Full frame - not used one but I don't get it

JohnN

Suspended / Banned
Messages
6,359
Name
John
Edit My Images
Yes
Hi,

With the rumors of the 1D V popping up all over it again made me start thinking about full frame cameras.

Now of course I get that it gives a wider image, but most of the time I want to be closer to the subject anyway and should I have need of a wider angle I could get a 10-20mm or similar (in fact I had one but didn't use it enough to keep), so I ask you good people what am I missing about the full frame that makes it worth not only loosing reach but also cost so much more?

Ta.
 
Last edited:
Well, Canon 1D series is 1.3x crop (1Ds is full frame) and the 10-20 lenses won't fit, or won't work properly if they do.

But that aside, the thing about full frame is image quality. The sensor is bigger, like almost twice the size of 1.3x and more than twice the size of 1.6x (1.4x would be exactly half the area). More photons collected for cleaner colours, less noise/higher ISO, and because the lenses don't have to work so hard, better sharpness. Apart from that, and not everybody needs that level of IQ, full frame is pretty much all downsides - much more expensive, bigger/heavier, especially the lenses for equivalent framing with long focal lengths.

Edit: don't let high pixel counts fool you. More pixels is good, but bigger pixels is better.
 
Last edited:
Loads of 'full-frame' threads on here already, but to summarise: they're expensive because the sensor (the most expensive part of the camera) is bigger, in fact as big as it's possible to be in a 35mm-based DSLR - 36 x 24mm, the same size and shape as a frame of 35mm film.

Without going into the technical reasons why, a bigger sensor means better high ISO performance and overall image quality.

Full-frame DSLRs are physically bigger than models with smaller sensors, which might suit you better if you have large hands.

As well as the camera itself being more expensive, a full-frame sensor needs physically larger lenses with more glass, which tends to get very costly when you're looking at telephoto lenses, 200mm and above.

As you've pointed out, the bigger the sensor, the wider a given lens effectively is, and also the less depth of field it has at a given aperture. (To illustrate the latter point, my other half's Canon G9 has a tiny sensor by DSLR standards. Even though it has selectable apertures all the way to f/2.8, in practice they make very little difference to the depth of field - everything in the shot is usually sharp, front to back. On my Canon 5D mk II with its full-frame sensor, you have to make a conscious effort if you want to achieve front-to-back sharpness, i.e. selecting f/8 or higher and careful focusing. Focusing on something in the foreground will invariably result in an out-of-focus background, and the depth of field you get at large apertures - f/2.8 and below - is very shallow.)

It's interesting that we already have a generation of photographers who have only ever used digital compacts or basic DSLR's with small sensors. Back in the days of film (not so long ago!), everyone with a 35mm camera shot "full frame". As an ex-film user myself, the main attraction to me of paying the extra for a full-frame DSLR is that it feels and behaves just like a film camera. I've got used to, say, a 50mm lens having a certain field of view, and I wouln't want to stick it on a DSLR and find that it's now behaving effectively like a 65mm or 80mm lens. Of course, if you need the extra reach, then magnification factor you get with a smaller sensor will work in your favour.

A.
 
Last edited:
...should I have need of a wider angle I could get a 10-20mm or similar (in fact I had one but didn't use it enough to keep), so I ask you good people what am I missing about the full frame that makes it worth not only loosing reach but also cost so much more?

The wide angle thing... A 10-20mm on APS-C will look something like 16-32mm on full frame, (it won't fit,) but you can get a 12-24mm for full frame... but then again you can get a 8-16mm for APS-C so I suppose you can get a just about equally wide image from either full frame and APS-C.

All in all I'm not really too sure what full frame offers the amateur photographer.

I've done test shots with my 20D, 5D and G1 (so that's APS-C, FF and MFT) and to be honest there's next to nothing in it at lower ISO's. The 20D and slightly more so the 5D give marginally better higher ISO performance and the margins increase as the ISO rises but at normal print sizes and viewing distances other than arguing about noise levels I'd argue that the differences are really rather marginal.

Re DoF. The smaller sensor cameras do effectively give deeper DoF as they generally have different lenses and you generally take the shot from a different camera to subject distance and as creative DoF is a nice thing to have available I personally would not want to use anything smaller than MFT, which I think gives pretty good DoF options especially when mated to a nice f0.95 lens.
 
Last edited:
[..]

Of course, if you need the extra reach, then magnification factor you get with a smaller sensor will work in your favour.

A.

Without wishing to cause trouble, it's a crop factor not a magnification factor. You can get the same view by cropping a full-frame image.

Anthony.
 
Without wishing to cause trouble, it's a crop factor not a magnification factor. You can get the same view by cropping a full-frame image.
Precisely. A 200mm lens on a full frame and crop will give the same image when the ff image is cropped to the same field of view as the crop.

