- Messages
- 3,696
- Name
- Frank
- Edit My Images
- Yes
As the MkIV has a pixel density exactly the same as the 40D, I'd guess that the image size would be the same as the 40D.
Edit: Whoops, beaten by Tim. And his post is more informative!
Canon announces EOS 7D ***Official Discussion thread***
- Thread starter Admin
- Start date
As the MkIV has a pixel density exactly the same as the 40D, I'd guess that the image size would be the same as the 40D.
Edit: Whoops, beaten by Tim. And his post is more informative!
- Messages
- 7,179
- Name
- Edward Bray
- Edit My Images
- Yes
Is it not something to do with printing size ?
- Messages
- 3,696
- Name
- Frank
- Edit My Images
- Yes
Not a lot. It's a demonstration of that all-important, but oft-ignored, specification called pixels-per-bird (PBB). The more pixels you have covering your bird the more detail you'll have (assuming a perfect image from the lens).
CT
TPer Emeritus
- Messages
- 26,617
- Edit My Images
- Yes
Guys, the actual maximum size at which you can reproduce your bird is nothing to do with pixel count. It's purely to do with the maximum file size which the camera outputs. Regardless of sensor size, the bird will always be the same size on the sensor for a given camera to subject distance, but the larger the 1:1 file which the camera outputs, the larger will be the bird proportionally to that file size.
Where pixel density does count is in the resolution of the image, particularly when it comes to cropping.
Tim's had the advantage of doing 100% crops from cameras he owns, whereas I've had to provide my examples by adjusting image sizes to provide my chart. Assuning Tim's cropping is accurate, then it looks entirely possible that I've ballsed up in the case of the 1D3, but the fact remains that his other crops just confirm the accuracy of my assimilation.
Where pixel density does count is in the resolution of the image, particularly when it comes to cropping.
Tim's had the advantage of doing 100% crops from cameras he owns, whereas I've had to provide my examples by adjusting image sizes to provide my chart. Assuning Tim's cropping is accurate, then it looks entirely possible that I've ballsed up in the case of the 1D3, but the fact remains that his other crops just confirm the accuracy of my assimilation.
CT
TPer Emeritus
- Messages
- 26,617
- Edit My Images
- Yes
I love it when you talk technical Cedric
Touche'Ed we need the full max file size from the the 1DMK4 to do an approximation of where it sits in the cropping stakes.
CT
TPer Emeritus
- Messages
- 26,617
- Edit My Images
- Yes
Where does he say it's better than a 5DMK2? What he actually says is...
"It can be hard to deduce any conclusion from a bunch of low resolution jpegs on a blog, but this camera will be my main street/travel camera, there is no doubt about that. It produces lovely, very film like files, with loads of detail and tonal range; I just love the 'feel' of the files at 1600 iso - they make me feel nostalgic!!
I can't wait to get my hands on this camera and shoot some RAW files at a wedding. Hopefully that should happen within the next week or so. I can see this also becoming part of my normal wedding gear, for that extra reach in church, and also as a first rate back up camera. If you currently shoot with something like a 40D, you will definitely want this camera ;-)"
- Messages
- 6,729
- Name
- Jim
- Edit My Images
- Yes
Doesn't image size depend on the total number of pixels rather than the density? -
Edit - It's ok I worked it out
- Messages
- 6,481
- Name
- Tim
- Edit My Images
- No
Doesn't image size depend on the total number of pixels rather than the density? -
Edit - It's ok I worked it out
Well done
For those who haven't figured it out yet, here is an example....Suppose you have a 400mm prime lens and you are shooting a subject (bird if you like) at a distance and you can't get closer. Let's say that the image of the bird as captured on the sensor is 10mm high. It doesn't matter which camera you use, how many pixels it has or what size the sensor is. The bird will be 10mm high on the sensor.
Now, suppose you want to make a print from your image at 300 ppi. How big will the bird be in print? There is one formula that can be applied to all the cameras to work out the height of the printed bird. The formula is....
Printed Height (inches) = subject height(mm) / sensor height(mm) x sensor height(pixels) / 300(ppi)
If we feed that data into a sausage machine here are the results....
