Why go full frame?

Hogloff wrote in post #13178725
Have you used a 5d2 or all your opinions on it also ignorant?

I stayed at a Holliday Inn Express last night....

Higgs Boson wrote in post #13178073
It's not anyone's job except the buyer to decide if they have enough money.

If everyone says full frame is better, maybe that's because there is scientific/empirical evidence for that..... People have to make choices for "next best" every day due to lack of funds. If you want to convince yourself that "next best" is actually best and anything more is overkill to justify your own personal decisions and abilities, that's fine, but that is opinion and is not motivated or based in proof, only feeling and should not be considered as anything but personal anectdote by potential buyers.

It's really more a question of providing enough valid (and hopefully reasonably objective) information so that people can decide if crop is really good enough for their needs, or whether they really need full frame. "Best" for one person may be "second best" for another. Some of the advantages of full frame will be meaningless to some people, and important to others. Similarly, crops have some advantages over full frame, and this may be important to some people, but not others. In other words, in both cases the "advantages/disadvanta​ges" are relative to the needs/wants/standards of the potential buyer. There's probably way too much opinion on both sides of the debate. A dispassionate listing of realistic, useable advantages/disadvantag​es of both sensor sizes should be of more use to someone asking if they should buy full frame than "buy a full frame, it blows everything else away" posts and posts endlessly discussing technical minutiae, such as how much DOF can stand on the head of a full frame pin versus a crop pin when the moon is blue. For the most part, that "dispassionate listing of realistic, useable advantages/disadvantag​es of both sensor sizes" was covered in the first page or so of this bloated thread. It then spun out of control.

Although I have a crop camera, I'm perfectly willing to state to anyone who asks that current generation full frame cameras can do certain things better/easier than crop cameras, at least under some shooting circumstances or for certain requirements. I'd be very happy to own a full frame camera, but I'm also happy with my crop camera. On the other hand, it would be appreciated if full frame advocates would admit that crop cameras can do certain things better/easier than, or at least as well as, full frame cameras, at least under some shooting circumstances or for certain requirements. One size doesn't fit all.

kcbrown wrote in post #13178817
The characteristic differences between full frame and crop are well known. That's not the problem.

The problem is that many who encourage the buyer to go full frame do so by saying things like "full frame image quality blows crop image quality out of the water!" and other nonsense like that.

But the fact of the matter is:

1. Full frame has about a stop shallower depth of field for a given aperture, distance to subject, and angle of view.
   
2. Full frame gives you about a stop better high ISO performance as long as you're willing to use a shallower depth of field to get it.
   
3. Full frame gives you sharper images out of the camera in some (perhaps many) conditions -- this depends greatly on the lens being used on each camera. The sensor, primarily through its resolution, places a hard upper bound on the amount of detail that can be recorded. The rest depends on the lens.
   
4. Full frame will give you creamier tones at low ISO, probably by about a stop (I don't know if anyone has actually tested this, but the physics of it all suggests that this will be the amount of difference).

Now, the question of the year is: how much of a difference to the buyer do those differences make? This is the point Lloyd (Picturecrazy) is making, and he is right on the money.


If "best" were always considered independent of cost and independent of the needs of the individual, and you full frame advocates who claim to want maximum image quality really meant what you say, then you wouldn't be shooting full frame, you'd be shooting medium format or large format. You'd be shooting with cameras that cost, at a minimum, $50K.

"Best" always has to account for the individual's situation. If you want to eliminate that and ask which camera gives you the "best" image quality, then that quite obviously has to go to large format. There can be no disagreement on that.

I dare say that none of you guys who spout the advantages of full frame because it has the "best" image quality are shooting large format. Why not? Simple: because it does not suit your actual needs. Which is to say: image quality is not really the thing you care about most. It is merely one of many attributes that you care about, not the least of which is cost (for if cost really weren't a consideration then you'd be willing to plunk $100K down on a medium format system for your landscapes and such, right?).


Now that we've dispensed with that, we can get back to the real discussion, which is: at what point do the advantages of full frame outweigh the disadvantages? Make no mistake, full frame, most especially as implemented by Canon in the 5D series, does have disadvantages relative to their crop offerings. All of those things have to be carefully considered when making a gear change if one is to acquire a setup that optimally meets his needs.

