Upgrade to Full Frame Inevitable?

[sk66]

You really think this difference is down to red channel info only?

Nope. It's also due to the DR/color depth advantage the D800 has at lower ISO's (it goes away pretty quickly at higher ISO's).

My personal opinion of the D800 is from field use, not "measurements" or reviews or anything else I've read. If/when I'm able to use the D800 to benefit, I will/do. I have an antique pocket watch I want to do some macros of and it will be the camera I choose...well, I'm also going to try the 16MP V2 adapted to the macro lens for the 2.7x... I'm not entirely certain how that comparison will turn out.

 

[Pookeyhead]

Nope. It's also due to the DR/color depth advantage the D800 has at lower ISO's (it goes away pretty quickly at higher ISO's).

My personal opinion of the D800 is from field use, not "measurements" or reviews or anything else I've read. If/when I'm able to use the D800 to benefit, I will/do. I have an antique pocket watch I want to do some macros of and it will be the camera I choose...well, I'm also going to try the 16MP V2 adapted to the macro lens for the 2.7x... I'm not entirely certain how that comparison will turn out.

So it's sharpness with the same lens is down to it's bit depth and dynamic range? Really? So if I do the same test with a ISO3200 set of RAWs the D4 will be sharper.. is that what you're saying?

 

[Pookeyhead]

...again... 2A0 print size. 100% crops.

ISO3200 @ f11

D4

D800E

errr....

Nope

 

[sk66]

The same lens also delivers 28MP compared to the 15MP of the D4 and the 17MP of the D7100 when it's mounted on a D800... So where does that leave your argument?

Exactly where I said... larger isn't "always" better in a strict "sharpness" measurement, but for most of us/most of the time it really is. It also shows that COC isn't constant for a given sensor size. At least not with a lens capable of very high resolutions.
While lab measurements are only that, and they ignore everything else, and I might not even entirely agree with *how* they determine their results. They do make their tests on raw files and it's empirical/repeatable testing of greater reliability than most things you will find on the web.

If I want to judge a piece of equipment.. I take photographs with it

As do I. Given the choice between using the D4 or the D7100 I will always choose the D4 (I don't think *I* would ever buy the D7100). And I will often choose the D4 over the D800... even though the D800 "can be" the better camera, it's usually not "better" for me and what I'm doing.

 

[Pookeyhead]

Exactly where I said... larger isn't "always" better in a strict "sharpness" measurement,

What you're seeing is DxO's flawed testing methodology. They're just reading off MTF charts without compensating for the magnification I reckon, so of course it's going to read higher. Like for like... meaning a 200mm on the D7100 and a 300mm on the D800, it would be a different story.

And I will often choose the D4 over the D800... even though the D800 "can be" the better camera, it's usually not "better" for me and what I'm doing.

I've also just demonstrated that the D800 kicks the D4's ass for sharpness at higher ISOs too. When exactly do you reckon the D4 is the better choice? So far as I'm concerned, it's when you need to the speed and durability it was designed for... not much else really. Whenever quality is the factor that matters, I'd ALWAYS be reaching for the D800. How many more actual images do I have to post before you realise? Keep looking the numbers if you want.... but the images disagree with you.

 

[sk66]

...again... 2A0 print size. 100% crops.
Nope

Boy, you really want me to love the D800 as much as you do... I don't even know how this got to be a debate on the D800 vs D4 anymore... They're the same sensor size...

IME, ISO results in good light by varying the aperture or SS are very different than ISO results when ISO is increased due to an actual lack of light. You're probably going to say "increasing the SS creates a lack of light," and I'm going to say "yes, but it's not really the same thing".... I'm tired of this.

I never said the D4 *is sharper* I said "it can be."

 

[sk66]

What you're seeing is DxO's flawed testing methodology. They're just reading off MTF charts without compensating for the magnification I reckon, so of course it's going to read higher. Like for like... meaning a 200mm on the D7100 and a 300mm on the D800, it would be a different story.

Well they aren't reading charts, they are testing the same as anyone with the equipment could do, and the results are compensated for sensor size (same size print/image). I don't imagine they are compensated for FOV/magnification in these tests. But they are in their body only tests (and the D800 rates the highest)

Keep looking the numbers if you want.... but the images disagree with you.

I'm not looking at numbers to make my decision. I usually have both cameras with me and I've used them both in many situations. The times the D4 will generally produce better images for me are when:
I don't have the stability the D800 requires (usually), when I have to use higher ISO's due to a lack of light (not uncommon), and when the subject is small/far/fast (common). Often I can get better results by adding a 2x TC and using the D4 than I can get by using the D800 and cropping slightly... you don't have to believe me, and I don't really need to convince you.

