PDA

View Full Version : The differences between formats(full frame vs APS-C)



arthurking83
22-08-2013, 2:40pm
I have a bit of time to start this thread, but I have to scoot off in a while to get the kids.

Anyone is welcome to add to the thread any extra info or knowledge they have on other differences not mentioned, and hopefully we can then incorporate all the collated info into the AP library for future reference.

Firstly lets discuss the differences between formats:

the two most common formats we enthusiasts seem to be more interested in are APS-C(of which there are at least two different versions), and full frame which are most prevalent in DSLR cameras.

Of course there are many other sensor formats and the most commonly used format for cameras is the small sensor used in point and shoot cameras which is usually referred too 1/2.3" format.
Phones with cameras have an even smaller sensor than this, and this is important for anyone willing to test it out for themselves.

As sensor sizes differential widens, the differences referred too here become much more obvious .. so to compare a 28mm equivalent photo captured with a smartphone with it's tiny sensor, against an equivalent focal length on a large format camera such as MF(or larger) will highlight these massive differences.
The reality will be that to achieve a similar field of view with both cameras, the phonecam uses something like a 4mm lens, whereas the MF will use a a lens with an approximately 50mm or more focal length.
I'm not entirely sure on the actual tech specs on camera phones, but the difference is huge.

And it's this focal length variation that we want to concentrate on for part of this thread.

The apparent differences between the most common formats (APS-C and full frame) is about a 1.5x factor between an APS-C frame and a full 35mm format frame.
The most common APS-C format is used by Pentax Sony and Nikon, and Canon's slightly smaller APS-C sensor size uses a 1.6x multiplier factor.

So in theory a 35mm lens on an APS-C camera will give approximately a similar FOV as a 52.5mm lens on a full frame camera.
But FOV is not the only difference between the formats when attempting to balance equivalence factors, because other elements of lens's rendering ability will also impact on this equivalence factor.

A lens is a lens is a lens, and there's no buts about this.
That is to say, a 50mm lens(which can actually be a 52.5mm lens) on full frame is nothing like a 35mm lens on APS-C, and to view these differences you need to compare overlays of the images shot with equivalence in mind to see this.
Casually looking at two images side by side doesn't reveal the differences clearly enough.(which I'll attempt to do).
Lenses have magnification which is a property of the focal length.
Another aspect of magnification comes from the lens's ability to focus closely too, but that is a topic for another day, as it's more to do with closeup/macro photography. We're more interested in this thread about a lens's abilities and applications in infinity distances .. not strictly at infinity, but lets say from about 1m to infinity.

As focal length varies between lenses, other aspects of the lens change apart from field of view. In fact, the FOV of the lens can actually be designed to be static even tho the focal length can be altered .. if the lens designer wishes too. But again this is not part of the scope of this thread, it's more centred around the idea that an Xmm lens on a crop camera is equivalent to an X/1.5mm lens on a full frame camera.

While they look similar they are different.

Part of this is that a lens distorts the image in many ways. These distortions are not barrel or pincusion types, but more along the lines of perspective distortions, and to my knowledge I've never seen this issue addressed by lens designs, nor many replies on the topic.

Perspective distortions manifest themselves in various ways, and for example a commonly seen and commented upon perspective distortion will be convergence and divergence where vertical parallel lines either converge or diverge depending on how the camera/lens is held.
Some lenses can account for this type of distortion which are called perspective control lenses .. usually seen in the form of tilt and shift(the shift is what can account for convergence and divergence.

Other commonly seen, but less commented on distortion types are called compression and extension.
The longer the focal length the more the image is compressed. This information is commonly used to factor in the correct lens being used for portraiture, where longer focal lengths are preferred over shorter focal lengths. A longer focal length compresses the image more, and a wider focal length extends it more .. compression and extension.

You commonly see compression as a distortion of reality if you shoot at say 300mm over a longer distance and as long as the far distance is clear enough to be legible in the frame, it looks as tho the far distance is not as far as it really is, if there is some subject matter in the closer or middle distances. Compression.
Used to good effect by people that use long focal length lenses(telephoto lenses)

Extension is the opposite distortion effect, and you will see this via the use of short focal length lenses, to capture a wider perspective.
The difference in perceived distance between the foreground and middle distance and deep background look vast by comparison to normal lenses(say in the 35-70mm range), and hugely dramatic if compared to a telephoto lens.
This is used too good effect by people that use short focal length lenses(Ultra Wide Angle lenses).

