The Impact of Different Sensor Types on Your Photography

Gavjenks

TPF Noob!
Joined
May 9, 2013
Messages
2,976
Reaction score
588
Location
Iowa City, IA
Can others edit my Photos
Photos OK to edit
Note: I consider this a draft. I am not a sensor engineer or anything. I just have done a lot of research. So I could be wrong about some things here. Feel free to comment if so so I can update it.


The thing I see most often misunderstood when people give reviews of camera bodies, or are looking to make purchasing decisions, are the specifications of the sensor. In this post, I will describe the different types of numbers that describe a sensor, and what each of them means in practical terms for your photography, to help you make an informed decision about what camera type is best for you.

The three main numbers are:

1) Overall Sensor Size, often referred to as "crop factor." This is described as a single number usually (the length of one side), because sensors tend to share the same aspect ratio of 4:3, so you don't need to talk about both X and Y dimensions.
2) Number of Pixels
3) Pixel "Pitch," which for our purposes can be thought of as essentially just the size of one pixel in actual physical space.

These three numbers are not independent of one another. If you know two of these numbers, you can calculate the third:

(Pixel Pitch) = (Sensor Size along one side) / (Number of Pixels along the same side)

Here are a few made up examples, using an imaginary square sensor shape:

$5obpmJI.jpg

On the left is a 2x crop factor, 16 pixel sensor. Size = let's just say 1", pixels = 16, pixel pitch = 1/4"
In the middle is a full frame, 16 pixel sensor. Size = 2", pixels = 16, pixel pitch = 1/2"
On the left is a full frame, 64 pixel sensor. Size = 2", pixels = 64, pixel pitch = 1/4"



When you go to buy a camera, usually they advertise #1 and #2. Number of pixels is the megapixel count, and sensor size is advertised by various terms and letter codes and numbers like so:
"Large format" = a sensor (or piece of film) 4" x 5" or larger
"Medium format" = a sensor (or piece of film) larger than standard 24mm x 36mm but smaller then 4" x 5"
"Full frame" or "35mm" = 24mm x 36mm, or something very very close to that.
"Crop sensor" = any of a large number of different sensor sizes smaller than 24mm x 36mm. Almost every company has a different size crop sensor, sometimes multiple different sizes. These are labeled by a "Crop factor" like 1.6 or 1.3. 1.6, for example, would mean that a full frame sensor is 60% larger along one side than the crop sensor, or 2.56x larger in area. Etc.

So what do each of these factors mean for your photography?

Sensor Size

$z5DZsJy.jpg

The above diagram shows information relevant to sensor size alone as a variable. In this diagram and all others, I am using fake sensor sizes, and a square shape, just to be more generic and to not get bogged down in details about specific brands or anything like that. The "Crop sensor" I am using is an imaginary 2x crop factor sensor (1/2 total length of full frame along one side).

On the left, you see a crop sensor in a camera fitted with a lens specifically designed for a crop sensor (The equivalent of something like a Canon EF-S lens). Since crop sensors are physically smaller than full frame sensors, a lens made for crop sensors does not have to project as large of an image on the back of the camera. This leads us to our first practical difference between sensor sizes:

1) Lenses can be made for smaller sensors for less money and will weigh less than lenses compatible with full frame sensors (less glass!).

But there is also a drawback to such lenses, which is that they can't be used with a full frame sensor. This is because they don't project enough light to actually cover the sensor, so all your corners would be pitch black. And also because mirrors on crop cameras are smaller, so the lenses designed for them sometimes stick further back and would hit the mirrors of full frame cameras:

2) Lenses made for smaller sensors are incompatible with full frame cameras (so you'd have to buy new if you got a full frame camera), whereas lenses made for larger sensors usually are compatible with either sensor type.

Also, looking at the image on the right, it is clear to see that

3) Crop sensors will make the same lens appear to be a longer focal length (by a factor of the crop amount). So a 20mm lens would seem like a 40mm lens on a 2x crop sensor. It's not actually changing the optics, though. It's just a cropped subsection of the same optics. It will change what focal lengths are useful to you practically, though.