 

[steveclem]

Are you having a conversation with yourself? Sorry, only jesting, I like full frame because a 50mmm lens becomes a 50mm lens.
I like APS-C because a 50mm lens becomes a 75mm lens, I use both types of camera and it suits all my requirements.

 

[arad85]

The "circle of confusion" for a lens *IS* the airy disk. The airy disk is solely a function of the aperture size (or "apparent size"). "Aperture" is a function of the opening size in relation to the focal length. Subject distance only has the effect as to if the lens is focused on it to start with.

No, the Airy disk is a diffraction based effect when passing through a constrained aperture. The Circle of Confusion is defined as an optical spot caused by a cone of light rays from a lens not coming to a perfect focus when imaging a point source because it is off the focal plane. They are two distinct things. Have a read through of: http://en.wikipedia.org/wiki/Circle_of_confusion and http://en.wikipedia.org/wiki/Airy_disk and you'll see that the CoC is defined as a combination of distance from focal plane and aperture (not even focal length as aperture used in the equation is measured in unit length) and defines the size of the "blob" that gets rendered on the sensor based on how far away it is from the focal plane. The Airy disk depends on distance and the wavelength of the light (which shows it to be diffraction based). Fundamentally, the diffraction limit for a camera system (i.e. combination of lens and sensor size) is where the Airy disk size becomes larger than the CoC diameter limit - that is, the aperture at which the diffraction effects become more noticeable than the CoC effects when printing. Note that the CoC diameter limit is the (theoretical as it is based on average visual acuity) threshold at which you become aware of the CoC causing a point not to be in focus on a print.

 

[Phil V]

Are you having a conversation with yourself? Sorry, only jesting, I like full frame because a 50mmm lens becomes a 50mm lens.
I like APS-C because a 50mm lens becomes a 75mm lens, I use both types of camera and it suits all my requirements.

A 50mm lens is always a 50mm lens. What changes is the angle of view. The focal length of the lens is constant.

 

[sk66]

They are two distinct things. Have a read through of:

I have read through that, and honestly I don't see the distinction you're making.
COC is a perceptual criterion for sharpness. At it's base it is dependent on how sharply a point light source can be rendered/projected and it is dependent on accuracy of focus (which is dependent on lens/wavelength/aperture/etc etc.
The airy disk is the exact same thing except that it is not a perceptual criterion, it is physics. The COC criterion is dependent on the airy disk physics.

Way too much here: http://www.luminous-landscape.com/tutorials/resolution.shtml

 

[sk66]

I like APS-C because a 50mm lens becomes a 75mm lens, I use both types of camera and it suits all my requirements.

Sometimes I use the D800 similarly...in DX mode for extra "reach" or "cropping in." But then it becomes a 16MP D7000 and I loose all of the benefits of it being the D800. The results are generally worse than using a 2x TC w/ the D4. Granted, this isn't always the case for every lens/TC combination but with the latest 400 f/2.8 and TC-E III it is. At least in my experience it is.
Most of the difference is due to the high stability/SS/panning accuracy requirements of the smaller pixels which are compounded by the long FL's. I simply can't often meet those requirements when shooting action.

 

[arad85]

I have read through that, and honestly I don't see the distinction you're making.

You're missing the subtleties then. CoC is the bluriness of a point in the image. A long way from the focal plane it is a big blur, close to the focal plane it is a small blur. Whether you can see that blur in the output image is determined by how big you print - i.e. how much you magnify the captured image. This magnification factor is incorporated into a threshold used to calculate depth of field which is called the Circle of Confusion diameter limit. This is a threshold above which an average person will see things as blurred in a standard sized print viewed from a standard distance. Change any of those assumptions and CoC diameter limit changes proportionately. This CoC diameter limit varies depending on sensor size as the smaller the sensor the more you have to enlarge the resultant image to get a given print.

ICOC is a perceptual criterion for sharpness. At it's base it is dependent on how sharply a point light source can be rendered/projected and it is dependent on accuracy of focus (which is dependent on lens/wavelength/aperture/etc etc.

No. CoC is a continuously varying value which determines how a point in the image is rendered. Other than optical issues in the lens, it is solely determined by aperture and distance from the focal plane. Think of it as the bokeh of the image - a wider aperture produces a bigger blur (bokeh) the further the point source is off the focal plane. It has nothing to do with how accurately you can focus a point with the lens.

Where people are going wrong - including both yourself and the luminance landscape article - is using the words CoC where in fact they mean the CoC diameter limit. Have a look here: http://photo.andysheen.co.uk/technical/depth-of-field-and-what-affects-it-explained/ for a more detailed explanation (without maths) which tries to explain it a bit more.

