Crop vs FF for Landscape

existent

existent

Messages
1,417
Name
Esen
Edit My Images
Yes
Is FF really much better than crop?

I am not talking about focal lengths as that's not really an issue with 10-22 on crop vs 16-35 on FF. But I hear about the Dynamic Range being greater on FF which should help with landscape shots.

so when you compare, for example, a 7D and a 5D, how big is the difference in IQ for landscape shots?
 
Unless you're really into pixel-peeping, you'll not notice a great deal of difference between either format.

I've owned a 5D and currently own a 7D and have shot numerous landscapes on both and if I'm honest, I haven't noticed much if any difference between the two.

Go with whatever you're most comfortable with. Most people don't tend to print a great number of huge pictures and if you only post online, you shouldn't be seeing any appreciable difference. :)

Si
 
existent said:
Is FF really much better than crop?

I am not talking about focal lengths as that's not really an issue with 10-22 on crop vs 16-35 on FF. But I hear about the Dynamic Range being greater on FF which should help with landscape shots.

so when you compare, for example, a 7D and a 5D, how big is the difference in IQ for landscape shots?
Click to expand...

Per pixel, the 5D is going to be better - depends how big you want to print as to whether this makes any difference. Normally cropping is not so much of an issue with landscapes - it's much easier to compose the shot at time of shooting. The DR will be better on the 5D, but not by a massive amount. You will probably be shooting on a tripod at the base ISO. dxomark have the DR difference at the base ISO of only 0.2 of a stop!

However, you can always get a lot more DR with HDR - exposure bracketing and tone mapping. The results can look perfectly natural - HDR doesn't have to look wacky! You can do that with your 450D and get way more DR than the 5D mk2 can get from a single exposure. HDR is obviously great for landscapes as they don't tend to move...

Andy
 
Like Si said it's more down to what you're comfortable shooting with and what kit you already own. Pointless buying a full frame camera body if all your lenses are crop fits! I used to own a crop body and a Sigma 10-20 but about 15 months ago got a 5D and 24-105mm lens and what and how I'm shooting hasn't changed a great deal! The only thing that has really changed is that the 5D is a lot more flexible in allowing me to make up for a talentless photographer on the other end ;)
 
From my limited experience dynamic range is only very slightly better on 1Ds2 vs 40D (12bit FF vs 14bit APS-C). DxO mark sensor shows similar data. However 1Ds2 is clearly better at ISO1600 and still very good at 3200 when 40D can only take mobile phone quality snaps. Up to ISO800 40D is very clean, and you need only ISO 100 and a good tripod for landscapes. 5D mk2 looks like the best one (but note AF issues), while 7D apparently has more noise even at ISO100, and lower colour sensitivity. 7D will also require very sharp glass, possibly all primes for the best results. With FF most L or equivalent zooms are good enough.
 
5d over 7d, we've been playing and the 7d is awesome at high iso but v poor at low iso (relative mind).

Brilliant camera for some stuff just not really landscapes
 
I did notice a dynamic range improvement when I got the 5D (mk1) but my main motivation was just that it felt right. Seemed right coming from a film SLR...

Landscape is more about the photographer than the camera :)
 
If you print big, or tend to crop quite hard, or pixel peep, or like really shallow depth of field, then full frame is considerably better.

But if none of those things is important, you'll never notice the difference, and crop has its own advantages - mainly long lens reach, and price.
 
Thanks for the replies everyone. Pretty much what I hoped to hear.

I used 7D as an example because that was the latest crop model (other than 550D) and not really familiar with nikon or sony models as well.

I am not planning on upgrading yet as I am happy with 450D and rather build up my lens collection for a while, but it's always good to have an idea on which way to focus. Until now I didn't really consider FF as it is beyond what I do (unless I win some amount from lottery :lol:).



AndyElliott said:
However, you can always get a lot more DR with HDR - exposure bracketing and tone mapping. The results can look perfectly natural - HDR doesn't have to look wacky! You can do that with your 450D and get way more DR than the 5D mk2 can get from a single exposure. HDR is obviously great for landscapes as they don't tend to move...

Andy
Click to expand...

I do use HDR sometimes when it is hard to get everything correctly exposed, but I am one of those people that like to get it right out of the camera as much as possible :D
 
Interesting replies, when I look at landscape and portraits shot with a 5d mkII they seem much richer than from the 7d, I posted this question a few weeks ago and got completely opposite replies to this saying the dynamic range was really a deal breaker on the 5dMkII.