Note also the gradients I drew onto the lenses. As you move further from the center of almost any lens, the optical quality of the lens decreases. The edge of the projected image will be usually darker, blurrier, will have more color fringing and coma, etc. than the center of the projected image. This leads us to our 3rd practical difference:

4) If you use a full frame lens with a crop size sensor, your image will be sharper on the edges than if you use a lens matched to your sensor size (either), because you are only using the highest quality central portion of the glass, which has the fewest optical flaws. You pay for this, however, by shelling out money for (and carrying the weight of) glass that you're not actually using.

(The above is not true if you use a crop size lens with a crop sensor)


Finally, if you take a smaller sensor size, with a particular lens, the world will look more zoomed in. Let's say you compensate for this and walk further away from the subject in order to get the same field of view as with the larger sensor. Being further away from a subject leads to a larger depth of field (just like mountains on the horizon are always in focus together or not, even though they may be a mile separated from one another).

Alternatively, you could zoom out to compensate and achieve the same basic composition as the full frame / change focal length. This is a better way to match the composition, because it doesn't change your perspective, but still, this will ALSO increase depth of field, because wider lenses in general have broader depth of field.

Thus:

5) Effectively / in practical usage, a larger sensor will have shallower depth of field for the same composition, because you have to move the camera or change focal length to match composition, and both of these change DOF.

These are the main practical differences that result from sensor size ALONE. Sensor size also tends to correlate with the other aspects of a sensor (pitch and pixels), and thus indirectly is related to other practical differences, but we will get to those next. The above differences are the ones that necessarily result PURELY from the size of the sensor itself.
k9
Number of Pixels

Number of pixels has just one practical effect on your photography in and of itself:

6) The more pixels, the more you can blow up your print size relative to the distance a person views it from. And similarly, the more you can crop down your image after taking it and still have a usable, printable final product.

Fewer pixels means less of an ability to show an tiny line or an edge, and thus less sharpness at the same print size. Thus, all other things equal (which they usually aren't), more pixels = more ability to show sharpness and detail. This means you can blow up the same image larger. Or you can use a smaller sub-section of your image and blow that up to normal size (same concept). This has the effect of giving you more leeway in adjusting your composition in post processing, and it also allows your camera to act like it has a longer lens on it, because when you crop a portion of your image and zoom it larger, it is an approximation of having had a longer lens (provided that you still have enough sharpness to print the zoomed in section, which more pixels tend to provide).

Anything over 10-12 megapixels is going to be undetectably different to the human eye at regular print sizes, so more megapixels does NOT mean better sharpness or image quality if you do everything right and are not intending to try to get more reach out of your lens by cropping. What more megapixels do for you, again, has more to do with giving you the flexibility to change things more after you've taken the shot (cropping and zooming to simulate longer focal length, or blowing up huge prints and such).

Also note that the whole "crop and zoom" thing will not work if your lens is not sharp enough to resolve that extra detail. Having 30 billion pixels is useless if you have a soft lens that isn't actually delivering different light to each pixel.

Pixel Pitch

Again, remember that pixel pitch is mathematically entirely determined by (sensor length on one side) / (number of pixels per side), so none of these three things are completely independent of one another. But here, I will describe the practical effect that larger versus smaller pixel pitch has, if all other variables are held constant.

$z5GzKDq.jpg

Here I drew a picture of two individual sensor units (a pixel would composite across several of these, in ways not worth explaining here), one from a sensor with large pixel pitch on the left, and one with a small pixel pitch on the right.

Note that these do not necessarily correspond to full frame or crop size sensors! It is possible to have small or large pixel pitch on either type of sensor. Usually, in the camera industry, full frames HAPPEN to have larger pixel pitches, but they don't technically HAVE to.

Anyway, the above image shows a situation with bright light. During the same exposure conditions and shutter speed, etc., the larger sensor might gather let's say 7 photons, and the smaller only 3. But both sensors generate about the same amount of noise from their electronic components and the process of converting from analog to digital (represented by a red line here). So the larger sensor got a signal to noise ratio o 7:1, while the smaller got a ratio of 3:1. The higher your signal to noise, the less noisy your final image will look. Basically, you expect some average number of "photons" of noise in any one sensor (in this case 1). But it may sometimes be 0 (darker than it should be), or 2 (brighter than it should be). These deviations are random and we don't know for sure when they happen. Thus leading to noise.