Where diffraction comes into it is that with a small CoC diameter limit (due to a small sensor) the greater the effects of diffraction. A system is said to be diffraction limited when the airy disk size exceeds the CoC diameter limit. Wider apertures will be governed by CoC size to determine if something is in focus, beyond the diffraction limit, the airy disk becomes bigger than the CoC diameter limit, so it dominates what is perceived as in focus or not.

The mistake you are making is assuming the perception of in or out of focus is captured when you take the picture. It isn't. It is ONLY perceived when you render it in a print or on screen...

 

[BBR]

Anyway op, I moved up from a d200 to a d3 and the difference in what I could see on my pc as well as print was frequently wow!

 

[Pookeyhead]

Boy, you really want me to love the D800 as much as you do

No.. I think you're deluded about your D4

I genuinely could not care less what camera you like or dislike. They're just tools. Seriously. I am baffled that you refuse to believe what your eyes are telling you though. First you say the D800 is put at a disadvantage with higher apertures... so I post files at higher apertures that proves you wrong.. then you say Ahh, well... that's down to the dynamic range and colour depth... it wouldn't be as good at higher ISOs... so I post files that demonstrate that it's still better at higher ISOs... and now you're saying Ahh well... that's not actually a higher ISO file necessitated by actual low light. LOL. You're incredible.

I never said the D4 *is sharper* I said "it can be."

I'd love to see the RAW files that demonstrate that. Feel free to dropbox and link to them any time you like.

 

[RedRobin]

*cough*

What does my signature below quoting Ansell Adams say?

And it hasn't changed since he said it many many decades ago.

So use whatever format and gear you feel most comfortable with.... Simples.

 

[sk66]

I'd love to see the RAW files that demonstrate that. Feel free to dropbox and link to them any time you like.

I'd be happy too... but I don't keep images that don't meet the cut...maybe next time.
Do you really think I would choose to use one camera over the other based upon anything other than results when I own both and have them immediately available? That would be pretty stupid...
I believe/feel/see that the D800 is a "studio camera" and there are many reasons/specs for this... we can disagree.
Not being metric, I didn't realize you were showing a 6ft image at 100%... you do realize how unrealistic that is don't you? At least for most... IMO it *is* realistic for a fine art photographer with larger print sizes. And situations where you can add light, use a tripod, use lower ISO's in lower light... But those are not "my" situations.

 

[Pookeyhead]

*cough*

What does my signature below quoting Ansell Adams say?

And what does mine say?

I just feel you use what's most appropriate.

 

[Blank_Canvas]

No, the Airy disk is a diffraction based effect when passing through a constrained aperture. The Circle of Confusion is defined as an optical spot caused by a cone of light rays from a lens not coming to a perfect focus when imaging a point source because it is off the focal plane. They are two distinct things. Have a read through of: http://en.wikipedia.org/wiki/Circle_of_confusion and http://en.wikipedia.org/wiki/Airy_disk and you'll see that the CoC is defined as a combination of distance from focal plane and aperture (not even focal length as aperture used in the equation is measured in unit length) and defines the size of the "blob" that gets rendered on the sensor based on how far away it is from the focal plane. The Airy disk depends on distance and the wavelength of the light (which shows it to be diffraction based). Fundamentally, the diffraction limit for a camera system (i.e. combination of lens and sensor size) is where the Airy disk size becomes larger than the CoC diameter limit - that is, the aperture at which the diffraction effects become more noticeable than the CoC effects when printing. Note that the CoC diameter limit is the (theoretical as it is based on average visual acuity) threshold at which you become aware of the CoC causing a point not to be in focus on a print.

So is the same as Boker ?

 

[arad85]

So is the same as Boker ?

No as the lens and aperture shape has a lot to do with it, but it is the physical phenomenon that causes it.....

 

[ChrisHeathcote]

I can't believe this conversation is still going round in circles

 

[Blank_Canvas]

Just out of idol curiosity, there are plenty of rumours on the web of Fuji making a full frame CSC this year, even though they have publicly claimed it won't happen. I believe it will happen so they don't dip out of the market and keep up with Sony etc. I digress, I'm no expert on the subject of sensors, sizes etc but I do know DX/FX lenses are completely different. My question then, if Fuji do produce a FF sensor would they still be compatible with the X series do you think ?

 

[Alastair]

My question then, if Fuji do produce a FF sensor would they still be compatible with the X series do you think ?

No. Fuji have answered this question already.

 

[Blank_Canvas]

No. Fuji have answered this question already.

That's an emphatic 'No' !

