Full frame cameras

[billtils]

Something similar can be said about pixels; two cameras of the same generation of sensors and pixel pitch should perform virtually identically.

The only time we can actually make out a difference is if we magnify the images from thos sensors. The full-frame image can be enlarged more before softening sets in than a crop frame sensor. So unless you are making large prints and then pixel peeping, one would be hard pressed to see any differences.

Manfred, Is that correct re ISO noise and FF -v- crop? It certainly wasn't re my crop D7100 and FF D750 (but both are pretty long in the tooth).

[Manfred M]

Noise is related to pixel size and hence pitch (i.e. the size of the individual sensor element). The D7100 has a pitch of 3.9 μm whereas the D750 is a lot larger 5.9 μm, so the low noise performance of the D750 should be significantly better than the slightly older D7100 (which uses one generation older technology as well).

[DanK]

Dan, Many more years ago than I care to confess to, there was an ad for Castrol motor oil, with the punchline "Oils ain't oils", and I have a faint memory (as in sort of recall the event but not sure where or when) that "pixels ain't pixels" and those on a FF have some advantages over crops, certainly ISO noise but perhaps more general IQ too? Is this still the case (if it ever was)?

Bill,

Let's leave aside for a moment one big complication, which is improvements in sensor technology over time.

At any one time, it is generally the case that larger pixels perform better because they collect more light. For that reason, most FF sensors, if they have room--which is to say, if they have lower pixel density--use larger photosites which yield better results (including less noise). The tradeoff for macro work (back to Joe's OP) is that you then get fewer pixels on the macro target.

In recent years, Nikon and Canon have introduced FF cameras that have pixel pitches similar to their crop cameras. One example is the one above: the 5DsR and the 7D2. In that case, for macro work (an image of a fixed size on the sensor), the FF has no advantage over the crop unless the wafer is made from a better process. If the sensors are cut from the same wafer, they will yield identical macro results after cropping.

The big advantage of those cameras is for other types of work, like landscape work, where you can compose so that the subject fills the frame. This gives you a huge number of pixels.

Since I have a 5DIII (bigger and better pixels than my 7D, but a lower pitch), I prefer the former for all macro that has low enough magnification that I can fill the frame, such as a good bit of flower macro. However, when I am at 1:1 and the size of the bug is what determines the size of the image on the sensor, I prefer my 7D which has many more, albeit lower quality, photosites on the subject.

Dan

[James G]

Your reasoning would work if the framing were the same on the two cameras, that is, if the bug took up the same proportion of the total sensor area on both cameras. However, it won't. it will take up exactly the same area.

Hi Dan... I stand corrected....
I went for the 5DsR after reading about the write-ups about Canon's decision to remove the high pass filter. I was able to hire one for a few days before buying and was surprised (pleasantly) to find that I was seeing less softening as I 'enlarged'/zoomed in. It was detectable when I examined my stained glass captures, though as Manfred indicated I was more or less pixel peeping . Arguably perfect! capture conditions for the test since I used a heavy duty tripod and filled the frame with the test window for both cameras. I seem to remember I even used mirror lock up to avoid any camera shake. After reviewing the results I guess I stopped thinking.

As regards captures in the wild, maybe I'm just getting better .... and attributing it to the technology.

[rpcrowe]

At the risk of repeating what I have probably stated in other postings, here is my take on Full Frame vs Cropped Sensor cameras...

The full frame camera is not a magic wand which will elevate mundane photo efforts to heights of grandeur

If, and only if, you are getting very good to excellent imagery with a crop camera (given that you are using top line lenses) you may see an improvement when switching to full frame - especially if you are making relatively large prints.

However, if your photos with a crop camera and top quality lenses are mediocre and boring, I strongly suspect that those mediocre and boring qualities will carry over to your full frame efforts.

I shoot with both full frame (a Canon 6D2) and a crop sensor (the Canon 7D2) and I like both of these cameras for different reasons. However, I don't see a terrible great difference in the quality images I post on the Internet. I don't do a lot of printing...

I entered the full frame field with a 5D2 because I use a 70-200mm f/4L IS lens to photograph my rescue dogs and I did not have enough room in my shooting area to back up far enough to shoot comparatively larger dogs using that lens on my 7D2. After-all, the equivalent focal length of the short side of that lens is 112mm when shooting with a crop format camera. The 70mm on a full frame camera gives me plenty of leeway in shooting room.