I'm interested too because I feel the need to upgrade to 5d in the future from my 7d
 
I am waiting to upgrade to FF anytime soon (waiting to see what Nikon do to replace the D700) either with the D700X or Canon 5DMkii.

The main advantage for me shooting landcsapes is the option to shoot wide with pro grade glass. Something I believe you can't do on a cropped sensor.

Rgds
 
joescrivens said:
Interesting replies, when I look at landscape and portraits shot with a 5d mkII they seem much richer than from the 7d, I posted this question a few weeks ago and got completely opposite replies to this saying the dynamic range was really a deal breaker on the 5dMkII.

I'm interested too because I feel the need to upgrade to 5d in the future from my 7d
Click to expand...

Looking at the dxomark tests (which only deal with the sensor - not with a lens attached) - the DR is better on the 5DmkII, but this is only a big difference at higher ISOs:

http://www.dxomark.com/index.php/en...329|0/(onglet)/0/(brand)/Canon/(brand2)/Canon

Even at higher ISOs, apparently the difference is only 1 stop. The tonal range and colour sensitivity are better for the 5D. It's a judgement call - the 5D will be better, but whether it's worth it given you would have to give up using your EF-S lenses, only you can decide.

Andy
 
mark_farrell said:
It's not just the dynamic range that you see with the 5DII, but what you can pull out using lightroom etc afterwards. Seemingly blown skies can usually be easily recovered, whereas on my old 450D, they would simply have been a white mess.
Click to expand...

that's an interesting point... is it simply because you get more data from bigger pixels?
 
existent said:
that's an interesting point... is it simply because you get more data from bigger pixels?
Click to expand...

I think that sort of thing is more down to the shape of the tone curve in the highlights area - if it flattens off not too quickly, rather than going straight to 100% of the sensor output, then it is easier to recover.
 
No idea why to be honest, and as I never had the two at the same time, I didn't take any direct comparison shots. But after getting used to the ND 'filter' in LR just pulling down a grey line on overexposed skies, it seems to recover more detail on similar shots on the 5DII. Obviously I can't directly compare like for like as sky and light may have been different, but it's certainly something I've noticed, and something my Nikonian friend also noticed when they went to a D700 from a D200
 
There are some excellent landscape photographers both on here and on Flickr... many of whom are using crop sensors. I think the skills of the actual photographer are far more important than the full-frame vs. crop issue. To prove a point, go and take a look at Scott's (digifrog) photostream. I can't see how his seascapes can get much better unless he went medium-format. :)

Even when I had a 5D, I can't say that my own work was any better for it. The quality of the lenses is far more important than sensor size in my opinion.

Cheers,
Si
 
Interesting thread this.....
I think there's more to it than dynamic range, diffraction and noise; I'm not sure what it is other than a textural richness to the images that you don't get on some models.
I'm not even sure this is a FF v crop difference.

Have a look at the 7D v 5DII comparisons in this review.
Astonishing!

I'm sure I've seen this in practice too......
Shooting waterfalls next to a 40D user and I had loads of texture in my whites from the moving water, whereas his were almost featureless.
This was so blatant we could see it full frame on the back of the cameras, no need to zoom in.
We checked our settings and found they were identical.
The non-whites were exposed the same, the only difference we could see on the back of our cameras was a huge difference in the whites.
 
I've got 30 x 40 inch landscape print on my table taken with my 40d, and it looks spot on to me.
 
DuncanDisorderly said:
Rich,
I don't doubt your photographic prowess for a moment :)

This threads about crop v FF,
but you are right, the photographer is way more important...
Click to expand...

:)

My point was that unless you are printing enormous prints crop frame does OK.

Incidentally I've also just got some nice landscape prints shot with my 4/3rds GF1.
 
Owning both full frame and crop sensor bodies I have to say that for landscape a full frame bodywith a 16-35 lens hammers into the dust any crop sensor body on a 10-22

I love it when asp-c users look through the viewfinder of my full frame kit with a 16-35 on the front and and go wow. They almost seem to fall in to the camera. This tells me that they have either forgotten about 35mm or (worse) have never experienced a proper 35mm film camera and therefore can be described as a digi-noob™

On the other hand I prefer to use a 70-200 on a crop sensor.

Horses for courses.

"My point was that unless you are printing enormous prints crop frame does OK."