So, ranking these in order of subject size the 7D has a commanding lead, with the 50D not far behind, while the 1D3 is very much the tail end Charlie. The 20D, 30D, 5D2 and 1Ds3 all produce the exact same sized subject in print. The 40D and the 1D4 will also produce the exact same sized subject in print. Of course, that isn't the whole story because we still have to consider noise, dynamic range, colour quality etc., but in simple terms the size advantage of high pixel density sensors, especially if you are focal length limited, is clear.
Now, if you are not focal length limited and have the freedom to frame/compose how ever you wish then, very simply, the camera with the most total pixels (5D2 and 1Ds3) will produce the largest print of all the cameras. In the case of the full frame cameras there should be an extra bonus in IQ due to larger, cleaner individual pixels.
- Messages
- 3,696
- Name
- Frank
- Edit My Images
- Yes
Well done
Nice one Tim! And at 6 in the morning!
- Messages
- 6,481
- Name
- Tim
- Edit My Images
- No
grumpybadger
Alan Rickman
- Messages
- 4,638
- Name
- Paul Beastall
- Edit My Images
- Yes
Nice one Tim. I have a very similar spreadsheet from a few years back comparing 30D/1DII/5D/1DsII which also includes minimum aperture for AF, that helped lead me to a 1DsII at the time.
- Messages
- 23,200
- Name
- Richard
- Edit My Images
- No
Good stuff Tim
Mathematics is one thing, but I think there's a danger that people might think all you need to get extraordinary reach with small, distant subjects is a 7D.
It's important to be aware that simply planting a large number of pixels over the subject doesn't automatically give sharp photos. The lens has to deliver an extraordinary level of resolution and one of the optical facts of life is that when resolution goes up, image contrast goes down pretty much in proportion. Basically, you need a heck of a sharp lens for the image to hold up at what are very high levels of magnification.
Mathematics is one thing, but I think there's a danger that people might think all you need to get extraordinary reach with small, distant subjects is a 7D.
It's important to be aware that simply planting a large number of pixels over the subject doesn't automatically give sharp photos. The lens has to deliver an extraordinary level of resolution and one of the optical facts of life is that when resolution goes up, image contrast goes down pretty much in proportion. Basically, you need a heck of a sharp lens for the image to hold up at what are very high levels of magnification.
- Messages
- 6,481
- Name
- Tim
- Edit My Images
- No
Hoppy, I provided the maths just to demonstrate that CT's images, as helpful as they might be, are not actually correct in their relative sizes. I totally agree that there is a lot more to achieving worthwhile/maximum reach than simply sticking a 7D on the end of your lens. I've said it before with respect to the 50D and, with the 7D, the challenges are magnified further.
In particular, with ever increasing resolving power of the sensor there will be flaws exposed elsewhere in the image capture/processing sequence. The effects of camera shake will be magnified by the 7D. The blur of subject motion will also be magnified. Focus errors will be more evident, as will the impact of diffraction if stopping down too much. Lens aberrations or weaknesses will also be more easily revealed. Add in pixel level noise and there are plenty of reasons why high pixel density may buy you nothing useful at all and will simply fill up your cards and hard drive quicker while slowing down your workflow.
Despite my table showing the 50D to have a huge reach advantage over the 1D3, I'd still rather use the 1D3 for anything that moves with any sort of pace, or anything small (needing a heavy crop) at ISOs from 400 and upwards. The 50D is great when stuck on a tripod and used with low ISO and pixel perfect focusing, but when there is the slightest chance of blur/shake/noise/misfocus those pixels are potentially wasted. All they leave me with is larger files that capture the errors more accurately and must be downsized to yield acceptable IQ. 1D3 files are very useable at 100%. It's not as easy to say that about 50D files, and once 50D files are resized to, say, 50% you suddenly end up with the equivalent of a 3.75 megapixel file, which starts to look embarrassingly feeble when stacked against a 1D3's lush pixels and the reach advantage is very much reversed.