I normally shoot with two cameras when I am doing walk-around shooting. I like the 17-55mm lens on my 7D2 and like the 70-200mm f/4L IS lens on my 6D2. I switched over to the 6D2 from the 5D2 in an effort to get improved autofocus and because I got a really good deal on that camera... The full frame - crop sensor combination is great. Because, that way, I don't have a gap between the equivalent focal lengths of the two lenses. The 17-55mm is an equivalent to 22/7mm to 88mm and the 70-200mm on a full frame camera gives me a small bit of overlap between the two rigs which I enjoy.

BTW: did I get improved autofocus? Yes and no! The 6D2 focus array is skunched (my own word) at the center of the sensor making eye level AF difficult under some situations. However, the Dual Pixel autofocus in live view mode is terrific... On the crop side, the 7D2 autofocus is just short of miraculous...

[billtils]

Thanks Dan - seems like I should be getting into landscape .

On a more serious note, and entirely subjective (which is a shocking thing for a retired academic scientist to say) from day one of switching I have enjoyed my FF bodies more than the crop ones. Not easy to pin down but it may be as simple as the FF has more flexibility and works well in more situations than a crop would - or at least seems that way.

[joebranko]

Thanks everyone for your input and suggestions. There is a huge amount of info offered here and I will take sometime to absorb it all and decide. Thank you!

[joebranko]

My question to you would be what is it about going to a full-frame camera is going to give you better images you are getting with a crop frame camera?

I was under the impression that the greater number of pixels would produce sharper images?

If you make large format prints, you will definitely notice a difference, but if you just view on a computer screen or post on the internet, you won't see any difference.

Almost all of my pictures are on a screen
I have only made large format prints for posting on my wall.

[rpcrowe]

Unless you are shooting professionally, the ultimate criteria (for me anyway) is which gear do you enjoy shooting with the most...

I have liked almost all of my digital cameras (the one Rebel that I owned is the exception), however, there are some lenses that I really enjoy using more than others.

[Manfred M]

I was under the impression that the greater number of pixels would produce sharper images?

While that can be true, it depends on what you do with them. A standard computer screen has a resolution of somewhere around 1920 x 1080, or 2MP. If you view the output of a 20MP camera on the screen, it is downsampled and 90% of the data that your camera has captured is throw away. So if you primarily display your images on a screen, there is little to justify more pixels.

Almost all of my pictures are on a screen
I have only made large format prints for posting on my wall.

I print a lot of 17" x 22" images using a printer with a native resolution of 360 DPI; which means I have to go the other way and upsample and create more pixels to print from a 4912 pixel x 7360 pixel (36 MP) image. In my case I can justify using the higher resolution I get with my FF camera because of my print making. Upsampling will result in a bit of pixel softening, although at the size I am printing, that will be fairly minor.

[inkista]

I was under the impression that the greater number of pixels would produce sharper images?...

You may want to look at a very old post of mine that side-by-sided some 21MP full frame (Canon 5DMkII) and 16MP micro four-thirds (Panasonic DMC-G3; essentially a 2x crop camera with a sensor smaller than APS-C) images, just to see what we mean about how at web sizes, you may not be able to easily see the difference.

I tend to say that full frame is an expensive sidegrade from APS-C, not so much a straightforward upgrade. Yes, you get better dynamic range and high ISO noise performance, and resolution. But the gear is bigger/heavier/more expensive, there are heavier storage/processing demands on your computer, and you may have to rebuy a lot of your lenses. Certainly, all of your existing lenses will change their character when used on full frame (do the crop math backwards). It might be worth it, depending on what/how you shoot, or it may not. That's up to you.

[William W]

This point has been sort of mentioned, but perhaps not actually nailed down. . .

I should have listed one other advantage for me having a 'full frame' DSLR. In point of fact it was the primary 'advantage' and the main reason why I bought an EOS 5D, many years ago:

I was working professionally, mostly Social Events and Weddings. I have several years experience in this area and in 2004 I was engaged to plan the 'cut over' of a major W&P Studio from Film to Digital.