What's the point of struggling to that location and waiting until you're under under that light and merely doing OK ?

If you can't afford medium or large format then it's safe to say that full frame is de rigeur for landscape.
 
Tyrone Shoelaces said:
What's the point of struggling to that location and waiting until you're under under that light and merely doing OK ?

If you can't afford medium or large format then it's safe to say that full frame is de rigeur for landscape.
Click to expand...

OK= Undiscernable by most.
 
Tyrone Shoelaces said:
...I love it when asp-c users look through the viewfinder of my full frame kit with a 16-35 on the front and and go wow...
Click to expand...

Hmmm... I love it when people look at some of my landscapes taken on my APS-C body and go 'wow'! Most of 'em don't give a toss what it was taken with as long as it looks nice! :)

I used to own a 5D with a 17-40mm F4L so I'm as aware of the view from the 5D as anyone else. I got rid of it in favour of the 7D and my EF-S 10-22mm and apart from having a markedly more bowed horizon on my full-frame landscapes, there's little to choose between any of 'em. :)

APS-C, full frame, four thirds, 1/1.7in sensor... doesn't matter a damn if you can't take a decent image! Flickr is full of AGNI's... All the Gear, No Idea!

Cheers,
Si
 
I'm with Rich on this too.....
I define acceptable image quality as being good enough to print A3 and view at arms length.
You can get away with murder.
In the past I've printed cracking landscapes from a 4Mp point and Shoot.
With my 5D2 I can get acceptable images in far more challenging conditions with plenty of leeway to allow manipulation and cropping in PP.

Pixel peeping on the other hand....
Forget all rules of thumb like hyperfocal distance and 1 / focal length - they will always look soft.
 
DuncanDisorderly said:
...I define acceptable image quality as being good enough to print A3 and view at arms length.
You can get away with murder...
Click to expand...

:agree:

Just a thought...

Doesn't having a full frame sensor in order to be able to crop heavily kind of defeat the object of having it though? ;)

I'm not disputing that in terms of overall nit-picking IQ, the full frame isn't a better prospect; I'm just pointing out that one of the regular responses from FF users on here seems to be about how much they can crop!

Cheers,
Si
 
DuncanDisorderly said:
I'm with Rich on this too.....
I define acceptable image quality as being good enough to print A3 and view at arms length.
You can get away with murder.
In the past I've printed cracking landscapes from a 4Mp point and Shoot.
With my 5D2 I can get acceptable images in far more challenging conditions with plenty of leeway to allow manipulation and cropping in PP.

Pixel peeping on the other hand....
Forget all rules of thumb like hyperfocal distance and 1 / focal length - they will always look soft.
Click to expand...

If you look at the calculations to make 'Depth of Field' there are various assumptions base on print size a viewing distance.

The usual benchmarks use a Circle of Confusion diameter of 30 microns (0.00003 meters) on a 35mm negative, based on what can be distinguished on an 8 x 12 inch print. The size of the circle of confusion varies directly with the size of the film or sensor (since images from a smaller format need to be magnified more to create the same size print). Digital cameras tend to have smaller image sensors and will have a smaller circle of confusion. Medium and large format film cameras will have a larger circle of confusion.

But obviously cropping or going for print sizes way different from the 8x12" print make the DOF calculation meaningless. What is true is, up to a point, larger enlargements will appear to have a more shallow DOF and vice versa.

Andy
 
Spiritflier said:
Doesn't having a full frame sensor in order to be able to crop heavily kind of defeat the object of having it though? ;)

I'm not disputing that in terms of overall nit-picking IQ, the full frame isn't a better prospect; I'm just pointing out that one of the regular responses from FF users on here seems to be about how much they can crop!

Cheers,
Si
Click to expand...

LOL - you may be right :D
Although I'm not sure this is a benefit of FF or simply having lots of pixels to play with.

With my 20D I felt I could only get away with the lightest of crops so always tried to get it right in-camera.
I still don't crop very often.
However, correcting my wonky horizons without significantly degrading IQ is something I rely on far too often ;)
 
Spiritflier said:
:agree:

Just a thought...

Doesn't having a full frame sensor in order to be able to crop heavily kind of defeat the object of having it though? ;)

I'm not disputing that in terms of overall nit-picking IQ, the full frame isn't a better prospect; I'm just pointing out that one of the regular responses from FF users on here seems to be about how much they can crop!

Cheers,
Si
Click to expand...