Well said.

stsva wrote in post #13178981
Well said.

No. Very well said.

David

AJSJones wrote in post #13165999
Perhaps you thought the original, possibly mundane-sounding, assertion "Crop will be enlarged more than FF to a particular print size" was trying to imply more than the simple geometry of the situation. It wasn't. It's because of all the other parameters that must be considered that the issue becomes complex. Sometimes the 1.6x might be trivial as a determining factor, other times it may be visible - but it's always there.

I fully agree that the laws of geometry are always in play. That's not really what I meant to dispute.

What I dispute is that they matter. And there is one simple reason for that:

In order to get the same shot between a crop camera and a full frame camera, the details recorded by the crop camera will be reduced in physical size by a factor of 1.6. So when you "enlarge" to print more for crop than for full frame, all you're really doing is compensating for the fact that the details have been reduced in size prior to being recorded by the sensor.

Which is to say, your 1mm line that you use as an example is being recorded from something in a scene in the outside world, the light of which is then passing through the lens and being resolved into the 1mm line on the sensor you refer to.

But in the case of the crop sensor, where the above sensor is a full frame sensor, the focal length used to capture the scene is 1.6x less and, therefore, the size of the line being projected onto the crop sensor is correspondingly thinner, by 1.6x. So instead of a 1mm line on the sensor, you now have a 0.625mm line.

That is the most meaningful comparison because it is what we, as photographers, are ultimately going to care about: capturing the same scene in the same way with two different cameras, and comparing the results.

So you're right in that the 1.6x crop image is going to be "magnified" more to achieve the same print size, but because the details being magnified are correspondingly smaller (when comparing like against like), the end result is essentially the same. And the nature of a digital capture is such that any analysis beyond that is going to depend greatly upon the characteristics of the lens and the characteristics of the sensors being compared.

The reason the geometry question matters for film, even when comparing a piece of 35mm film versus, say, a piece of APS-C film, is that generally the film being compared is the same and only the size differs. It is from that fact, and the fact that with film you're literally doing a geometric enlargement, that most people who have a great deal of film experience believe that the geometry matters when enlarging to print images from digital cameras with different formats.

Well, I'd say that the degree of "enlargement" only matters when it matters visually!

So, if you print an image from a crop sensor and the same framing with a full-frame sensor, at what point would it "matter"?

It's hard to say, although nobody would argue that an image taken with an 18MP P&S camera would "look worse" at a large print than an 7D image printed at the same large size -- after all, we all "know" that those little P&S sensors aren't fit for "serious" photography!

Anyway, something we always encounter in these discussions are people who are way too biased -- opinionated to the point where it's impossible to have a reasonable discussion! You see this with full-frame "advocates", you see it with crop "advocates". You see it with Raw "advocates", you see it with jpeg "advocates". You see it with Ford "advocates", you see it with Chevy "advocates", and the list goes on...

One can ignore people who are obviously so biased that there arguments are simply fallacies, the problem is that strongly opinionated people love to voice their arguments over and over again, hoping that the repitition will, what -- get people to believe them? And then the problem is that for every argument such as that people will be provoked to respond, and of those, some will be biased in the opposite direction.

So then we are not having a rational discussion about the relative strengths and weaknesses of the two sensor formats, instead it's two irrational sides butting heads!

Just sayin'!

But they are still being enlarged more! In post 173, the 7D and 5D2 DPReview tests were compared. The system* resolution of the 7D was about 100 LW/mm, while the 5D2 was about 80 (consistent with pixel pitch difference). Now, even with same framing, etc, whe the 7D image is printed to the same size as the one from the 5D2, say 16x24, the MTF 50 on the print will be showing 5 LW/mm for the 5D2image but for the 7D it will be 100/(16*1.6) i.e. a little less than 4 LW/mm. So the 5D2 produces a slightly higher detail print despite having a lower pixel density. I said earlier that while geometry is always a factor, it may or may not make a visble difference. In this example, the eye can often see the difference between 4 and 5 LW/mm, but at smaller print sizes, the geometry effect might be less visible.