In 2004 the Canon digital scene was . . . the 1Ds was released a couple of years before, and following the 10D, the 20D was released in 2004. The 20D (from a business point of view) was a light-year advancement on the 10D. Around the same time as the 20D, came the 1DsMkII, which was a great Camera, but from a W&P Studio's business’s perspective; simply, too expensive.

The underpinning logic of my proposal to the Studio was to use what I termed at the time 'a dual format kit'. My logic was that if we could fit out the Studio with a set of 20D Cameras and a set of less expensive ‘full frame’ Digital Cameras, we could drastically cut the expenditure that would be required on Lenses.

A salient point of the exercise to note is that the Studio was running Nikon (135) and Mamiya 645 SLR gear at that time, so, on my reckoning that Canon (at the time) was ahead of Nikon, in the DSLR race, it meant buying complete new sets of lenses: so any reduction in (initial) expenditure was very important.

It was in the wind that Canon was to release a (less expensive) 135 Format DSLR – and in 2005 that was the EOS 5D, with the 20D now both icons of Canon’s Camera history.

***

The bottom line:

The above is the back story – to address the question as it was asked - having a ‘dual format kit’ may be to some an ‘advantage’.

Obviously this means, in the context of this conversation, Joe keeping (or replacing) his APS-C and also buying a 'full frame' camera - thus having two cameras. It may also mean that Joe has to buy new lenses.

The ‘advantage’ is this, (best illustrated by one practical example) using the two cameras and two lenses we used:

With a ‘kit’ of a 20D and a 5D and a 16 to 35/2.8 and a 70 to 200/2.8 the Photographer had at disposal ‘the equivalent of’ (in 135 format FoV terms) 16 to 56 and 70 to 320 all at a Lens Speed of 2.8. The gap between 56 and 70 is insignificant. The lenses were always mounted and carried. The 16 to 35 was used on the 20D (kits were very quickly upgraded to 30D) as the main working camera (‘equiv 25 to 56’) and the 70 to 200 was on the 5D. This was the most expensive kit.

For some of the kits we didn’t buy a 70 to 200 preferring to get a 135/2 and a x1.4Extender EF, which was carried in the jacket pocket – OR – an 85/1.8. and initially we only bought one 16 to 35, using instead a fast prime for the wide end, for example a 35 (on the 20D as the main working lens) and swapped it t the 5D for any W/A shots as required.

The same leverage of having a large RANGE of Field of View, with only two or three lenses is still an advantage to some Photographers today - and the 'advantage' of having a second camera (as "back up").

I know some very keen photographers (not shooting professionally) who have "small kit" comprising something like a 7DMkII and a 5D MkIII and two or three very carefully chosen lenses and their kits are very, very, ‘flexible and expansive’ in terms of Field of View; Lens Speed and Redundancy and can cover a wide range of Photographic Tasks, easily.

WW

[george013]

I would like to add some more too.

First of all I think one has to distinguish between the optical differences of FF and crop and the practical, comparing specific camera's.
With optical I mean the aov at a certain distance with a certain focal length. That combination with an equal framing. Here is an image I showed not so long ago.

This image is taken with 15mm on a FF. It would be possible to take this one on a crop camera with 10mm at the same distance or at 15mm at a distance 1.5 times longer. In the latter case the framing of the first staple would be the same, but not of the others. The further away you go, the less differences in magnification of the other staples.
I don't think there is much difference between FF and crop camera at a longer distance. If you never shoot short distances you won't benefit of a FF. Optical.

George

[William W]

. . . First of all I think one has to distinguish between the optical differences of FF and crop and the practical, comparing specific camera's. . . With optical I mean the aov at a certain distance with a certain focal length. . . This image is taken with 15mm on a FF. It would be possible to take . . . at 15mm at a distance 1.5 times longer. In [this] case the framing of the first staple would be the same, but not of the others.

Yes.

I absolutely agree with this point.

For clarity, you (George) use the terms 'aov' (Angle Of View) and 'distance' (meaning Subject Distance).

I term the result of the element "Subject Distance" as "Perspective".

I do not want to debate the meaning of terms - I merely want to acknowledge that I understand George's point and that I agree with it.

WW

[xpatUSA]

This paper by Falk Lumo "true reasons for full frame" covers the subject in great depth, far deeper than expressed so far in this thread:

http://www.falklumo.com/lumolabs/art...alence/ff.html

His article on equivalence is worth reading first:

http://www.falklumo.com/lumolabs/art...nce/index.html
.

