Why are so many of you on 1.6 crop cameras?

HoppyUK said:
That goldfinch is a great example of the practical advantages of crop. And a very convincing case for live view focusing. I need to try that :)

How did you support the camera using live view Tim, with all that focal length on board? Thanks.
Click to expand...

Manfrotto 055MF3 tripod and Manfrotto 468MGRC2 head.

Here's the full image, resized to 800x533....

20090321_145844_5549_LR.jpg


Here's a 100% crop....

20090321_145844_5549_LR.jpg


If only I'd had a 600/f4 (£7,340) instead of a poxy zoom and teleconverter. Then I might have done the camera some justice.....

By the way, viewed at normal levels of enlargement (7"x5" print from the full, uncropped image, viewed at 12" away) the DOF on this would have been no more than 4cm. Viewed at 100%, as this is, the DOF is around 0.8cm, which is simply not enough to get the bill and the eye in sharp focus at the same time.
 
Thanks Tim. Nice head :) I would like a Wimberly gimbal but can't justify it, so was going to get the Manfrotto mecano jobbie, but that nice hydro ball of your looks nice too. Is that your head of choice for this kind of thing?

Just a thought, but do you think you ran out of depth of field here? I agree that the bill isn't quite pin, but looking at the wire there looks like there might just be enough DoF. In which case it's subject movement. I mention this because I've got quite a few pics of birds with nice sharp bodies and blurred heads - they move some damn quick!

On the depth of field thing, it is perhaps relevant to note in this thread that this is the only area where crop will always be different to full frame. Crop gives more depth of field, f/number for f/number, and the more you crop it, as here, the more DoF you get and the more benefit there is to cramming more pixels into a smaller area.
 
tdodd said:
I have a 30D, 40D, 50D and 1D3. On Saturday I went birding and I took the 50D and 1D3 together with my 100-400 and 1.4X teleconverter. I left home with the lens on the 1D3 but when I got to the bird reserve I saw there was nothing large enough or close enough for the 1D3 to make a worthwhile shot, and little going on in the way of action to demand the 1D3 AF performance.

Without taking one shot with the 1D3 I swapped the lens over and tried for some smaller, perched birds with the 50D. That still wasn't enough so I added the teleconverter.

Here is a sample from the day, taken with the 50D at 560mm, f/8 (wide open), 1/640, 400 ISO. This has been cropped to 1/4 of the frame and then resized to 50%. By using the 50D instead of the 1D3 I placed 2.27X as many pixels over my subject, thus picking up detail the 1D3 would have missed entirely. To get a similar sized image from the 1D3 I would have had to crop and resize to 100%, and it would still have been a bit smaller.

3372735537_9a489a6a2f_o.jpg


It's a shame it isn't a little sharper, although this has had no edits other than the crop and some output sharpening on resize and conversion to JPEG. I guess I could have tweaked my capture sharpening first. I should also have stopped down a bit. You can see how narrow the DOF is from the wire upon which the bird is perched. Then there is the wide open softness of the zoom to consider as well. I could have traded 1/3 stop or 2/3 stop of aperture for a little reduction in shutter speed. This was focused using Live View Contrast Detect AF, a feature which the 1D3 does not even have.

There is clearly room for different sensor sizes and pixel densities. Higher densities give you more detail, at the expense of noise. Lower densities give less noise but less detail as well. Larger sensors require longer glass to create an image of equal relative size on the sensor and to gain any value at all from the larger sensor. That costs big bucks for the camera and bigger bucks for the glass.

There is also extra bulk and weight to consider. When I go on vacation, with photography in mind, I like to take a couple of bodies - one to have long glass, like my 100-400. for wildlife - and one to have a wide angle for landscapes or a medium zoom for people etc.. The extra bulk and weight of the 1D3 is significant and for long range/cropping the low pixel density is an issue, compared to the 50D. For such trips the 1D3 will stay at home and I'll take my 40D and 50D.
Click to expand...

Hallelujah! (y)
 
HoppyUK said:
I would like a Wimberly gimbal but can't justify it, so was going to get the Manfrotto mecano jobbie, but that nice hydro ball of your looks nice too. Is that your head of choice for this kind of thing?
Click to expand...
It's my only head, so it's not like I have an alternative :)

If by "Mecano" you mean the Manfrotto 393 gimbal head then by all accounts that is a very good alternative to a Wimberley gimbal at far lower cost. There are also the Indian knock-offs on eBay, which seem to get pretty positive comments that I've seen on POTN....

http://search.ebay.co.uk/search/sea...action=compare&copagenum=1&coentrypage=search

http://photography-on-the.net/forum/showthread.php?t=664600&highlight=gimbal

http://photography-on-the.net/forum/showthread.php?t=484987&highlight=gimbal

Back to my photo, at 1/640 it's true that head movement could be an issue, albeit slim. This is one out of several shots and the bird was perched there for a while and really not moving about hugely. The figures calculated here - http://www.dofmaster.com/dofjs.html - supports my belief that the DOF was wafer thin and could not accommodate everything, based on an estimated subject distance of 7 meters. The focus on the wire pretty much confirms how thin it was. 8mm DOF is really not much, even on a bird of this size. Then there is the issue of the 100-400 at 400mm and wide open, with a teleconverter. That is not a recipe for a razor sharp image.
 