The larger sensor will have the more accurate color and lightness, because it is less "polluted" by random noise. A more overwhelming proportion of actual light makes the statistical impact of the noise smaller. Thus:

7) The larger the pixel pitch, the less noisy your images will be under the exact same exposure conditions.

So if you have a larger sensor size, in other words, you can crank up the ISO on your camera higher without suffering from noticeable noise. Higher ISO means you wait for fewer photons, and then just multiply what you got to amplify the signal to normal levels. But of course, you also amplify the noise! So having a higher signal:noise ratio means you can amplify further without suffering from as much noise. In even more direct practical terms:

7.5) (in other words,) At the same number of megapixels, a full frame camera would have about 2 stops more effective speed from ISO than a 2x crop factor camera

(2x crop @ same megapixels = 1/4 the sensor size, and thus 1/4th the signal to noise ratio and 4x more amplification pollution. 4x = 2 stops)

And finally,

8) Smaller pixel pitch sensors have better resolution than larger pixel pitch sensors, within the angle of view they cover, which means they get more out of the same telephoto lens.

This is because smaller pixel pitch crams more pixels of resolution into a smaller visual angle. An 18.1 megapixel 2x crop frame sensor has 18.1 megapixels of resolution in a narrow slice of the image circle, while an 18.1 megapixel full frame sensor would only have 1/4 as many pixels in that same angular slice of the image circle. So if you tried to crop and zoom just that portion of the full frame, you'd have a blurrier picture than the uncropped image from the smaller sensor.

$Zcs8enB.jpg


Careful!

I cannot stress enough that although these 3 factors are USUALLY CORRELATED, they do not always HAVE to be. Take a look at that same image I posted earlier:
$5obpmJI.jpg

The sensor on the right is a full frame sensor, but it actually has the SAME pixel pitch as the crop sensor on the left.

In this (made up) situation, the full frame sensor on the right would not have any better ISO performance than the crop sensor on the left.

A zoomed up picture of a blank wall would look something like this, under similar ISO and exposure across the three sensors, and prior to any software de-noising:

$DftydSw.jpg

This is why, for example, Canon's flagship professional 1DX camera only has 18.1 megapixels. If it had 46 megapixels instead, then it would not actually end up with any better ISO performance than, say, an entry level crop sensor rebel T2i, because the pixel pitch would be identical. The grain would be smaller, yes, which might matter in some cases, but basically, just as bad of noise.

if it had 46 megapixels, it would be much better able to adjust composition after a shot or crop and zoom (giving you the same telephoto performance of the crop sensor without needing to actually put on a longer lens!). But I guess Canon decided that high end pros would have more use for ISO performance than for flexibility in post processing.

(Also, better ISO performance saves you money on telephoto lenses anyway, because you only need to buy an f/4 lens instead of an f/2 lens for the same quality, if your ISO performance is 2x better. So you aren't even really compromising telephoto ability with the fewer pixels, in a way.)



That's all folks! At least all I can think of. If I missed something that is not covered above or made a grave error, please let me know.
 
Last edited:

amolitor

TPF Noob!
Joined
May 18, 2012
Messages
6,320
Reaction score
2,131
Location
Virginia
Can others edit my Photos
Photos OK to edit
The discussion of pixel pitch and noise is correct as far as it goes. But ignores the biggest practical factor which is sensor generation. A newer sensor will in general have less noise than an older generation one. This produces some very confusing results, since pitch tends to go down over generations, but so does noise.
 
OP
Gavjenks

Gavjenks

TPF Noob!
Joined
May 9, 2013
Messages
2,976
Reaction score
588
Location
Iowa City, IA
Can others edit my Photos
Photos OK to edit
Well sure, but that's not really a difference between sensors per se, because both the crop and the full frames will benefit equally from later generation technology.

Dunno... I will ponder this / see if I can find data anywhere on how much it actually has improved for that reason.
 

