This is the part of the article that confused me.

And no wonder - I think you can see what I was saying regarding the language. To be honest, I am not entirely sure what he is saying either, but he is most definitely not claiming that 10MP is not 10MP. He's trying to calculate how many MP you'd need on a FF camera to give you the same effective resolution as 10MP does on APS-C.

Basically, you are making far too much of all this. If you don't use a 35mm camera, there is absolutely no reason to ever think about crop factors, except to make the mental translation when talking to others about their 35mm camera. That is, if someone else says they used a 135mm lens to shoot a given subject with their 35mm film camer, know that this means you'd want a 90mm. That's the only time you ever need to think about it.

[newarts]

Discussions of "resolution" often get confusing because we leave out the characteristics of the observation system.

Say we take photos with a K20D & K10D that look the same through the viewfinder (the lens is perfect, the sensors are the same size, and the K20D has 1.5 times more pixels than the K10D)

If you make uncropped 4x6" prints and view them unaided, the maximum separation between points on the print you'll resolve is what your eye can see (maybe 0.25mm); any higher resolution in the previous optical chain is immaterial & the prints will appear essentially identical.

Roughly speaking, an uncropped 4x6" print from a K20D and from a K10D will show about the same resolution because each display pixel contains more than one sensor pixel. In effect, the small print and our eyes comprise the weakest link in the optical chain (each print pixel will hold about 4 K10D pixels or 6 K20D pixels).

The minimum separation you can discern between points on the print can be referred back to the subject by the optical magnification. The optical magnification is close to Focal_length/Distance for non-macro work. So in this case the smallest separation you can discern on the print corresponds to 0.25mm*Distance/Focal_length on the subject.

But if you enlarge the images such that each display pixel contains just one sensor pixel, the K20D's print will be Sqrt(15/10)=> 22% larger on a side; the smallest separation on the original subject you'll be able to discern will be 22% smaller (ie. the resolution referred to the subject is 22% better.)

The short answer is that given a perfect lens, the K20D is capable of making bigger prints with a better resolution that a K10D.

I hope this clarified rather than further obscured things.
Dave

[*isteve]

And that was precisely my point, although worded differently. Lens resolution is definitely what I'm talking about. The smaller image means more magnification of the image, which pushes the lens' own resolution harder. If you have the same lens on 35mm and APS-C and then take advantage of the "crop factor" and print both images at 4x6", you'll have fewer lines of resolution in the APS-C image. In particular, you'll have the same number of lines of resolution you have in the central portion of the 35mm image, but stretched out to fill a larger area - meaning lower effective resolution.

No this is simply untrue. Its not a "smaller image". Its a 12MP image.

The resolution of the output is a combination of sensor and lens resolution, true, but only if the image resolution is limited entirely by the lens will the overall resolution be related directly to crop factor. You would need an APS sensor with a lot more than 15MP - or a truly appaling lens - for that to be the case.

In most cases, the lens is NOT the primary limiting factor in a DSLR. Also lens resoltion is not absolute, but simply diminishes in contrast the smaller the detail becomes. As a result, you may see a bit less micro contrast on a high res APSC sensor if you used a cheap lens but thats about it.

Other factors come into play also. For one thing, the system with the larger pixel pitch will need a stronger AA filter to avoid moire, and this will reduce its effective resolution. Similarly, lenses designed specifically for APS generally have higher resolution because they need less correction for spherical aberation. In the end, it all balances out, hence 12MP is pretty much 12MP whatever the format.

In the end with any half decent glass you can compare RAW files from 12MP APS and FF cameras till you are blue in the face but you will see no practical difference in the amount of detail captured.

Nikon D2X (12 MP) APSC format with 50mm prime
Horizontal resolution 2400/2800 (absolute and extinction)
Vertical resolution 2000/2500

Canon 5D (12 MP) FF format with 50mm prime
Horizontal resolution 2300/2500
Vertical resolution 2000/2500

[FunkyMonk]

And no wonder - I think you can see what I was saying regarding the language. To be honest, I am not entirely sure what he is saying either, but he is most definitely not claiming that 10MP is not 10MP. He's trying to calculate how many MP you'd need on a FF camera to give you the same effective resolution as 10MP does on APS-C.

Basically, you are making far too much of all this. If you don't use a 35mm camera, there is absolutely no reason to ever think about crop factors, except to make the mental translation when talking to others about their 35mm camera. That is, if someone else says they used a 135mm lens to shoot a given subject with their 35mm film camer, know that this means you'd want a 90mm. That's the only time you ever need to think about it.

The only reason I was concerned about it was that I was asked about this in a job interview today. He was a little concerned about a couple of the 35mm lens that I have and whether or not you lose resolution. I told him that it didn't affect the resolution because the sensor was still 10.1 megapixel but I wanted to be sure that I wasn't misinformed. Anyways I was hired for a Senior Photographer position with Wedding Media Group today. The owner is a Nikon guy but I have the K10D. So I ordered the Sigma 18-50mm f2.8 and the K20D today. So I'll be having some fun with the new stuff next week. Thanks for the info.