Ooo! A Side-Kick copy for £75! Got to have one of those I think. Cheers Tim.

In think that you maybe did run out of DoF on the beak, but I reckon it is at least 3-4cm and not as narrow as you have calculated. Assuming your format was effectively reduced to about 10mm x 7mm in the first cropped image, this is what the DoF calculator suggests. (I just put in 4/3rds format and guessed a bit more.) Of course the 100% image is more testing.
 
HoppyUK said:
Ooo! A Side-Kick copy for £75! Got to have one of those I think. Cheers Tim.

In think that you maybe did run out of DoF on the beak, but I reckon it is at least 3-4cm and not as narrow as you have calculated. Assuming your format was effectively reduced to about 10mm x 7mm in the first cropped image, this is what the DoF calculator suggests. (I just put in 4/3rds format and guessed a bit more.) Of course the 100% image is more testing.
Click to expand...

DOF works in a linear fashion with respect to image magnification. The more you magnify the image the shallower your DOF becomes. Remember that DOF is only a zone of acceptable sharpness. The more you magnify an image the more you magnify softness in the image. There is only one plane of absolute sharp focus (ignoring field curvature). If you double the image size you double the softness at places other than the exact focal plane, and thus you halve the DOF. DOF calculations apply for an image rendered at a particular size (normally around 7"x5" to 10"x8" when viewed from 10" to 12" away. I don't know the exact figures but that's a good ballpark.

When I view a 50D image at 100% on *my* monitor (133 PPI display) that is equivalent to displaying the whole uncropped image at a size of 4752/133 = 36" x 24". That's on *my* monitor, which has a pretty high pixel density (1920x1200 17" screen). On a lower density monitor the same image viewed at 100% might be more like 48"x32". Well that's around 5X as large (or larger) as a regular 10x8 or 7x5 print. Thus the 4cm that was calculated for an image viewed at "normal" size and distance becomes 4cm/5 = 0.8cm when viewed at 100% on a monitor from 12" away. Now the original image I posted was at only 50% so it would follow that the DOF is around 1.6cm in that version of the image.

Regardless of the calculations, the falloff of DOF is visible on the wire and to my eyes it looks like about the only area of the wire that is sharp is right at the feet of the bird. It's hard to see how far behind the left foot the sharpness is maintained but to my eyes it is not much at all. 1.6cm for that image size seems about right to me. How big is a goldfinch foot?
 
Sorry if I've misunderstood you Tim, but for depth of field, what happens in practise is that final output size (print or VDU) doesn't make much difference, whereas changing image format makes a big difference. What you have done by cropping heavily on the bird is effectively to reduce your camera format, and in doing so increased your depth of field. Depth of field calcs are confusing as doing this upsets all the other parameters so it's hard to get the right answer.

I think it is accepted that full frame delivers narrower depth of field than crop (1.28 stops less in the case of Canon 1.6x) and that with the tiny formats of compact cameras there is so much depth of field it's almost impossible to get anything out of focus! It is logical therefore that when you crop an image still further to an effectively smaller format, depth of field goes up.
 
I'm afraid, Hoppy, that I think you are wrong. Plug in numbers to the DOF calculator for a 400mm lens at f/5.6 and 1,000cm. Then change the format from 40D to 1D3 to 1Ds3 and you will see the DOF increase as the format size goes up, from 12.9cm to 15.6cm to 20.4cm respectively. Better yet, here are the figures....

MWSnap%202009-03-23%2C%2018_41_03.jpg


MWSnap%202009-03-23%2C%2018_41_13.jpg


MWSnap%202009-03-23%2C%2018_41_24.jpg


Note that the difference in DOF between the 50D and the 1Ds3 is a factor of almost exactly 1.6X (20.4/12.8 = 1.58). In other words, the DOF varies in direct relation to the size of the captured image. It follows, that if I started with a 50D image and then cropped it to 1/4 the area - i.e. halved the linear dimensions, that my DOF was actually half that stated for a 50D without cropping. Of course, that would be for an image reproduced at "normal" viewing sizes from "normal" viewing distance.