Derrel

Mr. Rain Cloud
Joined
Jul 23, 2009
Messages
48,227
Reaction score
18,929
Location
USA
Website
www.pbase.com
Can others edit my Photos
Photos OK to edit
Re, your point #4:4) If you use a full frame lens with a crop size sensor, your image will be sharper on the edges than if you use a lens matched to your sensor size (either), because you are only using the highest quality central portion of the glass, which has the fewest optical flaws. You pay for this, however, by shelling out money for (and carrying the weight of) glass that you're not actually using.

I do not think that has been proven to be true, but it has been repeated over and over and over, but from what I have seen, the highest resolution figures for various lenses that have been tested have always been recorded when those lenses have been shot and tested on full-frame sensors, and not on crop-sensor cameras.

Not sure what you mean by the rather vague term "sharper", but I am in my comment above assuming that by "sharper", you meant to say "will yield higher resolution" or something like that. "Sharpness" really does not have a very good definition. Resolution, or contrast can be easily measured and quantified, while "sharpness" is a rather vague term that can be affected by many different image characteristics, like noise, grain, acutance,resolution, contrast, and so on.

Again, I do not think that the "better center/sweet spot" argument is correct; I think it is actually an old folk tale, well-conceived in theory, but not really in line with what actually happens when the same lenses are tested on both APS-C and FF sensors...the FF sensor + FX lens combo always out-resolves the FX lens + APS-C sensor pairing, all across the frame.
 
OP
Gavjenks

Gavjenks

TPF Noob!
Joined
May 9, 2013
Messages
2,976
Reaction score
588
Location
Iowa City, IA
Can others edit my Photos
Photos OK to edit
Re, your point #4:4) If you use a full frame lens with a crop size sensor, your image will be sharper on the edges than if you use a lens matched to your sensor size (either), because you are only using the highest quality central portion of the glass, which has the fewest optical flaws. You pay for this, however, by shelling out money for (and carrying the weight of) glass that you're not actually using.

I do not think that has been proven to be true, but it has been repeated over and over and over, but from what I have seen, the highest resolution figures for various lenses that have been tested have always been recorded when those lenses have been shot and tested on full-frame sensors, and not on crop-sensor cameras.

Not sure what you mean by the rather vague term "sharper", but I am in my comment above assuming that by "sharper", you meant to say "will yield higher resolution" or something like that. "Sharpness" really does not have a very good definition. Resolution, or contrast can be easily measured and quantified, while "sharpness" is a rather vague term that can be affected by many different image characteristics, like noise, grain, acutance,resolution, contrast, and so on.

Again, I do not think that the "better center/sweet spot" argument is correct; I think it is actually an old folk tale, well-conceived in theory, but not really in line with what actually happens when the same lenses are tested on both APS-C and FF sensors...the FF sensor + FX lens combo always out-resolves the FX lens + APS-C sensor pairing, all across the frame.

You are correct that I am operating from the realm of theory with point #4.

However, the only way to properly test this would be to have:
1) Two cameras with equal pixel pitch, one full frame and one crop sensor. Pitch must be equal, because obviously, different pixel density will have an effect on resolving power, and so will light collection, blah blah. None of that is what this is about. Those are all confounding factors, and if you don't control them out, they could easily cause the two sensors to cancel out and look indistinguishable. But it still wouldn't disprove this theory if so!
2) Same lens should be used.
3) The shots should be lined up when taken so that the CORNERS hit the same part of the image together, since the corners are what this theory is about, not the middle (if you compare middles, they should both be equally sharp...). Choose an item with good texture, etc. for discerning resolving power and align it to the corner of each shot. Or a chart.
4) Then crop out a small portion of each corner with the same number of pixels each, and directly compare the crops.

I've never seen such a comparison. I would do it myself, but I don't own two cameras with equal pixel pitch and different size sensors. They do exist though! A perfect pair would be the Nikon D800 at 4.9 micron pitch and full frame vs. the Nikon D7000 at 4.8 micron pitch and crop frame. Both pretty modern, and both can test the same lens together. Slap a nice average piece of glass on both, and test as specified above.



Until I see a valid test proving one way or the other empirically, I am personally of the opinion that the possibility with the strongest theoretical basis should be assumed. And the position with the strongest theoretical basis is, I believe, that there IS a difference.
 