I remember Nikon publicly stating in the mid 1980s when they launched the F4s that they would never produce a camera that didn't fit their F lens system. All that went out of the window recently when Nikon launched the Nikon 1

Companies often keep their business strategy under wraps, circulate disinformation and don't always disclose their hand until they are ready. In addition, change and market competition also dictates new straggles and products.

I wouldn't be to cocksure if I was you....best to keep an open mind on these matters




Sent from my iPad using Talk Photography Forums

 

[ojo]

I can't believe this conversation is still going round in circles

Circles of cunfusion?

 

[Phil V]

That's an emphatic 'No' !

I remember Nikon publicly stating in the mid 1980s when they launched the F4s that they would never produce a camera that didn't fit their F lens system. All that went out of the window recently when Nikon launched the Nikon 1

Companies often keep their business strategy under wraps, circulate disinformation and don't always disclose their hand until they are ready. In addition, change and market competition also dictates new straggles and products.

I wouldn't be to cocksure if I was you....best to keep an open mind on these matters




Sent from my iPad using Talk Photography Forums

But there's a difference between No we will never do something, and No they wouldn't work. The x series lenses are designed for crop sensors, like Canon EFS and whatever the Nikon nomenclature is (DX?). So Nikon and Canon and Sony can easily produce FF CSC cameras, because they have lenses to fit, but Fuji would need new lenses, the x ones wouldn't work.

 

[Blank_Canvas]

But there's a difference between No we will never do something, and No they wouldn't work. The x series lenses are designed for crop sensors, like Canon EFS and whatever the Nikon nomenclature is (DX?). So Nikon and Canon and Sony can easily produce FF CSC cameras, because they have lenses to fit, but Fuji would need new lenses, the x ones wouldn't work.

Thanks Phil, that's answered my question. So if Fuji did produce a FF version it would be outside their X Series, just as Sony have just done with their Sony Alpha 7. Interesting camera that, according to The Amateur Photographer magazine review, it's the only camera that produces image resolution on par with the Nikon D800 / 800e.

It must be a very difficult time for camera manufactures at the moment knowing which way to go.


Sent from my iPad using Talk Photography Forums

 

[sk66]

You're missing the subtleties then. CoC is the bluriness of a point in the image.

So your saying COC is a characteristic of a lack of focus but has nothing to due with the ability to focus. And that COC (blurriness) is a characteristic of the lens and not dependent on viewing the image. None of that makes any sense.

COC only comes into discussion regarding DOF/sharpness and as such it is *not* a characteristic of the lens. It is entirely dependent on how the image is displayed/viewed whether at 1:1 on the sensor or some other print size/distance. The only difference between COC and COC diameter limit is that "the limit" is a somewhat arbitrary definition of *when* something is "acceptably sharp."
None of your references/links actually say anything different.

Where people are going wrong - including both yourself and the luminance landscape article

You do realize who created that article don't you?
Michael Reichmann,
Norman Koren (creator of Imatest)
Roger N. Clark (Ph.D, Clarkvision, NASA)
Brian A. Wandell (Ph.D, Stanford, digital imaging and neuro science),
Nathan Myhrvold (Ph.D, Cambridge, theoretical physics and inventor)

You'll have to forgive me for believing them over you...

 

[arad85]

So your saying COC is a characteristic of a lack of focus but has nothing to due with the ability to focus. And that COC (blurriness) is a characteristic of the lens and not dependent on viewing the image. None of that makes any sense.

You're mixing things again. CoC is the circle that is projected on the sensor by a point source that is off the focal plane. Nothing more, nothing less. It is nothing to do with how well the lens can focus (other than if a lens can't focus, a point source becomes a circle and points off the focal plane become slightly larger).

COC only comes into discussion regarding DOF/sharpness and as such it is *not* a characteristic of the lens.

NO! CoC diameter limit is used for DoF calculations. It is a threshold on CoC which predicts when we humans see things as in or out of focus and changes depending on viewing conditions. Using CoC when CoC d/l is meant is very common, but not correct.

And as to what it is, I'll say it again... CoC is a characteristic of the aperture size and how far off the focal plane something is. No more, no less... Look at the equations that define it on the Wiki page - it's simple trigonometry.

You do realize who created that article don't you?
Michael Reichmann,
Norman Koren (creator of Imatest)
Roger N. Clark (Ph.D, Clarkvision, NASA)
Brian A. Wandell (Ph.D, Stanford, digital imaging and neuro science),
Nathan Myhrvold (Ph.D, Cambridge, theoretical physics and inventor)

You'll have to forgive me for believing them over you...

You know... I don't really care who wrote the article. Their use of the terminology is incorrect if they don't use the term CoC diameter limit when discussing depth of field..... It confuses people and seems to have confused you too...

I've led you to the water... up to you if you want to drink it