The size of the sensor has got little to do with anything in many respects. That's just a convenient starting point to begin the calculations. What matters in the final/actual calculation is how large the captured image is, physically, before you enlarge it to your print size. Cropping a full frame image to the same size as a crop camera will give an identical DOF to an image shot with a cropper in the first place. The calculations certainly don't care what size your sensor was to begin with - what matters is what size your image was (i.e. after cropping) just before you magnified it to create your enlarged print.

The reason the people think the DOF reduces as you increase sensor size is that you have to increase focal length in order to frame the scene identically. e.g. to shoot with a 50mm lens on a 1.6X cropper you'd need an 80cm lens to shoot the same image/scene/composition from the same distance on a full frame camera. Plug in numbers for a 50mm lens on a 50D at 100cm and f/5.6 and then do the same for a 1Ds3 with an 80cm lens, also at 100cm and f/5.6, and then you get a DOF of 8.18cm for the cropper and a DOF of just 4.88cm on the full frame camera. That's the relationship people normally understand. Here are the calculated results....

MWSnap%202009-03-23%2C%2018_45_04.jpg


MWSnap%202009-03-23%2C%2018_45_18.jpg


The reason that the DOF is actually greater on full frame, for the same focal length, aperture and subject distance is because you don't need to magnify a full frame mage as much to reach your final size (e.g. 30"x20") compared with the magnification you need to get a crop sensor image up to the same size. The larger sensor needs less magnification so the DOF is greater. Cropping the image, as I did, did make my image even smaller, thus needing more magnification than the whole image, but it did not alter the focal length I used, or the aperture, or the subject distance. Thus reducing the image size by cropping also reduced the DOF.


Consider this example - you have a picture printed at 10x8. From your viewing distance (let's assume 12") it looks sharp so long as the blur width of any edge does not exceed 0.1mm (invented figure for the sake of example). Anything less than that will look like a sharply defined edge. Anything greater than 0.1mm looks soft and blurry. Now suppose you want that image printed at 20x16 instead, but you still want to view it from the same distance. Now, any edge that did have a blur width of 0.1mm will instead have a blur width of 0.2mm it will look soft. Anything that previously had a blur with of 0.05mm will now have a blur width of 0.1mm and will continue to look sharp. But basically, by doubling the magnification of the original image you have doubled the edge blur at every point within the scene. Some things that did look sharp at 10x8 will no longer do so at 20x16 - unless you double your viewing distance. The fact is that the final physical magnification at which you view an image makes a big difference to DOF calculations. That's why people who view 15MP or 21MP images at 100% on their screen and complain of soft images are perhaps missing the technical aspects of optical physics. As I said, a 50D image viewed at 100% on my monitor is equivalent to a virtual image size of 36"x24". Viewed on my HDTV at 100% that would be equivalent to blowing the image up to 88"x59". Given people tend to view their monitors from fairly close distances, that is an enormous level of magnification compared to more conventional printed image sizes, for which the typically published DOF calculators were designed to provide figures.
 
Tim, only just seen you post. Thanks for that - and I'm rushing out. Maybe I've got this around my neck - it is confusing when you swap all these different parameters about! Back later :)

Edit: Back now. Apologies Tim - I got things back to front. Thanks for troubling to post that detailed reply :)

Cheers.
 
HoppyUK said:
You mean the comparisons between the 30D and 1D cameras posted many weeks ago? What do you mean by a 1D? If you refer to the original 1D camera of 2001, it was only 4mp on a 1.3x crop. I don't see the relevance :thinking:

To compare like with like as far as we reasonably can, today you should compare something like a 50D and a 5D2.
Click to expand...

No... the 1dmk2N as clearly labelled in the first link of that post ;)

Both cameras 8mp, actually the 1D very slightly higher mp. The 1D's image falls apart on fine detail as soon as image cropping comes in to play.
 
To answer the original question, because I'm poor.

Full frame would be lovely but the body alone would cost more than I've ever spent on all my gear put together.

The crop factor isn't really an issue for me personally anyway as I rarely go wide enough to warrant a full frame body.
 
cracking write ups so far

just wondering "tdodd" even though i'm sure the results will be respective, you're using 400mm in you're calculations (whereas you've said you've used a 400mm + 1.4x TC = 560mm)

just pointing that out, thanks for all the info though :)

drew
 
buckas said:
cracking write ups so far

just wondering "tdodd" even though i'm sure the results will be respective, you're using 400mm in you're calculations (whereas you've said you've used a 400mm + 1.4x TC = 560mm)

just pointing that out, thanks for all the info though :)

drew
Click to expand...
The example I used in post #288 is just a random choice and not intended to be related to my goldfinch image in any way.

