Full Frame advantage real?

[gdanmitchell]

Not sure, but the OP included the following in his post: "I was just reading an article on luminous landscape about FF and it said something about people that go for FF get it and those that don't, don't. I don't get it."

Thus making the reference relevant.

Dan

In general (though not every time) you can trust what you read on
Luminous Landscape a lot more than what you read from Ken Rockwell...

what did Luminous Landscape say about full frame advantage?
(link?)

Just sayin'

Dan

--
---
G Dan Mitchell
SF Bay Area
http://www.gdanmitchell.com/

--
---
G Dan Mitchell
SF Bay Area
http://www.gdanmitchell.com/

 

[gdanmitchell]

JoeMama wrote:

"n any case, the superior sharpness of FF will not be apparent if the glass is sharp enough to resolve the pixels of the higher density sensor. In other words, at the f/9 those test pics were taken at, you will see no sharpness advantage for FF."

Not exactly.

It s true that any "imperfections" produced by the lens are the same measured dimensions. However, because the sensor is larger on the FF body, that dimension covers a smaller area of the image on a percentage basis. In other words, a given distortion is "smaller" on FF.

In addition, if the FF camera has the same number of photosites, it is less likely that certain types of distortion (e.g. - out of focus blur) will spill to adjacent pixels.

All of this is particularly true (and much less theoretical) if the photographer frames the subject so that it "fills the frame" the same way on crop and FF and uses the best available lenses on both. The FF image will be "better" as a print.

Of course, one has to admit that this "betterness" may not be apparent in all situations - for example you probably won't see it in letter size prints or in web images posted as jpgs. You will be able to see it at some point in larger prints.

While I'm writing (and not necessarily in reference to JoeMama's post) there is a lot of hot air being blown regarding DOF issues. People speak as if DOF is a binary - e.g. subjects within the DOF zone are "in focus" and those out of the DOF zone are "out of focus." It doesn't work that way.

There is a plane (more or less) of optimum focus at any aperture. Everything outside of that plane is out of focus. The rate at which focus degrades is greater with larger apertures, and less with smaller apertures - but stuff outside the zone is still "out of focus." What DOF scales suggest is how far in front/behind the zone of optimal focus an object can be and be acceptably out of focus such that a some assumed enlargement size most viewers under certain conditions will not notice the OOF. Whew.

You cannot compare formats in a vacuum. It is critical to remember that the "out of focusness" of a crop image will be enlarged more to achieve a given print size than the "out of focusness" of a full frame image, since the latter starts larger. Another way to think of it is that the absolute dimensions of an OOF area on a crop sensor will be a larger percentage of the overall image than on a larger sensor.

Frankly, when I'm out making photographs I essentially never do DOF calculations. Do people actually do this!? I look in the viewfinder. I compose. I select the right lens. I do manual DOF preview. I take multiple exposures at different apertures if necessary.

Finally, for certain kinds of photography the smaller apertures that one can use on FF have some other benefits besides DOF control. Issues that affect virtually all lenses to some extent, such as vignetting and corner softness, can be virtually eliminated at small apertures such as f/11 and f/16. If you do landscape and similar subjects, you are perhaps already familiar with this.

Take care,

Dan

(Who apologizes in advance for not reading every message in the whole thread before posting.)
--
---
G Dan Mitchell
SF Bay Area
http://www.gdanmitchell.com/

 

[Radio Denver]

Well I prefer full frame bodies and I get it. My crop sensor bodies sit and never get used. One reason I think is ....oh, never mind. If you don't get it, no use in trying to explain it.

All I can say is that I've personally never met anyone that went to a full frame body and then decided it sucked and went back to a crop sensor body because it gave them better results.

Not sure, but the OP included the following in his post: "I was just
reading an article on luminous landscape about FF and it said
something about people that go for FF get it and those that don't,
don't. I don't get it."

--
Visit me at

http://www.have-camera-will-travel.com

 

[Dan62326]

Well I prefer full frame bodies and I get it. My crop sensor bodies
sit and never get used.

Which is because FF is what's best for your shooting style.

One reason I think is ....oh, never mind.
If you don't get it, no use in trying to explain it.

I'm trying to understand it by intelligently discussing it with other members.