What changes is solely the field of view. When people say a lens is equivalent to a lens 1.6x as long on a full frame, they are saying (although they may not realise it) that the field of view is equivalent. Any extra "reach" people perceive is obtained through smaller pixels and the ability to crop harder that gives you. Whether that gives you anything in reality (as no matter how many pixels you have, the sensor area taken up by an object through a given lens is exactly the same on any format camera) is another discussion.
 
Hmn, thank you all and very interesting.

One day I may try one to see exactly how much better the IQ is and how DoF is effected, but with the mark V processor coming with its much better ISO control and presumably less noise at lower ISO settings it may knock out one of the three advantages.
 
When the 5DII was new (a long time ago in a land far far away) I conducted a test comparing the 20D with the 5DII.
It turns out that their pixel density is almost exactly the same, so comparing a 20D frame with a cropped 5DII image will give the same number of pixels and the same field of view.
So they are directly comparable....

The results were interesting!
At ISO 100 there was only a difference in fine detail when viewed at 100%; the 5DII was better, but not by much, and not enough to affect an A3 print.
Differences are explainable by factors like improvements in technology of anti-aliasing filter and IR filter.

However... The 5DII image was cropped to less than half the available pixels and at anything other than ISO 100 the difference was so huge that there was no meaningful comparison.
I didn't compare retention of texture in the highlights (e.g. milky water); but my impression is that the 5DII is vastly superior. Other people have also noted the 5DII's exceptional fine detail, such as this detailed comparison of the 7D with the 5DII.
 
When the 5DII was new (a long time ago in a land far far away) I conducted a test comparing the 20D with the 5DII.
It turns out that their pixel density is almost exactly the same, so comparing a 20D frame with a cropped 5DII image will give the same number of pixels and the same field of view.
So they are directly comparable....

However... The 5DII image was cropped to less than half the available pixels and at anything other than ISO 100 the difference was so huge that there was no meaningful comparison.

Interesting that you found that as I've done the same comparison and at low to mid ISO I didn't find the difference huge at all. At low to mid ISO I found lens differences far easier to spot.

It makes sense to crop the full frame image to compare it to a crop APS-C image and this should be a perfectly fair comparison as cropping an image can never ever reduce its quality, only its size, and I'm therefore a little confused that you seem to be saying that at 100% there's just a shade in it in favour of the 5DII but full image difference is so huge there's no comparison. I'll have to have a think about that one as from an image quality point of view rather than field of view and other things point of view it just doesn't make sense to me.
 
Alan,
There are two parts to my response and I think you may have merged them...
Probably my fault, at times my English is a tad ropey.

For the same lens, same focal length and same aperture, used on both the 20D and 5DII.

1) Viewed at 100% and ISO 100, the 5DII has slightly more fine detail; but the difference is very subtle and won't affect an A3 print.

2) Viewed at 100% and higher ISOs, the 5DII is significantly better.

Hope it is clearer this time :)
 
Hmn, thank you all and very interesting.

One day I may try one to see exactly how much better the IQ is and how DoF is effected, but with the mark V processor coming with its much better ISO control and presumably less noise at lower ISO settings it may knock out one of the three advantages.

FF will still have an advantage, when the Mark V processor is applied. Bigger simply means better in terms of image quality. It was always thus.

It's less about lots of pixels, and more about photon capture and lens MTF characteristics. Bigger formats are just better at all these things.
 
Alan,
There are two parts to my response and I think you may have merged them...
Probably my fault, at times my English is a tad ropey.

For the same lens, same focal length and same aperture, used on both the 20D and 5DII.

1) Viewed at 100% and ISO 100, the 5DII has slightly more fine detail; but the difference is very subtle and won't affect an A3 print.

2) Viewed at 100% and higher ISOs, the 5DII is significantly better.

Hope it is clearer this time :)

I see. :) Personally I'm perfectly happy to use my 20D at ISO 800 without worry beyond that things may very well be different. I don't have easy access to a 5DII but past comparisons didn't show a world of difference but we were using different lenses which I'm sure helped the 20D and hindered the 5DII (20D+50-mm f1.4 seemed to compare very well against 5DII+24-70 f2.8, much to the distress of the 5DII owner :D.) With my 5D I find that there's quite a bit of noise at 1600/3200 but they look better than the 20D shots after NR but in reality to me they don't look night and day different at normal print sizes.
 
When the 5DII was new (a long time ago in a land far far away) I conducted a test comparing the 20D with the 5DII.
It turns out that their pixel density is almost exactly the same, so comparing a 20D frame with a cropped 5DII image will give the same number of pixels and the same field of view.
So they are directly comparable....