Cropping is a bad idea. No question. But I will admit that it's one of the reasons I've just swapped to full frame.

I shoot a lot of portraits and while I always try to fill the frame, I often find that there's a shot in there that looks better cropped quite hard, and in a way that I didn't see at the time, or didn't dare zoom in to, for whatever reason.

And there have been other times when shooting two children together that one of them has a great expression that can be pulled out as a good solo portrait. I've done that a few times with my 40D, but it only makes a small print before it looks sub standard.

Here's another idea, though I've not tried it myself. I've seen portrait photographers shoot on medium format film, on a tripod, with quite a wide field of view. They then don't need to even look through the viewfinder to know that they've got the shot, even if the subject moves quite a bit. In fact, one technique is to wander around chatting to the subject, firing by remote release, waiting for just the right relaxed pose and expression. You can then crop any framing you like afterwards, with ample quality. David Bailey does it.

Edit: Duncan, heavy cropping has got more to do with basic big sensor area than number of pixels. Ultimately it's about the number of photons you can collect. For example, 7D vs 5D2, similar number of pixels, 5D2 blitzes it on big enlargements/crops. Although I would agree than below A4 it is very hard to see the difference. Some compacts have 14mp but they can't hold a candle to larger sensor cameras.
 
AndyElliott said:
If you look at the calculations to make 'Depth of Field' there are various assumptions base on print size a viewing distance.

The usual benchmarks use a Circle of Confusion diameter of 30 microns (0.00003 meters) on a 35mm negative, based on what can be distinguished on an 8 x 12 inch print. The size of the circle of confusion varies directly with the size of the film or sensor (since images from a smaller format need to be magnified more to create the same size print). Digital cameras tend to have smaller image sensors and will have a smaller circle of confusion. Medium and large format film cameras will have a larger circle of confusion.

But obviously cropping or going for print sizes way different from the 8x12" print make the DOF calculation meaningless. What is true is, up to a point, larger enlargements will appear to have a more shallow DOF and vice versa.

Andy
Click to expand...

Andy,
I'm not going to disagree with you other that to point out that the CoC tables indicate that a crop sensor will max out a 8"x10" print at f16.
But that doesn't deter us from printing bigger than 8x10 or using f22 when we need to.

I think it's about acceptable sharpness versus being properly sharp.
and there is nothing wrong with an acceptably sharp print.

Since you have brought up this subject and back on topic....
That table indicates to print 8"x10" the maximum aperture on a crop sensor is f16, while FF can get away with f22 :D:D:D
The table conveniently explains why P&S cameras never go over f8 and large format can use stupid f numbers.
 
Tyrone Shoelaces said:
I love it when asp-c users look through the viewfinder of my full frame kit with a 16-35 on the front and and go wow. They almost seem to fall in to the camera. This tells me that they have either forgotten about 35mm or (worse) have never experienced a proper 35mm film camera and therefore can be described as a digi-noob™
Click to expand...

Being a digi-noob (:D), why does 16mm on FF looks better than a 10mm on a crop? It surely is similar field of view. But as you said, I never looked through the viewfinder of a FF, so I can only go by what people say about it :).

DuncanDisorderly said:
Andy,
I'm not going to disagree with you other that to point out that the CoC tables indicate that a crop sensor will max out a 8"x10" print at f16.
But that doesn't deter us from printing bigger than 8x10 or using f22 when we need to.

I think it's about acceptable sharpness versus being properly sharp.
and there is nothing wrong with an acceptably sharp print.

Since you have brought up this subject and back on topic....
That table indicates to print 8"x10" the maximum aperture on a crop sensor is f16, while FF can get away with f22 :D:D:D
The table conveniently explains why P&S cameras never go over f8 and large format can use stupid f numbers.
Click to expand...

When I first got my slr, I used to shoot at f16, and f22 all the time thinking that's what I needed to shoot at to get everything sharp. But as that link suggests you start to get diffraction after f16 (even at f16 with some lenses).

But, since we are talking landscape, most of the time, the lens will be a UWA, or WA, which already has very large dof even at larger apertures. I now shoot at f8 or f11 mostly (unless I am trying to do long exposures), and everything seems to be sharp in the pictures.

So I don't really see f16 vs f22 for crop vs FF such a big issue.
 
existent said:
Being a digi-noob (:D), why does 16mm on FF looks better than a 10mm on a crop? It surely is similar field of view. But as you said, I never looked through the viewfinder of a FF, so I can only go by what people say about it :)
Click to expand...