For all these reasons I am holding off on the 7D, although it is surely a tempting camera, but I have today added a 5D2 to my kit list as I think that will do far more to expand my photographic options than yet another 1.6X cropper.
p.s. I'm sure everyone understands the guideline of minimum shutter speeds to hand hold a lens of 1 / (focal length x crop factor). Well that might hold true for 10x8 print but for a 7D file viewed at 100% that would be equal to blowing up the image to 39"x26" on my monitor, or ~4X larger than a 10x8. Well for such files to look sharp at 100%, when shooting handheld, you'd need (theoretically) shutter speeds of at least 1 / (focal length x crop factor x 4) which for a 400mm lens would suggest a shutter speed of around 1/2500. In brilliant sunshine, at f/5.6, you'd need ~200 ISO to achieve that, which is easy enough, and with good IQ, but if the light drops even a little then you could well be at 400 ISO, 800 or even more. Well I've seen 100% crops from the 7D at 400 ISO and to my eyes they look no prettier than those from the 50D at 400 ISO, and absolutely not good enough to use at 100%, when striving for detail in a small subject like a bird in the distance. Personally I think that 18MP in a cropper is far more than is beneficial when shooting BIF in anything but superb light.
p.p.s That previous paragraph is, once again, basically maths, which may not tell the full story, especially in real world shooting, but again it is something to think about.
In particular, with ever increasing resolving power of the sensor there will be flaws exposed elsewhere in the image capture/processing sequence. The effects of camera shake will be magnified by the 7D. The blur of subject motion will also be magnified. Focus errors will be more evident, as will the impact of diffraction if stopping down too much. Lens aberrations or weaknesses will also be more easily revealed. Add in pixel level noise and there are plenty of reasons why high pixel density may buy you nothing useful at all and will simply fill up your cards and hard drive quicker while slowing down your workflow.
Despite my table showing the 50D to have a huge reach advantage over the 1D3, I'd still rather use the 1D3 for anything that moves with any sort of pace, or anything small (needing a heavy crop) at ISOs from 400 and upwards. The 50D is great when stuck on a tripod and used with low ISO and pixel perfect focusing, but when there is the slightest chance of blur/shake/noise/misfocus those pixels are potentially wasted. All they leave me with is larger files that capture the errors more accurately and must be downsized to yield acceptable IQ. 1D3 files are very useable at 100%. It's not as easy to say that about 50D files, and once 50D files are resized to, say, 50% you suddenly end up with the equivalent of a 3.75 megapixel file, which starts to look embarrassingly feeble when stacked against a 1D3's lush pixels and the reach advantage is very much reversed.
For all these reasons I am holding off on the 7D, although it is surely a tempting camera, but I have today added a 5D2 to my kit list as I think that will do far more to expand my photographic options than yet another 1.6X cropper.
p.s. I'm sure everyone understands the guideline of minimum shutter speeds to hand hold a lens of 1 / (focal length x crop factor). Well that might hold true for 10x8 print but for a 7D file viewed at 100% that would be equal to blowing up the image to 39"x26" on my monitor, or ~4X larger than a 10x8. Well for such files to look sharp at 100%, when shooting handheld, you'd need (theoretically) shutter speeds of at least 1 / (focal length x crop factor x 4) which for a 400mm lens would suggest a shutter speed of around 1/2500. In brilliant sunshine, at f/5.6, you'd need ~200 ISO to achieve that, which is easy enough, and with good IQ, but if the light drops even a little then you could well be at 400 ISO, 800 or even more. Well I've seen 100% crops from the 7D at 400 ISO and to my eyes they look no prettier than those from the 50D at 400 ISO, and absolutely not good enough to use at 100%, when striving for detail in a small subject like a bird in the distance. Personally I think that 18MP in a cropper is far more than is beneficial when shooting BIF in anything but superb light.
p.p.s That previous paragraph is, once again, basically maths, which may not tell the full story, especially in real world shooting, but again it is something to think about.