All I can say is that I've personally never met anyone that went to a
full frame body and then decided it sucked and went back to a crop
sensor body because it gave them better results.

I've never met anyone that went to a crop and then decided it sucked and went back to a FF sensor body because it gave them better results.

 

[Dan62326]

It s true that any "imperfections" produced by the lens are the same
measured dimensions. However, because the sensor is larger on the FF
body, that dimension covers a smaller area of the image on a
percentage basis. In other words, a given distortion is "smaller" on
FF.

So does this explain why you can use a lens that is 1.6 times less sharp than you could on a 1.6 crop sensor?

In addition, if the FF camera has the same number of photosites, it
is less likely that certain types of distortion (e.g. - out of focus
blur) will spill to adjacent pixels.

I'm re-reading that and I'm having trouble understanding what you said.

All of this is particularly true (and much less theoretical) if the
photographer frames the subject so that it "fills the frame" the same
way on crop and FF and uses the best available lenses on both. The FF
image will be "better" as a print.

If the 1.6 crop lens was 1.6 times sharper than the FF lens, would they be equal?

It is critical to remember
that the "out of focusness" of a crop image will be enlarged more to
achieve a given print size than the "out of focusness" of a full
frame image, since the latter starts larger.

Does this mean that the background blur of the image from the crop sensor will look the same as the background blur from the FF image (when the image is printed at the same size)?

Another way to think of
it is that the absolute dimensions of an OOF area on a crop sensor
will be a larger percentage of the overall image than on a larger
sensor.

You lost me there.

 

[Dan62326]

has no clue how to test a camera or lens
properly, and simply doesn't even know how to use the gear. His tests
are hysterically bad!

Isn't that how most people would compare different cameras? I doubt the average person is going to process RAW files optimally using the perfect amount of noise reduction and sharpening. Although I would say that he should have provided results of other cameras with varying degrees of high ISO noise reduction.

But often, he's doing is basically what anyone would do if they brought their memory card to the store to shoot a few test images.

 

[Dan62326]

a 35L is a 35mm lens. Not a 50mm something.
a 135L remains a great portrait lens.

That doesn't say anything...at all...to me. Besides, the 85 1.2 is better. You want 35mm...just get a wider lens. If the desired lens does not exist, then FF is for you.

same is true for the 17-40, 24-105L ( there is mo such lens for
cropped sensors now and this lens alone is worth FF )

Isn't that what the 10-22 is for?

24-105...true..at least yet. Sony has a 16-80 though so we know it's possible.

And that´s all without taking shallow DOF, noise or per pixel
sharpness into account.

So far all this talk about the above seems to be myths and misunderstandings.

 

[vadims]

subj

 

[Dan62326]

I'm a firm believer that the differences in operation
between cameras matters far more for most people than the differences
in IQ.

I totally agree....I think. Would you place high ISO noise reduction under operation?

After studying samples from the D300 at ISO3200 and 6400, I've come to the conclusion that I can use shots at those settings.

The 5D has no ISO6400 setting. Manually pushing ISO3200 shots is too much trouble. And the 5D doesn't have different high ISO noise reduction settings to chose from right? I don't like adding an extra step to my workflow, but would do it if it weren't too complicated. I like being able to chose how my picture looks in-camera and then shoot JPG.

Speaking for myself, I'm a minimalist, I guess, in search of maximum
IQ for the money I can afford to spend.

For me, IQ is of a lower priority than features that enable me to get the shot...like the AF system - which is why the 51-point AF of the D300 interests me, subjective quality of the high ISO noise reduction system (enabling me to shoot at higher ISO's without having to deal with 3rd party noise reduction software and such), and sometimes auto white balance (to which I have to declare Fujifilm as the clear winner by a large margin).

From all I've read about
Nikon, the only operational advantages that system has that would
serve me better than Canon are on-demand grid lines and Auto ISO (I
really, really, really want Auto-ISO). And those two operational
advantages of Nikon mean all of squat to me compared to what the 5D
offers me in terms of shallow DOF and high ISO performance. In
addition to that, Canon has three indispensible lenses to me: 24 /
1.4L, 50 / 1.2L, and 100 / 2 (the 100 / 2 has USM, which makes it far
more useful to me than Nikon's counterpart).