So how does this make a difference to the topic of formats, lenses and equivalence?

it should be fairly obvious, but it may need to be explained anyhow:

Using a 35mm lens on a cropped format camera will obviously cause an amount of extension compared to using a 52.5mm lens on a 35mm camera format.
The perceived similarity of total FOV between the two different approaches may look the same to the casual observer tho.

And in opposition, using a 52.5mm lens on a 35mm format camera will concomitantly provide some compression if compared to a 35mm lens on a cropped camera.
This is a product of the magnification property of the lenses .. because a 50mm lens magnifies more so than a 35mm lens does(at distances).
As said before, magnification can be made equal between different focal lengths, but this is dabbling into macro photography realms.
So for equal subject distances, the focal length of the lens will magnify a scene differently.

I have a quick couple of images I took to show how the same image is actually ever so slightly different.
In the real world those differences may not be much to the 'average Josephine', but in some aspects of photography it can cause a major difference, which may result in huge issues!

Two sample images, posted large:

102828102829

Two images seemingly the same looking. Viewing them like this tho and you miss the point.
Better option is to download them and view them via your preferred software, and view them overlaid onto each other.
Most software has a feature whereby you can flick between the two images with your keyboard arrow keys, and the displayed image simply replaces the previous one.
For this Fastone's FSViewer is about as good as it gets. No clutter and extremely quick. I use ViewNX2 to do this too.

You will see the differences as you flick between the two images.

The orange level in the immediate foreground doesn't change by much between the two frames. Maybe a couple of percent difference in terms of framing this way or that.
The Eclipse bottle in the middle distance(which is actually close in) doesn't change by much either. It may look it if you flick from frame to frame, but in the liveview screen on the camera, I used a loupe to set the size of the bottle to be equal in both the Dx frame and Fx frame, as well as the orange level in the foreground.

That is, subject size was the attempt to keep constant here(this is the focus point too).
The background extends and compresses depending on the frame you are looking at.

The difference is not large, but it is a difference, and this is the key factor for some of us lens equivalence nazi's(note I'm included in this group) for responding to some posts on the topic of lens equivalence and format differences.

A lens is a lens is a lens, and the format makes no difference to the lens's properties. The lens does! :p

There's so much more to add to this post and thread, but I gott'a scoot and play Mr Mum for a bit.
I'll try to revisit later today if possible but this weekend will be busy so I may not come back for a few days after that.

Note too tho that DOF in the image(whilst not scientifically done!) is about two stops difference too. Exif is intact in both images and the obvious cue to differentiate the full frame image is the massive vignetting at the lower corners!
Note too that camera did not move between frames. Set on tripod and lens was zoomed between focal lengths and captured in both their respective format modes(Dx and Fx). Lens had to be refocused on the subject between zooms.
The not so 1.5x difference between the two focal lengths is most likely due to differences in focal length reporting due to lens design.

Lenses are not exactly as most manufacturers state. That is your 24-70, may actually be a 26-65mm or something. So the 62mm focal length reported by the lens may be something else in reality.
As already said, info about this exists eleswhere, but a lenses focal length is an approximation and quoted within a set of tolerances allowed to the manufacturer.
I think I vaguely remember that the Nikon 50/1.2 is actually 52mm in focal length(as measured at infinity).

All questions and related info is welcomed and encouraged .. remembering that no question is silly or dumb, and no info is unacceptable. Any incorrect info will be addressed accordingly tho.

arthurking83
22-08-2013, 6:14pm
I've added an embedded gif image for those that would like to see the subtle differences, but don't want or need to download the images.


http://cdn.makeagif.com/media/8-22-2013/Ap52UE.gif

Tannin
22-08-2013, 6:25pm
Cheers Arthur. I have long regarded this - http://www.josephjamesphotography.com/equivalence/ - as the touchstone article on this subject. So far as I know, it still is.

arthurking83
22-08-2013, 6:48pm
So far, as can be seen in the very small differences between the two images from the different formats, for most of us these differences are fairly small and probably not going to make or break a decision either for or against a format.

Nor is it completely untrue that a Xmm lens on APS-C looks similar to how a X/1.5mm lens on full frame .. but it is important to note that the focal length of the lens is still the focal length of the lens.

Although, for the easiest method to explain how the formats will impact the lenses, it's not too far off the mark to describe the use of say a 35mm lens on a cropped body as equivalent to a 50mm lens on a full frame camera.