The results were interesting!
At ISO 100 there was only a difference in fine detail when viewed at 100%; the 5DII was better, but not by much, and not enough to affect an A3 print.
Differences are explainable by factors like improvements in technology of anti-aliasing filter and IR filter.

However... The 5DII image was cropped to less than half the available pixels and at anything other than ISO 100 the difference was so huge that there was no meaningful comparison.
I didn't compare retention of texture in the highlights (e.g. milky water); but my impression is that the 5DII is vastly superior. Other people have also noted the 5DII's exceptional fine detail, such as this detailed comparison of the 7D with the 5DII.

Very interesting read and seeing the 5DII vs 7D comparison it stood out a mile, although on that test the 7D seemed much softer than I recall it being when I had one.
 
I see. :) Personally I'm perfectly happy to use my 20D at ISO 800 without worry beyond that things may very well be different. I don't have easy access to a 5DII but past comparisons didn't show a world of difference but we were using different lenses which I'm sure helped the 20D and hindered the 5DII (20D+50-mm f1.4 seemed to compare very well against 5DII+24-70 f2.8, much to the distress of the 5DII owner :D.) With my 5D I find that there's quite a bit of noise at 1600/3200 but they look better than the 20D shots after NR but in reality to me they don't look night and day different at normal print sizes.

Alan, the link posted by Duncan Disorderly (this one http://rolandlim.wordpress.com/2009/11/06/canon-eos-7d-review/ ) has a good illustration of the 5D2 vs 7D, and as the OP notes above, the 5D2 is much better.

It's about half way down that very long page, headed 'Comparison with 5D Mark II'. The author says he doesn't know why the difference is so apparent, but if you scroll further down several commentors have explained why, and it's simply down to lens MTF (all other things being more or less equal).

The thing we call 'sharpness' is actually a combination of resolution (the fineness of detail) and contrast (how clearly those details are shown) which is the basis of Modulation Transfer Function in a lens. It's a fact of physics that as resolution goes up, contrast goes down, and since the 7D's sensor is less than half the size of the 5D2 with roughly the same number of pixels, it's asking the lens to work more than twice as hard for the same resolution. Which it can't do without losing contrast, and that's the clearly visible difference. You can see it in an A4 print easily enough.
 
Just for fun and I know it's next to impossible at these sizes anyway... and these have been strangled by photobucket...

IMG_4342.jpg


_1000704rc.jpg


_1010270.jpg


IMG_7909.jpg


... but on my screen and in print I can't honestly see any significant difference between MFT, APS-C and FF at low to medium ISO :D
 
Last edited:
Just for fun and I know it's next to impossible at these sizes anyway... and these have been strangled by photobucket...

<snip>

... but on my screen and in print I can't honestly see any significant difference between MFT, APS-C and FF at low to medium ISO :D

Those shots make a very good point. How much image quality do you actually need?

If you shoot at medium ISOs and only output smallish (say A4 max) or only on screen, as many people do, then you're unlikely to see much benefit from full frame.

It's only when you start making big canvases, or shooting high ISO in poor light, or need to crop hard that FF starts to really shine, which is why wedding photographers favour it.

I've never tried a 7D or processed any shots myself and without wishing to cause offence to owners I've often been underwhelmed by 7D shots I've viewed on line :D

Yes! Which supports my other point - it's much more about the size of the pixels rather than the sheer number of them.

The reason I ended up with a 5D2 is I went to buy a 7D and tested it against my 40D. And the 7D was better, but not by much at all. Then I tried a 5D2 simply because Jessops had one sitting there, and was amazed at how much better it was. That turned out to be quite an expensive little test, but the difference is that while the two cameras have roughly the same number of pixels (7D 18mp vs 5D2 21mp) the 5D2's pixels are twice the size. So it's capturing twice as many photons and the lens is working at half the resolution so contrast is correspondingly higher (inherant lens contrast is not the same stuff as raising the contrast in post processing, which is artificial). The result is sharper pictures, with less noise and cleaner colours :thumbs:
 
Last edited:
...in poor light....

Is there any other type in the UK ;)

Funny you should mention wedding photography, I commented only ten minutes ago to my wife after looking through the long review that is why they always seem to have full frames at weddings. Poor light + no flash = high ISO
 
...The result is sharper pictures, with less noise and cleaner colours :thumbs:

The holy grail!

By the way how is the autofocus speed on the 5DII?

Not that I can afford it but were I to get a second body I could consider FF, possibly the 5DIII when that comes out (or 7D II) - I've got a saving fund going.
 
thats good not much noise visible noise at that res, now don't you go trying to convert me to the dark side ;)
 
The holy grail!

By the way how is the autofocus speed on the 5DII?