My indispensable lens is the 18-200mm VR. And since there is just slightly over 1-stop of difference in noise, with fancy noise reduction software (whether in-camera or 3rd party) it could narrow the gap (depending on what you shoot of course - and personal preference). And as you might notice in the other messages, I'm still trying to get a handle of this difference in DOF, but right now, to me it's more a luxury than a requirement.

That said, if you could spring for a D3, I'd really push that option.

Not at this time =) It's too expensive.

Seems like a killer camera.

Yeah, and to me the usability at the higher ISO's is too cool.

And, to counter my indispensible Canon
lenses, Nikon has a 14-24 / 2.8, which, by all accounts sets a new
standard in wide lenses by a large margin, a new 24-70 / 2.8 which
may lay waste to Canon's 24-70 / 2.8L in the same manner, and a 100 /
2.8 VR macro that is, from what I've heard, a killer lens.

The only lens that really interests me on the Canon side is the 24-105 F4 IS. But I have shot a wedding with a 24-70mm F2.8 on a crop body and I often found myself wanting to zoom-in more. I didn't have trouble at the wide end even though it was only 24mm.

There was a lens that interested me for a while. The Tamron 28-105mm F2.8
But I think there were some sharpness/contrast issues with it.

Oh well, I guess what I'm trying to think of is a way to not have to carry two cameras around when I want F2.8 zooms but I guess that will never happen. That's why my favorite lens right now is the 18-200mm and the VR enables me to hand-hold with lower shutter speeds than I would even with an F2.8 zoom (without IS/VR).

Then again, there's
the upcoming Sony FF DSLR.

Really? Cool...looking forward to the specs

If its sensor performs as well as the
Nion D3's (and there's no reason to think it won't),

Let's hope they don't apply noise reduction to the RAW files the way the A700 does.

and has
in-camera IS, then I would even more strongly consider that system
instead.

I really wonder how effective in-camera IS would be with a 300mm lens. Wouldn't the sensor have to move quite a bit? I'll have to see.

They already have a 35 / 1.4 and 135 / 1.8, so just a few
more wide to mid fast primes and I get IS on them to boot. Neither
Canon nor Nikon seem to think that fast primes need IS unless you're
at or above 200mm (200 / 2 IS and 200 / 2 VR).

And for me, I'd have to add that having a stabilized 85mm F1.4 would be very nice.

One nice Canon crop zoom that caught my eye was the 17-55mm F2.8 IS.

And you're right. I wish they would add IS/VR versions of all their lenses.

 

[joe mama]

It s true that any "imperfections" produced by the lens are the same
measured dimensions. However, because the sensor is larger on the FF
body, that dimension covers a smaller area of the image on a
percentage basis. In other words, a given distortion is "smaller" on
FF.

So does this explain why you can use a lens that is 1.6 times less
sharp than you could on a 1.6 crop sensor?

Yes, but there's a much simpler way of saying it. Read the essay yet? 'Cause it's answered in it:

http://www.josephjamesphotography.com/equivalence/#myths

Go down to Myth #5: Larger sensors require sharper glass and have softer edges.

I'll quote it here for you:

"Consider the following analogy: we have two rectangles each with the same number of squares on them covering the entire area. If we are throwing baseballs at the squares, we need to be more accurate when throwing at the smaller target than the larger target. In fact, we can be more precise: if we call the ratio of the sides of the large rectangles the crop factor, then we have to be more accurate by a factor of the crop factor for the smaller target than for the larger target. Of course, we only need to be as accurate as is necessary to hit the squares, and more accuracy doesn't really improve anything.

In other words, as long as glass is "sharp enough" to hit the pixels, additional sharpness does not help. Of course, as we know, glass does not have the same sharpness across the image. Thus, it may be "sharp enough" in the center, but too soft on the edges. This is what happens when comparing, for example, 4/3 lenses with 35mm FF lenses. The 4/3 lenses are sharper than the 35mm lenses (on average), but they need to be sharper to hit the smaller pixels of their sensors. And while 35mm FF glass is easily "sharp enough" for the center of the image, it is not "sharp enough" for the edges, despite the larger pixels. Thus, at some point, sharper glass on a smaller sensor may outperform less sharp glass on a larger sensor in the corners, but the amount of the corners where this reversal in sharpness occurs is dependent on which lenses are being compared, and must be taken on a lens by lens basis. In the evidence section of this essay there are examples detailing this issue. However, in almost all cases, if not all (I've not compared every lens combination), the image from the larger sensor system will be sharper overall. What do I mean by "sharper overall"? Well, if the smaller sensor system is sharper in the 5% extreme corners of an image, that would amount to a 400 pixel square in each corners of a Canon 5D, and the evidence suggests the difference is even less.