*lineal resolution of the sensor (for sensor and lens combined -I have not studied their methodology in detail.)

jonneymendoza wrote in post #13163806
Also, IMO the 50mm lens on a crop is useless for my needs as its not wide enough to be used as a general walk-around lens and not long enough for shooting stuff in a distance

Funny but now i have a crop sensor the 50 have proved to be the most useful lens for my still life shots and certain close ups!

^^^
I agree with you, however Jonney, does kind of have a point in that a 50 on a crop is a little tight for a walkabout lens, but then you just get a 30/35mm.

yea it is. its like saying many use a 85mm on a FF as a walkabout lens when they dont and usually is the 0 or 35mm that gets used a lot more.

I just dont see the 50mm good for crop IMO. it jus doesnt cater for distance shooting or wide angle shooting.

i dunno what the 50mm is trying to be in the crop business as it falls betwene these two lines i.e not wide enough and not long enough

AJSJones wrote in post #13180494
But they are still being enlarged more! In post 173, the 7D and 5D2 DPReview tests were compared. The system* resolution of the 7D was about 100 LW/mm, while the 5D2 was about 80 (consistent with pixel pitch difference).

No, the 7D resolution is considerably higher than 100 LW/mm. Its absolute resolution is 2500 LPH, which over 14.9mm is 168 LW/mm. The 5D2's absolute resolution is 2800 LPH (that's giving it the benefit of the doubt, as dpreview shows its vertical resolution as 2700 LPH), which over 24mm is 117 LW/mm.


Sizing the 5D2's image to 16x24 gets you 6.9 LW/mm, while sizing the 7D's image to 16x24 gets you 6.2 LW/mm.

Now, the linear resolution ratio (ratio of linear pixels) of the 5D2 to the 7D is 1.0833, while the ratio of the actual measured resolutions is 1.12. So the larger photosites of the 5D2 in reality buys you only a 3.3% advantage in detail retention linearly, if the lens on the 7D is up to the task. Which is to say, that's the amount of advantage that having the larger photosites themselves buys you. You still gain the 8.3% linear advantage of greater sensor resolution (i.e., number of pixels).


Still think the difference would be easily visible if you use a good lens on the 7D (like they did when performing these tests), if the 5D2 didn't have a pixel resolution advantage?

I KNOW the difference is easily visible. So do MANY other people here who have or do use both and can see both side by side. If you have the ability to do so, please do, yourself, in Lightroom or similar, not an 800X600 file posted here.

The 7D looks like a Monet and the 5D looks like a photograph.

kcbrown wrote in post #13180895
Still think the difference would be easily visible if you use a good lens on the 7D (like they did when performing these tests), if the 5D2 didn't have a pixel resolution advantage?

I still think the 5D2 would be clearly sharper in terms on MTF figures even without a mega pixel advantage (the lens is clearly the limiting factor), as even with an extremely sharp lens at it's sharpest setting the 5Dii is producing 40ish% better MTF numbers with only a 15% mega pixel advantage.

However, I'm not saying there will be as much of an appreciable difference between the two in the real world in such a case, as the image will be very sharp in both cases and probably difficult to tell the two apart, but it would make a noticeable difference if your shooting a lens or using an aperture that isn't extremely sharp.

rhys216 wrote in post #13180682
^^^
I agree with you, however Jonney, does kind of have a point in that a 50 on a crop is a little tight for a walkabout lens, but then you just get a 30/35mm.

Well for a walkbout lens (nice phrase that!) I have a 19-35 (30-50 equiv)

jonneymendoza wrote in post #13180721
yea it is. its like saying many use a 85mm on a FF as a walkabout lens when they dont and usually is the 0 or 35mm that gets used a lot more.

I just dont see the 50mm good for crop IMO. it jus doesnt cater for distance shooting or wide angle shooting.

i dunno what the 50mm is trying to be in the crop business as it falls betwene these two lines i.e not wide enough and not long enough

Its fantastic for studio shots, table top kind of thing. My 100mm macro is out of range except when doing macro, my 19-35 wont let me close enough. The 50 nails it!