Not that I can afford it but were I to get a second body I could consider FF, possibly the 5DIII when that comes out (or 7D II) - I've got a saving fund going.

5D2 gets a bad press on AF, but there's nothing wrong with it at all. It's just that the 7D is exceptionally good after Canon threw the kitchen sink at it ;) IMHO a 7D and 5D2 make a fantastic pairing - 7D has speed and (pixel) reach for action and long lens stuff, 5D2 delivers fabulous IQ.

high iso will never beat decent light but its nice to have the backup heres one at iso 10000 from the d700.
straight from camera

<snip>

Makes the point again :D Nikon D700 has a feeble (;) haha) pixel count at a mere 13mp, but it's full frame and thrashes any crop format DSLR on everything, even those with more pixels, when it comes to image quality - sharpness, noise, colour, the lot.
 
Last edited:
Those shots make a very good point. How much image quality do you actually need?

Well, yes, obviously I agree. I don't tend to need fantastic IQ and I don't normally do very heavy crops (I had a quick look at that link and he's nit picking at 100% and the chances are that I'll do that once or twice a year) and I've never printed larger than A3.

I do see differences in contrast and saturation, and of course in noise too but I personally think that the comments about full frame that we've all read or even made ourselves from time to time about FF being night and day better and giving pop out 3D images are perhaps over egging things a little and leading perhaps to some having unrealistic expectations and disappointment.

As long as the framing and look can be achieved and noise can be controlled and equalised I personally find that slight differences in contrast and saturation and the like melt away in post capture processing and I personally tend to find it hard to remember what shot was taken with what camera as long as I can't see the file name.

One more thing. The first comparative test I did between my G1+14-42mm and 20D with 50mm f1.4 gave two lovely images and it was only upon close inspection that I could say with some confidence that the 20D was better but upon fitting a Voigtlander 25mm f0.95 to the G1 it's now a much closer match, proving to me that at low to middling ISO the lens is probably a bigger deciding factor in image quality than the sensor even if you want to look very closely.

_1000749.jpg


IMG_4430.jpg
 
Makes the point again :D Nikon D700 has a feeble (;) haha) pixel count at a mere 13mp, but it's full frame and thrashes any crop format DSLR on everything, even those with more pixels, when it comes to image quality - sharpness, noise, colour, the lot.

but don't let high pixel counts fool you(5d2):D. More pixels is good(5d2):D, but bigger pixels is better (d700):D:1st::D
 
scottthehat said:
heres a crop, come to the darkside you will like it.

Scott, can you stop banging on about your bloody D700....







...You're making me very jealous :lol:
 
but don't let high pixel counts fool you(5d2):D. More pixels is good(5d2):D, but bigger pixels is better (d700):D:1st::D

I thought I said that several times already, though it needs repeating ;)
 
Flash In The Pan said:
It's cheap enough that I'm sure you could come up a convincing argument for your bosses to buy it for you :lol:

.....hmmmm, ny recent budget proposal has been met with silence so I'm thinking that trying to convince them to buy me a memory card might be hard at the moment:lol:
 
Aaaand...

Couple of 7D shots..

6400 ISO


feeder by tonky8203, on Flickr

12800 ISO


cleaner by tonky8203, on Flickr

Shot in raw - saved as TIFFS - resized and saved as jpegs - no editing whatsoever.
 
CBA reading the whole thread sorry. :)

But just to weigh in anyway, I have a 5D mk I and a 7D - despite the 7D being the much more modern camera with many more features etc, I prefer using the 5D I by a long way. The images just look better, they have a certain 'magic' to them.
 
I suppose the question for me is could you get rid of the 7D altogether and use the 5D for everything, wildlife, macro, portrait etc etc, relying on improved quality and heavier cropping. That of course pretends that AF speed is equal (I know the 7D is very good, but I've a 60D which is still pretty good)
 
I suppose the question for me is could you get rid of the 7D altogether and use the 5D for everything, wildlife, macro, portrait etc etc, relying on improved quality and heavier cropping. That of course pretends that AF speed is equal (I know the 7D is very good, but I've a 60D which is still pretty good)

Not really - or all wildlife photographers would be using full frame, but the vast majority use crop bodies for the increased reach and cropping ability.

To be fair, as the game has moved on the goalposts are moving a bit with the overall general trend toward more pixels across all formats. The 1DMK4 got an unheard of pixel increase for the 1 series making it a very useable wildlife camera while not having the ultimate reach of a 7D if you want to sacrifice a bit of reach.

If the 1DS4 when it comes is around 30mp which seems likely, then it will definitely be worthy of serious consideration for someone who wants an all round camera including wildlife with a 50% crop still leaving a 15mp image .
 
Back
Top