All of this is particularly true (and much less theoretical) if the
photographer frames the subject so that it "fills the frame" the same
way on crop and FF and uses the best available lenses on both. The FF
image will be "better" as a print.

If the 1.6 crop lens was 1.6 times sharper than the FF lens, would
they be equal?

If the 1.6x sensor and FF sensor had the same number of pixels, had the same design and efficiency, and we shot the same scene at the same FOV, DOF, and shutter speed, then, yes, if the 1.6x lens was 1.6x times sharper everywhere in the image circle, they'd be as equal as you'd ever be able to get two images from different format systems.

Of course, in practice, we don't get any of those ifs in terms of the lenses.

--
--joe

http://www.josephjamesphotography.com
http://www.pbase.com/joemama/

 

[joe mama]

It is critical to remember
that the "out of focusness" of a crop image will be enlarged more to
achieve a given print size than the "out of focusness" of a full
frame image, since the latter starts larger.

Does this mean that the background blur of the image from the crop
sensor will look the same as the background blur from the FF image
(when the image is printed at the same size)?

For the same FOV and DOF, yes.

Anyway, your asking hypothetical questions about hypothetical systems, which, of course, is all fine and dandy, as it helps form an understanding about the principles at work, just as working a high school ballistics problem that ignores the effects of wind resistance and a curved surface helps illustrate the principles of Newtonian Mechanics.

But eventually, we must shoot our rocket into orbit though an atmosphere and around a curved Earth. To that end, let's discuss some of the important "real world" differences between crop and FF:

1) Larger sensors have less shot noise and more read noise. What this means is that larger sensor systems produce much cleaner images at low ISOs when there is ample light, but when upping the ISO to maintain the same DOF and shutter speed in a limited light environment, the darker areas of the image from a larger sensor will exhibit more noise.

2) The MTFs of some FF lenses exhibits a sudden drop at the exteme edges. This means that cropped sensor systems using FF glass will not experience this sudden drop, as this portion of the image is cropped out, and the extreme corners (5% of the total image area) may be sharper, even though the rest of the image will be softer.

3) While I know of no example where cropped glass is sharper by a factor of the crop factor, most cropped glass is sharper than their FF counterparts. Thus, while FF will still be sharper overall, the difference is not as exteme as some make it out to be.

4) The fact that glass on 1.6x has to be 1.6x times sharper on FF does not mean that FF images are 1.6x sharper than 1.6x. Remember, as long as the glass is sharp enough to resolve the pixels, being sharper gives no additional advantage. Thus, if the glass can resolve all the pixels on 1.6x, and can also resolve all the pixels on FF, there is no advantage in sharpness for FF, but there is an advantage for whichever sensor has more pixels.

Let's discuss this last point a bit further. Let's say we have FF glass that is sharp enough to resolve 12 MP on 1.6x. Then that lens will produce the same sharpness of image on 1.6x as it does on FF if both sensors have 12 MP. However, if the glass is sharp enough to resolve 12 MP on 1.6x, then it is also sharp enough to resolve 31 MP on FF, so if a FF sensor had 31 MP, then it would obviously have a big advantage.

On the other hand, let's say that the glass is only sharp enough to resolve 12 MP on FF. Then it will not be sharp enough to resolve 12 MP on 1.6x. However, if a 1.6x counterpart is made for 1.6x, and it is 1.6x times sharper than the FF lens, then both the FF sensor and 1.6x sensor will have the same sharpness and be able to resolve the same detail.

In practice, of course, lenses lie somewhere between these two extremes, with interesting twists, such as the sudden MTF falloff at the extreme edges of some FF lenses mentioned earlier.