Last edited:
OP
Gavjenks

Gavjenks

TPF Noob!
Joined
May 9, 2013
Messages
2,976
Reaction score
588
Location
Iowa City, IA
Can others edit my Photos
Photos OK to edit
To be absolutely clear about my proposed test, a graphical depiction of it:
$KG9OnHR.jpg
Both crops will end up with same pixel pitch, and same number of pixels, making the ONLY variable changing the size of the sensor itself, and thus the position of the monkey in the optical projection. I.e. the path that the sensed photons took through the lens (through the "sweet spot" or not)

I predict that the Monkey in the crop on the right will be noticeably more free of lens aberrations of various sorts than the monkey in the crop on the left. And it should show higher resolving power.

A lens of average or lower quality should be used for this test, in order to guarantee the detection of any differences, should they exist. Also, it should probably be shot wide open. If super high quality glass is used with almost no aberrations anywhere, it sort of undermines the whole idea.

After all, the whole point of this "myth" is that you can get more quality out of the glass with a crop sensor. But if it's nearly perfect to begin with end to end, then there's no more quality to squeeze out in the first place.
 
Last edited:

Judobreaker

TPF Supporters
Supporting Member
Joined
Apr 4, 2011
Messages
969
Reaction score
228
Location
Holland
Website
www.rvanantwerpen.com
Can others edit my Photos
Photos OK to edit
In your test you're filling the frame with the same scene (a scene with the monkey bottom left).
Yes, the monkey on the right will possibly be better quality than the one on the left, but that is not because of your myth.

You're stating that a crop sensor will get more quality out of the glass than a crop sensor.
I disagree.
Because you're using a crop sensor you're shooting the monkey more towards the center of the glass. Everybody knows that the quality of a lens gets better towards the center.
However, that does not mean the crop sensor gets better quality out of the glass. The full frame sensor gets the same quality towards the center, it just also gets the stuff that's more towards the edges of the glass.
Say you'd walk backwards with a full frame sensor and shoot the scene so that it is the same size on the sensor as it is on the crop frame sensor, and then you crop it.
Now you'll have the same scene through the same parts of the glass. Now you can compare one sensor to another because you've ruled out all other variables.

What I'm trying to say is that glass quality is a constant. No matter the sensor, a lens is of a certain quality and it will not change because of the sensor.
I think the only thing on a sensor determining how much quality of the glass you will see is pixel pitch. The smaller your pixel pitch, the more you'll see the glass quality.
If you use bad glass on a sensor with a large pixel pitch you will not notice it as much as on a sensor with a small pixel pitch, as the one with the small pitch will be able to show how unsharp the lens actually is.

Now in your example you're using a D800 and a D7000.
The pixel pitch of a D7000 is smaller than the D800 so basically it is probably true you can get more quality out of the glass with a D7000. ^^
 

Derrel

Mr. Rain Cloud
Joined
Jul 23, 2009
Messages
48,227
Reaction score
18,929
Location
USA
Website
www.pbase.com
Can others edit my Photos
Photos OK to edit
Look at the many,many lens testing and camera-testing sites... you will see that your theory is incorrect. The better performance is obtained with the larger-sensor cameras. Every time. This issue has come up elesewhere. Theory is nice, but sometimes pipedreams prove to be just that.

Your assesment of theoretical validity sounds good, but it is seriously flawed. It's a nice wive's tale, but it ignores a number of factors. The HIGEST resolution figures come from using a lens on a larger sensor. Using a lens designed for 24x36mm capture on a smaller capture medium does not somehow automatically make "sharper" images. In fact, the opposite is true. You are free to speculate as to why.

Why don't you head on over to the big testing sites, then get back to us with a new Point #4. One based on reality. Not theory.
 
OP
Gavjenks

Gavjenks

TPF Noob!
Joined
May 9, 2013
Messages
2,976
Reaction score
588
Location
Iowa City, IA
Can others edit my Photos
Photos OK to edit
What I'm trying to say is that glass quality is a constant. No matter the sensor, a lens is of a certain quality and it will not change because of the sensor.
Yeah I know... that's the whole point.