You will find the approximate DOF I quoted for the goldfinch (4cm at normal viewing size) is based on a focal length of 560mm at f/8 and 7m, guesstimated from the DOF for a 600mm lens at DOFMaster, since there is no 560mm lens listed....

20090616_151254_0009_LR.jpg
 
HoppyUK said:
On the depth of field thing, it is perhaps relevant to note in this thread that this is the only area where crop will always be different to full frame. Crop gives more depth of field, f/number for f/number.
Click to expand...

In answer to the OP's question, this for me was the strongest reason for choosing the 5. I just love the whole background blur thing so FF was a must. Although, with F1.8 glass it can be a royal pain in the backside.
 
for me it's about budget.
cropped cameras are cheaper on the canon side of things
also you get a bit more out of your lenses if you are on a budget like me, as the cropped sensors use the centre "sweet" part of the glass so I hear.
would love a 5D but for that I'd better upgrade much of my glass too, to get the most out of the body. not an option!
 
I shoot for magazines and unless we spend £1000s on changing printees and buying in better paper stock, there's little in the way of need for a fancy-Dan 24mp full-frame camera - the quality just won't be apparent, which is why I've consoled myself to using a crop camera. I also get more from my leong lenses, which is a big thing for me.
 
I use my D200 simply because it suits my way of shooting and I love it the only way I would change would be to move onto the D2X or Xs as I live in the real word and simply couldnt justify a D3 as all my dosh gets eaten by the moths in my wallet
 
More to the point, why don't Nikon & Canon bother making crop sensor camera's? is it purely the cost of making them?

Given the choice I would go for a FF but the cost is daft.
 
Dal said:
More to the point, why don't Nikon & Canon bother making crop sensor camera's? is it purely the cost of making them?

Given the choice I would go for a FF but the cost is daft.
Click to expand...

You'll have to explain that one. Your D50 is a crop sensor camera isn't it? :thinking:
 
Dal said:
More to the point, why don't Nikon & Canon bother making crop sensor camera's? is it purely the cost of making them?
Given the choice I would go for a FF but the cost is daft.
Click to expand...

Dal, just for reference - Crop sensor bodies ;)

EOS 300/350/400/450/500/10/20/30/40/50D
Nikon D40/50/60/70/70s/80/90/100/200/300/5000
All Olympus, Pentax and Lumix - Sony does 'em too (A900 is FF though)
 
Dal said:
More to the point, why don't Nikon & Canon bother making crop sensor camera's? is it purely the cost of making them?

Given the choice I would go for a FF but the cost is daft.
Click to expand...
Errrm... they do...!!

Confused about that post.:thinking:
 
It could be the sensor Dal was reffering to as I am sure but could be wrong ...didnt sony make the sensors for Nikon or used to, I know the D1 sensor was sony
 
Hi
In reference to your question i use full frame, 1.3 crop & 1.6 crop sensors in canon cameras the reason i use these is a matter of subject distance and being able to use a faster shutter speed when needed ie 600mm F4 i.s. on full frame 780mm f4 on 1.3 crop & 960mm F4 on 1.6 crop sensor without using teleconvertors as pixel count grows with developement i may change my mind but at the moment i think the separate sensors win over cropping images especialy with wildlife and bird photography
Regards
Lost
 
Don't know if someone has mentioned this yet (it's a long thhread !) but many people prefer 1.6 ratio's because it is a more aesthetic frame shape. The 1.6 ratio is called the fibonacci ratio, or golden rectangle. It's worth a google (and also gives you more options composition wise rather than just the default 'rule of thirds')
 
Huujuu said:
Don't know if someone has mentioned this yet (it's a long thhread !) but many people prefer 1.6 ratio's because it is a more aesthetic frame shape. The 1.6 ratio is called the fibonacci ratio, or golden rectangle. It's worth a google (and also gives you more options composition wise rather than just the default 'rule of thirds')
Click to expand...

All DSLR sensors are 3:2 ratio - full frame, 1.6x crop, 1.5x crop. Only 4/3rds system is different, eg 4:3.
 
kardo said:
The D300 has a sony sensor
Click to expand...
& the D40/D40X/D60/D3X.
Chances are that some of the others are fabbed by Sony as well.

In reply to the OP it's primarily because of cost & also for some (tele shooters) the crop factor is a benefit.
 
Hi again
Another reason i use crop sensors is down to frames per second on full frame canon cameras 5d mkii & 1Ds mkiii you get 3.9 & 5 frames with 1d mkii & mkiii you get 8.5 & 10 frames a second with a 40d & 50d just over 6 frames a second. For a bird in flight more frames a second is very useful to get wings in exact position you want also the same applies with sport when the action is quickly unfolding in front of you.
Regards
Lost
 
You must log in or register to reply here.
Back
Top