- Messages
- 23,200
- Name
- Richard
- Edit My Images
- No
Cheers Tim
I think we are both agreeing that the concept of 'digital zoom' is no free lunch and for sure a good big un will always win out. The appeal to me is that with the 7D my occasional forays into birding may be more successful. I just don't want all the impractical bulk of say a 1D4 with a massive prime, and I am certainly not prepared to pay for it just to get used a few times a year. That's what is most appealing to me about the 7D - the reduced size, weight and cost, but still doing a passable job (for what I want). I just need to see what it can do for my own eyes first. (But CT is doing a grand job )
I don't understand why you are saying subject blur is increased with cropping, or that focus errors will be more evident when depth of field is increased by cropping (effectively to a smaller format). Compared to fitting a longer lens I mean.
I think we are both agreeing that the concept of 'digital zoom' is no free lunch and for sure a good big un will always win out. The appeal to me is that with the 7D my occasional forays into birding may be more successful. I just don't want all the impractical bulk of say a 1D4 with a massive prime, and I am certainly not prepared to pay for it just to get used a few times a year. That's what is most appealing to me about the 7D - the reduced size, weight and cost, but still doing a passable job (for what I want). I just need to see what it can do for my own eyes first. (But CT is doing a grand job
) I don't understand why you are saying subject blur is increased with cropping, or that focus errors will be more evident when depth of field is increased by cropping (effectively to a smaller format). Compared to fitting a longer lens I mean.
- Messages
- 6,481
- Name
- Tim
- Edit My Images
- No
What I am saying is that as pixel density increases then the sensor's ability to see desirable detail also increases. However, so does its ability to see and record blur, shake, diffraction and any other undesirable details. So, if those "nasties" are large enough to cover more than one pixel the benefit of that high degree of accuity offered by the sensor disappears. So, you have greater leeway with a 1D3 and its large, widely spaced pixels than with the tightly packed and unforgiving pixels of the 7D. The 7D will visibly record shake, blur and so on tht would not be picked up by the 1D3 at all (not visibly at least). I don't see much value in 18 megapixels that more finely record soft focus and blur than 10 megapixels that "filter" out those problems and give you smaller files right off the bat.
In other words, a 7D will demand far more from a photographer, his/her technique, lenses and perhaps even the lighting conditions than a 1D3 would. If one of those things is not up to snuff then 18 megapixels won't do you much good. That is not to say that there aren't other advantages over xxD bodies, such as the new AF system, AF microadjustment and new sensor technology, but I'm not sure that 18MP are any better than 15MP or 10MP for action shooting where movement is inevitable and some degree of blur/shake a likely reality. Now for landscapes, and other more relaxed styles of photography that is a different story, especially when using a tripod, MLU, remote release and excellent glass stopped down just the right amount.
- Messages
- 6,481
- Name
- Tim
- Edit My Images
- No
Well in my opinion it would certainly generate smaller file sizes that will take up less card space and disk space and should process quicker. Also, with some of your pixels consolidated before they leave the camera it should make pixel peeping at 100% a less stressful experience as each pixel should look cleaner (less noise) and sharper. So in short I believe it would "help out". But of course, once you've downsized within the camera there is no going back afterwards, so be careful you don't throw baby out with the bathwater.
Also, if you are switching often between fast action and relatively still action - e.g. from BIF to perched birds and back again - it might be quite irritating to think to keep swapping back and forth between raw and mraw. However, for something like soccer, especially in less than perfect light, I would be really surprised if there is any point in having the camera churn out full raw files throughout a game. There are many things moving in different directions - arms, legs, the ball, other players, so it's not like perfect panning will eliminate all motion blur, so unless you can shoot at fast shutter speeds and at lowish ISOs I'd save the card space.
Like I say, this is just an opinion based on my 50D experience. Personally I do plug on shooting full raw so that no data gets thrown away until I've had a chance to review it and decide how much NR to apply, how much cropping, how much sharpening etc. etc..
- Messages
- 2,196
- Edit My Images
- Yes
Well, yes and no, I set my quick-button to churn out sraw for the next photo. So in theory, if you have very flexible fingers, you could just hit that each time. It saves space. Might be difficult when on H-speed though
grumpybadger
Alan Rickman
- Messages
- 4,638
- Name
- Paul Beastall
- Edit My Images
- Yes
Slight change of tack but thought it was worth posting that I've been playing with the remote flash stuff and its pretty good. Seems to work as well as the ST-E2 so that's one less thing to carry. Wonder how long it will take for the FF Canons to come with an integral flash to enable them to drive remotes?