All other things equal, a smaller sensor will suffer less lens aberration, because it physically cannot possibly utilize the "bad" parts of the lens. Whereas a larger sensor will ALWAYS suffer some degree of lower quality in the outer half of its pixels, no matter what.

The fact that the glass quality is a constant is precisely why this remains true no matter what you try and do to stop it. Zooming, moving around, whatever, doesn't matter.

The only thing you can do to get the same lens optical quality across the whole image is to just crop to the middle of the image on your full frame. But if you do that, then there was no point in having a full frame sensor in the first place. And they are more expensive. So that's less optimal.



Yes, when you look at more typical comparisons, other factors start mattering, almost certainly more than this. But that doesn't make this not a factor. It's still always a factor (unless proven otherwise). Just maybe not the most important one.
The pixel pitch of a D7000 is smaller than the D800
It's 2% smaller pitch... that doesn't matter at all. I chose them specifically because they have almost completely identical pitches.




Look at the many,many lens testing and camera-testing sites... you will see that your theory is incorrect. The better performance is obtained with the larger-sensor cameras. Every time. This issue has come up elesewhere. Theory is nice, but sometimes pipedreams prove to be just that.
Where are you getting your information from?

Can you link to a test that actually properly tests this, controlling for pitch and pixel # like above? I searched for this for about an hour (including testing sites) and did not find anything that even remotely approached a well controlled test, although several people tried (badly).

14 different poorly controlled tests don't prove anything more than zero tests do...

I of course could simply have not found the good one(s) though. If you don't immediately have a link for me, then I will take the time to go post on a testing site though and see if anyone there can point me in the right direction, or do the test for me.
 

Derrel

Mr. Rain Cloud
Joined
Jul 23, 2009
Messages
48,227
Reaction score
18,929
Location
USA
Website
www.pbase.com
Can others edit my Photos
Photos OK to edit
You stated that you consider your work a draft. And yet, you defend it as if it's your thesis.
 
OP
Gavjenks

Gavjenks

TPF Noob!
Joined
May 9, 2013
Messages
2,976
Reaction score
588
Location
Iowa City, IA
Can others edit my Photos
Photos OK to edit
No, I just see no reason to change information based on speculation. My goal here is to provide a reference with the most accurate possible information. I have no vested interest in either answer being correct.

So if I can find well-controlled data that proves anything, then data always wins. If you have some, please link it. If not, I will seek it out when I have time. "I totally read somewhere once that this isn't true... I think..." does not = well controlled data.

And if I have NO well-controlled data yet, then theory wins. And theory alone obviously predicts that the EXACT same sized sensors looking at the EXACT same object at the EXACT same distance would be sharper when doing so in the middle of a piece of glass than on the edges.

If that turns out not to be the case, it would be an extraordinary finding. Doesn't mean it isn't true, but extraordinary claims require extraordinary evidence, as they say.
 

Judobreaker

TPF Supporters
Supporting Member
Joined
Apr 4, 2011
Messages
969
Reaction score
228
Location
Holland
Website
www.rvanantwerpen.com
Can others edit my Photos
Photos OK to edit
...

The only thing you can do to get the same lens optical quality across the whole image is to just crop to the middle of the image on your full frame. But if you do that, then there was no point in having a full frame sensor in the first place. And they are more expensive. So that's less optimal.

...

If you do the test your way you've got to be fair and look at the final display of the image.
Say we shoot the scene while filling the frame... Sure the monkey might be of less quality on the D800 when looking closely at 100%.
But now print the image, or resize it for web or something like that.
The D800 will actually have a lot more detail than the D7000 so in this case the D800 is definitely the winner.



...

It's 2% smaller pitch... that doesn't matter at all. I chose them specifically because they have almost completely identical pitches.

...

I know you won't be able to actually tell the difference, but this is purely theoretical. This is a theoretical discussion isn't it? :p
 
OP
Gavjenks

Gavjenks

TPF Noob!
Joined
May 9, 2013
Messages
2,976
Reaction score
588
Location
Iowa City, IA
Can others edit my Photos
Photos OK to edit
But now print the image, or resize it for web or something like that.
The D800 will actually have a lot more detail than the D7000 so in this case the D800 is definitely the winner.
Printing the full frame image at the same size is doing nothing but resampling it smaller. I.e., throwing out data (and not intelligently, either, just 1 out of every 2 pixels in order). It's not physically possible to gain detail from indiscriminately throwing away 1 out of every 2 pixels.