Lastly, as I've said many times before, none of this matters if your focus is not accurate enough to resolve the pixels, if the camera shake or motion blur is present (and don't think that's an issue only for very low light -- why else would landscape photographers be using a tripod and MLU?). Finally, there's print size. For an 8x12 print, 300 PPI only requires 8.64 MP. On the other hand, 300 PPI for a 12x18 print requires 19.44 MP. So, how big are you printing? Even so, just how big of a difference will 200 PPI, or even 150 PPI vs 300 PPI matter to you?

It's important to understand the advantages/disadvantages of the equipment. But this can all too easily be taken to an extreme where it is well past the point of having any meaningful purpose.

--
--joe

http://www.josephjamesphotography.com
http://www.pbase.com/joemama/

 

[joe mama]

I'm a firm believer that the differences in operation
between cameras matters far more for most people than the differences
in IQ.

I totally agree....I think. Would you place high ISO noise reduction
under operation?

I place it under "PP -- Post Processing" which is another factor to consider, in that some RAW converters may favor the RAW files of some cameras over others.

After studying samples from the D300 at ISO3200 and 6400, I've come
to the conclusion that I can use shots at those settings.
The 5D has no ISO6400 setting. Manually pushing ISO3200 shots is too
much trouble. And the 5D doesn't have different high ISO noise
reduction settings to chose from right? I don't like adding an extra
step to my workflow, but would do it if it weren't too complicated. I
like being able to chose how my picture looks in-camera and then
shoot JPG.

Now that falls under "operation" and you're 100% correct in your analysis. However, I would advise against jpg, since memory is cheap, and RAW converters can help a lot for when you screw something up, or just need the most out of a particular image.

Speaking for myself, I'm a minimalist, I guess, in search of maximum
IQ for the money I can afford to spend.

For me, IQ is of a lower priority than features that enable me to get
the shot...like the AF system - which is why the 51-point AF of the
D300 interests me, subjective quality of the high ISO noise reduction
system (enabling me to shoot at higher ISO's without having to deal
with 3rd party noise reduction software and such), and sometimes auto
white balance (to which I have to declare Fujifilm as the clear
winner by a large margin).

100% agree.

My indispensable lens is the 18-200mm VR. And since there is just
slightly over 1-stop of difference in noise, with fancy noise
reduction software (whether in-camera or 3rd party) it could narrow
the gap (depending on what you shoot of course - and personal
preference). And as you might notice in the other messages, I'm still
trying to get a handle of this difference in DOF, but right now, to
me it's more a luxury than a requirement.

Canon may soon release an 18-200 IS. Regardless, from all you're saying, I'm thinking the D3 has your name on it. : )

That said, if you could spring for a D3, I'd really push that option.

Not at this time =) It's too expensive.

Seems like a killer camera.

Yeah, and to me the usability at the higher ISO's is too cool.

Well, if the D3 is too pricey, then it seems like you're in a position of choosing a D300, 5D, or 5DII. If you're not in a rush, you may wish to wait and see what the 5DII brings to the table. I know it's on my list of things to get. : )

The only lens that really interests me on the Canon side is the
24-105 F4 IS. But I have shot a wedding with a 24-70mm F2.8 on a crop
body and I often found myself wanting to zoom-in more. I didn't have
trouble at the wide end even though it was only 24mm.

There was a lens that interested me for a while. The Tamron 28-105mm
F2.8 But I think there were some sharpness/contrast issues with it.

For the record, I don't shoot weddings. But I have shot two: the first with a Canon G3 (4 MP compact digicam -- I was the backup photographer and did it just for the free food) and the second was with a 5D and 50 / 1.2L (I didn't use any other lenses, actually). It's hard to imagine two larger extremes in shooting than this.

In both cases, the people raved over my photos, although, and this must be obvious, there was a huge difference in IQ between them. So, while I wouldn't even consider a 5D + 24-105 / 4L IS to shoot a wedding, I shot with far "worse" and got rave reviews.

Again, it's important to put "IQ" into context with "expectations".

I really wonder how effective in-camera IS would be with a 300mm
lens. Wouldn't the sensor have to move quite a bit? I'll have to see.