Where this lax use of parallelism becomes an issue is for example in the macro photography world.
Other examples of photography genres that will become confusing could also be in forensic photography, where a measure of scale in the image could (most likely would!) be an important issue .. similarly to macro photography.

This is where strict and correct usage of lens focal lengths should be used. ie. in that a 100mm lens was used on a APS-C camera to catalog the rare and almost extinct phatobi.
If this convention is broken and the cataloger of the image used a 150mm lens on an 135 format camera(or vice versa) then the sense of scale is lost if the incorrect details are supplied with the image.

How?(and or why):

If the image is shot with a 100mm lens on a Dx camera, for the phatobi to fill the frame in such as way then the phatobi will have a specific dimension set.
If the cataloger has mis reported the catalog image as shot with an 'equivalent' lens of 150mm, which then the observer then assumes a 135 format camera was used, filling the frame in the same manner, the size of the phatobi can be calculated (say) for scientific purposes but the observer will be wrong in their calculations.

if the phatobi fills the frame perfectly from end to end(horizontally), then the observer could conclude that the phatobi will be so big relative to the 36mm width of the full frame 135 sensor used to catalog the phatobi.
Unknown to the observer tho is that the cataloger has mis reported the image of the phatobi to have been captured with a 150mm equivalent lens.
As the APS-C format is only 24mm across from one side to the other, you can easily see how in a scientific sense, the size of the phatobi will be 1.5x larger than it actually is.

This is easy to do as well as an experiment.
take out your macro lens. mount it too your full frame camera, shoot a ruler at 1:1 magnification and the distance scale on your ruler will show approx 36mm.
Do the same on an APS-C camera, and it will only show 24(ish) mm of the distance scale .. depending on your APS-C format.

Actually the way macro photography works, the actual focal length of the lens has no bearing on the magnification anyhow. The focal length of the lens will only really affect working distance, but again a digression of topics.

So you could imagine how important it is for this topic(of forensics, or scientific photography) for actual lens and camera specs must be reported correctly and maintained to a standard definition.

equivalence has no place in these fields of photography!

I think to myself why should it be any different for any other photography genre .. it only seems to cause confusion anyhow, as it is with many people coming in from the cold with confusing info about their 50mm lens being an 80mm lens and so on .....

I suppose the question is, do we or should we really care about this equivalence issue anyhow. A 35mm lens is a 35mm lens, and there's not much you can do about it. On a 135 format camera it gives a slightly wide perspective(with an amount of wide angle distortion), yet on an APS-C camera it gives a narrower perspective, with slightly less distortion but not entirely! Only the very distorted outer edges of the frame will be cut off from what the lens can actually see.

ROA44
24-08-2013, 8:35am
thank you Arthur for the time and effort you put into topics such as this I know I certainly appreciate it.

So is it correct to say that the focal length often referred to is not the best and FOV and Perspective would be better.

Also if you set up a FF & Crop camera with the same lens taking the same image off the same positioned tripod what differences could one expect to notice.

arthurking83
24-08-2013, 5:34pm
.....

Also if you set up a FF & Crop camera with the same lens taking the same image off the same positioned tripod what differences could one expect to notice.

My two sample images were made exactly in this manner.

The idea was to use the zoom capability of a 24-70mm lens at focal lengths that are usually in a range where distortion was minimal ... ie. 35-50mm range.
The shots were captured 35mm in crop mode(full frame Nikons have various crop modes, one of which is Dx format .. and the other image was then zoomed in to a longer focal length to mimic this apparent equivalence factor.

Theoretically, with an approx 1.5x crop factor, the focal length of the full frame image should have been about 52mm, yet the Tamron lens had to be set to 62mm.
This could have been an optical anomaly design of the Tamron lens too tho.

So, technically speaking, this is about as close as I could get to approximating the two different formats and the equivalence factor.
I don't have access to two prime lenses that can achieve the same so called FOV, from the same viewpoint .. but even if I did, I suspect that the images would look approximately similar anyhow.

Thanks for the added info to Tony too.
I've bookmarked the link to read in further depth when I get some time. Had a quick browse the other day, and there was way too much calculus looking stuff that made my head explode! :p

enseth
24-08-2013, 6:56pm
Thanks Arthur for pointing us towards FSViewer. Great program. I have been looking for something this good for ages. And the price is good too!