Set it up with 550EX and 430EX with different ratio between the two and all controlled from the back of the camera - does what it says on the tin.
The only downside is that I can't find a way to tweak settings using the PC via USB, which would be really useful for what I am trying to do at the moment. Although a good thing is that I have found remote Liveview on the laptop really useful for setting the camera up too...
Set it up with 550EX and 430EX with different ratio between the two and all controlled from the back of the camera - does what it says on the tin.
The only downside is that I can't find a way to tweak settings using the PC via USB, which would be really useful for what I am trying to do at the moment. Although a good thing is that I have found remote Liveview on the laptop really useful for setting the camera up too...
- Messages
- 23,200
- Name
- Richard
- Edit My Images
- No
Well put Tim
One thing that I think will catch out people using the high cropping potential of these new cameras, is shutter speed. If you are using a 400mm lens on a 7D, the rule of thumbs states that 1/640sec should be your slowest possible shutter speed to avoid shake.
If you then crop the image to quarter the image area (same magnification as fitting a 2x tele-con and 800mm focal length equivalent) you need to double that again to 1/1280sec and I would say 1/2000 just to be sure, as a minimum.
Even then, camera shake does not go away - it is only the effect that is reduced to an acceptable level. Thanks heavens for IS I say (again).
- Messages
- 23,200
- Name
- Richard
- Edit My Images
- No
I'm not sure the 1-Series cameras will ever get an on-board flash. I think both Canon and Nikon take the view that if you are sufficiently committed to have a very top end camera, then you will have a decent flash to go with it.
Certainly I think the 580EX is better than either the ST-E2 (surely overdue for an update) or any on-board flash in one very important area - range. With the 580 you can both change the direction of the head and also zoom it to give the pre-flash maximum intensity.
grumpybadger
Alan Rickman
- Messages
- 4,638
- Name
- Paul Beastall
- Edit My Images
- Yes
i deliberately said FF and not 1-series (especially bearing in mind the bulk of 1-series sold are not FF ). Another string to the D700's bow is the inbuilt flash and ability to use it as a remote commander.
I have no experience of a 580 (my 550's are broken so haven't replaced them) but the 550 is better than the ST-E2. But, the problem is I rarely carry it with me and so, sometimes when detached flash would be nice, I'm stuffed. At least the 7D can give a useful fall back...
). Another string to the D700's bow is the inbuilt flash and ability to use it as a remote commander.I have no experience of a 580 (my 550's are broken so haven't replaced them) but the 550 is better than the ST-E2. But, the problem is I rarely carry it with me and so, sometimes when detached flash would be nice, I'm stuffed. At least the 7D can give a useful fall back...
- Messages
- 3,696
- Name
- Frank
- Edit My Images
- Yes
Or you could get exactly the same effect by re-sizing your 18MP image down to 10MP. Why one should want to do that I'm not sure?. A 10MP sesnor will not show more detail than an 18MP sensor. Card and disk space are pretty cheap nowadays, so I don't see that as being a problem either. And I don't think that processing time would be changed that much (not with a decent machine to do it on).
CT
TPer Emeritus
- Messages
- 26,617
- Edit My Images
- Yes
A 50D file cropped by 50% is a 21.5mb TIFF file.
I've picked this image because the light was really dire as you can probaby see, and it was taken at 800 ISO. It's straight from the RAW via a TIFF with no sharpening, noise reductuion or other processing at all.
Full frame
Some 100% crops from the image....
I venture to say that under the circumstances, the noise for 800 ISO is pretty good. In a 1:1 print from this file, the noise wouldn't even be visiblle at normal viewing distance, and looking for it would be the equivalent of sticking your nose down on a huge exhibiition print from film looking for grain.Having said that, it would be two minutes work to run noise reduction on this image and it would make a pretty impressive print in my opinion with minimal processing.