[billtils]

This paper by Falk Lumo "true reasons for full frame" covers the subject in great depth, far deeper than expressed so far in this thread:

http://www.falklumo.com/lumolabs/art...alence/ff.html

His article on equivalence is worth reading first:

http://www.falklumo.com/lumolabs/art...nce/index.html
.

Thanks for these links Ted. Enjoyed reading.

[xpatUSA]

<> If you view the output of a 20MP camera on the screen, it is downsampled and 90% of the data that your camera has captured is [thrown] away.

Not strictly thrown away, Manfred, because most re-sampling algorithms use data distant (in the original image) from the target pixel. Perhaps only the "Nearest Neighbor" algorithm actually tosses out such distant data when downsizing.

So if you primarily display your images on a screen, there is little to justify more pixels.

My cameras can record at several true (on-sensor raw binning) resolutions including full- and half-size.

I can usually get a "better" half-size image by down-sampling a full-size capture with the likes of Lanczos3 as opposed using to the binned version and that is because the binning is simply averaging four pixels into one, like bi-linear - nothing fancy.

[george013]

This paper by Falk Lumo "true reasons for full frame" covers the subject in great depth, far deeper than expressed so far in this thread:

http://www.falklumo.com/lumolabs/art...alence/ff.html

His article on equivalence is worth reading first:

http://www.falklumo.com/lumolabs/art...nce/index.html
.

From your second link

Of course, nothing about the real world changes when we use a different sensor, and in particular there is no change in:

the desired distance between the object and the lens
the size of the object to be photographed
the wavelength of light

It seems he is using a constant object distance in his calculations. As a photographer I prefer a constant framing.
Beside that there is the question of perspective again. The always mentioned phrase :

Perspective: This is an easy item to address because “perspective” depends only on the position of the camera lens and is unaffected by the focal length.

What is that dammed perspective.

George

[Manfred M]

This paper by Falk Lumo "true reasons for full frame" covers the subject in great depth, far deeper than expressed so far in this thread:

http://www.falklumo.com/lumolabs/art...alence/ff.html

His article on equivalence is worth reading first:

http://www.falklumo.com/lumolabs/art...nce/index.html
.

Thanks for posting Ted, these articles are a good read, but unless I missed something, the author has not gone far enough because the ultimate indicator of image quality is the output, rather than what happens at the sensor. On screens (even on 4K screens) images are generally downsampled so that they fit. Once we get to 8K screens, the picture is not quite as clear as only the highest resolution images will still have to be downsampled.

In the old view camera days, a camera with a 8" x 10" negative or 11" x 14" negative would be contact printed and a very crisp and clean full size image would result. Smaller negative sizes, whether they came out of a small format cut film camera, medium format camera or 35mm camera would have to be enlarged. In general, the view was that that a larger negative resulted in a higher quality print because the image would not have to be enlarged as much, hiding defects and resolution issues. The industry's "dirty little secret" was that lenses on larger format cameras did not have to be a "good" as smaller format cameras because (a) the difference in the amount of enlargement let them get away with it and (b) view cameras, especially larger sizes had to be shot at very small apertures (f/32 - f/64) because of the extremely narrow depth of field.

In terms of cost of producing cameras and lenses in full-frame and crop frame he also discounts two major cost drivers; design life and production quantities. In a world where everything is "equal", the cost of advantage of a full frame camera would not be that much higher than a crop frame; but they are anything but equal. While Falk does mention the cost of a the sensor, he ignores amortization of development and startup costs, use of more recent technology in in full frame cameras as well as lower production quantities that generally result in having to use more expensive manufacturing processes. Full-frame cameras are generally aimed at the pro or prosumer market, so these users will require a more robust build and a longer life (shutter actuations), both of which will drive up manufacturing costs.

Lenses are no different; the lenses built for the high end user will tend to be faster (maximum aperture), more robust, have more optical elements and will be built in smaller volumes than the "kit" lenses found on consumer cameras. All of these factors will drive up the unit cost of the full frame lenses.

[pschlute]

I always find it mildly amusing that many other forums are full of discussions over FF versus aps-c, but strangely very little is posted about FF or aps-c versus Medium Format (or any other format for that matter)