Don't know, don't care. My longest lens is 150mm. : )

In any event, you could ask about it in the other forums with in-camera IS DSLRs (make sure to ask for 100% crops) and get an idea.

And for me, I'd have to add that having a stabilized 85mm F1.4 would
be very nice.

You and Lee Jay (ljfinger in this forum) both. He desperately wants such a lens. As for me, I'm clammoring for a 100 / 1.8L IS instead.

One nice Canon crop zoom that caught my eye was the 17-55mm F2.8 IS.

Nothing but rave reviews.

And you're right. I wish they would add IS/VR versions of all their
lenses.

I trust you're not going to hold your breath waiting for that to happen, right? : )

--
--joe

http://www.josephjamesphotography.com
http://www.pbase.com/joemama/

 

[Undah]

How often are we talking about cleaning the sensor, and will a blower do or are you talking about wet cleaning?

If you can live with cleaning the image sensor on a regular basis and
the below average LCD then its a bargain...

 

[Amin Sabet]

I did even less testing than Ken, but I'll tell you my experience. I went from a 30D to a 5D and immediately noticed that my lenses performed better. They were sharper on the 5D. I didn't analyze the corner performance because the overall improvement was apparent enough. I don't know if the 5D shines because of a low AA filter or because of big pixels, but it was a clear step up for me. I also feel that Capture One 3.x brings out the best in Canon DSLR files, particularly those from the 5D.

--
http://aminphoto.blogspot.com
http://aminsabet.tumblr.com

 

[landmarks4711]

Not sure, but the OP included the following in his post: "I was just

now that's what I call "believe"... lol

reading an article on luminous landscape about FF and it said
something about people that go for FF get it and those that don't,
don't. I don't get it."

Thus making the reference relevant.

Dan

In general (though not every time) you can trust what you read on
Luminous Landscape a lot more than what you read from Ken Rockwell...

what did Luminous Landscape say about full frame advantage?
(link?)

Just sayin'

Dan

--
---
G Dan Mitchell
SF Bay Area
http://www.gdanmitchell.com/

--
---
G Dan Mitchell
SF Bay Area
http://www.gdanmitchell.com/

 

[iano]

There are two separate questions here:

q1) does full frame inherently offer better images than cropped

q2) does the full frame advantage translate a better picture than cropped in all cases

Q1- yes.

Q2- No

For q1, yes a larger sensor has the ability to capture more light and greater detail, resolution and dynamic range. However physically larger lenses are also in the same circumstances. A particular 35 mm telephoto 35mm lens may be large, but the medium format equivalent will always be larger and even more expensive possiblty to the point that noone makes one. Larger is always better- but more expensive and inconvenient. Choosing the right trade off for the job is the challenge.

For q2, just because a camera is has a larger sensor does not necessarily translate to better. Technology moves on and newer cameras with a smaller sensor can be better than older technology larger sensor. Even at the same release time, the best crop may be better than the worst full frame.

So in the long term usually the best full frame will be better than the best crop camera and the best medium format will be better again. But for a given budget the best solution may vary.

 

[Joachim Gerstl]

Hi Dan,

Have you ever dried a full frame camera?

--
Joachim
http://www.joachimgerstl.com

 

[Amin Sabet]

a 35L is a 35mm lens. Not a 50mm something.
a 135L remains a great portrait lens.

That doesn't say anything...at all...to me. Besides, the 85 1.2 is
better.

How is the 85L better? It's less sharp wide open, heavier, more expensive, focuses slower, and the bokeh is a wash.

--
http://aminphoto.blogspot.com
http://aminsabet.tumblr.com

 

[Dan62326]

How is the 85L better? It's less sharp wide open, heavier, more
expensive, focuses slower, and the bokeh is a wash.

The 85 is less sharp wide open because it is WIDE OPEN when wide open. They're both about as sharp at F2. I'd rather have F1.2 available for available light photography.

And the 85 might be a little heavier, but it doesn't stick-out as much so the weight distribution is different. I don't really know how it would affect ergonomics in real life though.

But I was just pointing out that there was a lens available for the crop bodies that gives that angle of view.

 

[Dan62326]

Hi Dan,

Have you ever dried a full frame camera?

Nope. I'd love to. I'm here to find out about FF and if it's practical for me at this time.