If you mean it looks better just through a FF viewfinder, it does a bit because it's bigger, and brighter at f/2.8 on a 16-35L. Field of view is the same. It's just better in the same way that a bigger, brighter telly looks better than a smaller one.

When I first got my slr, I used to shoot at f16, and f22 all the time thinking that's what I needed to shoot at to get everything sharp. But as that link suggests you start to get diffraction after f16 (even at f16 with some lenses).
Click to expand...

Diffraction is present at all apertures, limiting sharpness. It's only when the increasing effect of diffraction overtakes the improvement in other aberrations caused by using a higher f/number, that the net result is an overall deterioration.

The better the lens is corrected for those other aberrations, the earlier diffraction can be detected, albeit at a higher level. In practise, the best lenses on crop format cameras deliver peak sharpness around f/5.6, and at f/8 on full frame.

However, we're talking about a very high level of sharpness there and while it's downhill after that, I was always quite happy with the sharpness I got at f/11 on my 40D crop. At f/16 it was going off noticeably, and at f/22 it was pretty horrible TBH.

But, since we are talking landscape, most of the time, the lens will be a UWA, or WA, which already has very large dof even at larger apertures. I now shoot at f8 or f11 mostly (unless I am trying to do long exposures), and everything seems to be sharp in the pictures.

So I don't really see f16 vs f22 for crop vs FF such a big issue.
Click to expand...

I don't think it's a big issue either, certainly very easy to avoid. What's more significant I think, is what you can do at the other end of the scale, ie the ability to create more shallow depth of field with low f/numbers on full frame compared to crop. If you want those really low f/number big bokeh effects, full frame has an advantage there.
 
HoppyUK said:
Diffraction is present at all apertures, limiting sharpness. It's only when the increasing effect of diffraction overtakes the improvement in other aberrations caused by using a higher f/number, that the net result is an overall deterioration.

The better the lens is corrected for those other aberrations, the earlier diffraction can be detected, albeit at a higher level. In practise, the best lenses on crop format cameras deliver peak sharpness around f/5.6, and at f/8 on full frame.

However, we're talking about a very high level of sharpness there and while it's downhill after that, I was always quite happy with the sharpness I got at f/11 on my 40D crop. At f/16 it was going off noticeably, and at f/22 it was pretty horrible TBH.
Click to expand...

I didn't know that you would get the sharpest image at f/5.6 for crop. But the DOF may not cover everything at that aperture, so as you said, you are better of stopping down to f/8 or f/11.
 
existent said:
I didn't know that you would get the sharpest image at f/5.6 for crop. But the DOF may not cover everything at that aperture, so as you said, you are better of stopping down to f/8 or f/11.
Click to expand...

It depends on the lens. The ideal (diffraction limited) lens would be sharpest wide open, hence some good teles are really excellent at f/2.8 and get only slightly better at f/5.6.
DOF is key for landscapes. f/11-16 still retains good sharpness on APS-C, and past f/18 everything suddenly becomes very soft. FF can handle f/22, but it also requires longer lenses (shallower DOF) for the same FOV. On average the results are very similar in print up to A3 or even A2.
Good corners performance is very important and is just as likely to ruin the photo. 17-40L has some trouble at 17mm, and so does 16-35mm at 16mm.
 
Three reasons I prefer FF to crop for landscapes
1) Coming from a film background, it feels more natural, and I find it easier to pre-visualise scenes dependant upon the lens I'm using, always found that more difficult on a crop camera.
2) I can shoot at a true 12mm, crop cameras can't go this wide.
3) I frequently let f22 take the strain, crop cameras really fall off a lot sooner than FF.
 
When people step down for dof they often go too far. It's worth keeping in mind what dof you need and what settings obtain it and not using using a needlessly small aperture.

Regarding focal length, it's also worth remembering that longer lenses do not give you less dof. If the image is the same size lenses with different lengths give the same dof. By zooming in all you are doing is making the out of focus area more obvious.
 
It's getting a bit confusing now.. :D

I know that dof is smaller on full frame. Is it one stop difference?

So lets take 10mm (crop) / 16mm (ff) for example.

10mm crop fov will be equivalent to 16mm on FF.

Now, what about dof.. Is the same f-stop going to give the same dof? or do you need smaller aperture (higher f number)?
 
You must log in or register to reply here.
Back
Top