Nor is it possible to overtake an image that started out with fewer lens distortions (color fringing, coma, or spherical aberration) simply from throwing away 1 out of every 2 pixels from the lower quality 100% image.

It may be possible to hide it well enough so that you can't tell with the naked eye. But that should be the best you can hope for: equality. There shouldn't be any possible way for the D800 to get BETTER results.

And whether you can even achieve equality or not is likely going to depend on just how bad that lens is. I would guess that you can probably achieve equality with a nice mid to high range Nikon lens. But quite possibly not with a trashy Rokinon lens.

(Also for the record, I never claimed higher detail. Only fewer lens distortions)
 

amolitor

TPF Noob!
Joined
May 18, 2012
Messages
6,320
Reaction score
2,131
Location
Virginia
Can others edit my Photos
Photos OK to edit
So you've got a couple areas where you want to stick to a sort pure hypothetical world, noise and center sharpness, where the pure world of theory gives answers that are different from reality.

On the one hand I think limiting the scope of discussion is valuable, trim away stuff so that we can really dig in a bit without having to talk about everything in the universe.

On the other hand, reality isn't really complying.

The trouble with these things, and this sort of discussion has been around for ages (which film is sharpest?) is that the actual answer is always about the whole system. The only way to get meaningful answers is to stick the lens on a camera and measure that. You can say meaningful things in general terms, but they have to be somewhat vague, and qualified as tendencies, not absolutes. Your essay is fine as a "look, these are general tendencies and when you're comparing apples to apples, they tend to be true, but be careful because you are usually comparing apples to donkeys or something" piece. As such it's quite nice and complete.

The section of DoF seems abrupt, comparatively, but it's a hard one to summarize in a couple of paragraphs.
 
OP
Gavjenks

Gavjenks

TPF Noob!
Joined
May 9, 2013
Messages
2,976
Reaction score
588
Location
Iowa City, IA
Can others edit my Photos
Photos OK to edit
Okay, I made the depth of field explanation a lot more.. erm, in depth.

Anyway, as for reality versus theory-land: I'm not sure I agree that there's a disconnect. Yes, in reality, the whole system is always what matters, but I believe that in most cases, you can determine the performance of the system by essentially just adding up the sum of its isolated, theory-land parts.

Which parts you add up / take into consideration depend on what you want to do with your camera at the end of the day, though, and will be different for different people. Which is why it is useful to isolate the parts in the first place, so that they can be reassembled as needed for custom purposes.

For instance, let's say i want to shoot sports. Well, let's add up the various factors above that are relevant to that:
1) Crop lenses are cheaper.
2) Even if I can't find a crop lens of high enough quality for my needs, the higher resolving power of most (not all) crop sensors per inch of sensor means I will get more out of full size telephotos as well, which can save me a lot of $$, since lenses get very very expensive as you go from 200mm to 400mm or onward, and I may often only by using a small-medium central portion of the frame anyway.
3) The effectively greater DOF of a crop frame will help give a little leeway for getting focus on active subjects.
4) Crop sensors are cheaper.
5) A crop sensor, however, is slower than a full frame with larger pitch, which will make it harder to freeze motion if I ever have poor lighting at all.

Useful Conclusions:
If I have infinite money, full frame is the winner, because #1-4 are mainly about saving money, so they don't matter to me, and I'll just go for the extra convenience of ISO speed + typically higher FPS in most real world full frames, etc.
If I am on a budget and shoot sports with plenty of light, crop sensor seems to be a clear winner for me, however (especially with the 8 FPS 7D!).
If I am on a budget and shoot sports in low-ish light, then it's sort of a wash / sensor doesn't tell you much.





Yes, trying out the real thing is a much more effective method of making a decision than the process I outline here, but the whole point of this is as a reference guide for people considering buying different camera bodies in the first place. In other words, they don't HAVE the various equipment in hand to try each thing out in person and compare. But by adding up the pieces, they can deduce the best choice for them instead.
 

Most reactions

ClickASnap

New Topics

Top