It's not difficult to achieve results like this from the 50D, in fact I could find many better examples taken under more favourable light, I just hope people aren't unduly influenced by some of the comments here now - if you're not getting a good percentage of usable 1:1 shots from your 50D, the fault doen't lie with Canon, it's in your technique or the gear you're using.
- Messages
- 6,481
- Name
- Tim
- Edit My Images
- No
When I see a blotchy blue sky at 400 ISO, when exposing correctly, then it figures that my subject is similarly affected by the blotchiness, even if it is not so readily visible. That means that fine detail is being obscured by the blotchiness. Now, when you go a good way towards filling your frame, as you have done in your example above, the noise is not that bad relative to the size of the details captured. But, if your subject is genuinely small in the frame then the details will be lost to the noise - even at 400 ISO. Yes, you can run NR and get a perfectly decent looking sky, but the bird will not be worth keeping.It's not difficult to achieve results like this from the 50D, in fact I could find many better examples taken under more favourable light, I just hope people aren't unduly influenced by some of the comments here now - if you're not getting a good percentage of usable 1:1 shots from your 50D, the fault doen't lie with Canon, it's in your technique or the gear you're using.
The fault with my technique in this situation would be my failure to fill the frame sufficiently in the first place. But I'm not talking about my technique. I'm talking about people in general relying on cameras like the 50D and 7D to allow massive cropping because they did not fill the frame enough. All I am saying is that massive cropping at higher ISOs is not going to yield stellar results. Yes, 400 ISO from a 50D viewed at 50% with a light dusting of NR should look good, but so would the output from a 1D3 at 400 ISO at 100%, giving the 1D3 more reach than the 50D in that situation.
By the way, you say you have used no NR in processing that image but the last time we discussed your non-use of NR it turned out that you were using NR all the time because DPP applies NR automatically unless you purposely disable it. So, not that it matters much, other than for accuracy, but I'd like to know whether your images really are being processed with no NR whatsoever, because to my eyes everything you post looks amazingly smooth for images without NR and I would love to know how you do it.
Here's an example from my 50D at 800 ISO, processed through Lightroom on default settings. The only edit was a WB tweak. Exposure, sharpening, NR and so on were unaltered....
Full frame :
100% crop :
I wonder why my background does not look as clean as yours.
Perhaps the flaw in my technique is that I use Lightroom instead of DPP. I've already seen comments that Lightroom 3 is doing a better job with 50D files than Lightroom 2.x.
Let's see how things look in DPP, firstly with NR sliders both set to zero :
and now with the NR set to default values (2,3) as chosen by DPP :
My background still doesn't seem as clean as yours, and with NR included the bird is looking quite soft, definitely lacking fine feather detail, so I guess my technique is pants. I'd be grateful for any pointers on how to match the noise levels you manage to achieve. I don't expect to match the sharpness of your 500/4 prime with my 100-400 but it would be great to see the low noise that you are achieving. How do you do it?
At 1600 ISO and with the subject a decent size in the frame, the full frame image is just about OK, when compressed down to an iddy biddy 800x533 file, but at 100% it is horrific and I really can't see that 15MP was any advantage at all because we did not capture more detail, simply more noise, perhaps contributing to the impression of "false" detail when downsized. Example, again with no edits....
Full frame :
100% crop :
Sure you can apply plenty of NR and get a smooth looking image, but (a) there is no detail in the bird even at 100%; (b) by the time you've applied NR and downscaled there will be even less. So how did having 15MP add any value in this shooting example? IMO all it did was to generate a huge file that needed to be downsized significantly in order to appear OK. Now this kind of thing is easier to do when shooting large, relatively featureless subjects, like motorsports, but not much help for bird feathers unless you can pretty much fill the frame in the first place.
- Messages
- 6,481
- Name
- Tim
- Edit My Images
- No
Regarding my comments about the difficulty of getting anything useful from 15MP or 18MP when shooting moving stuff, here is an example from my 1D3 at 400 ISO and with no edits whatsoever. The file is simply sharp and completely noise free....
Full frame :
100% crop :
I'm not sure the screen print has done it any favours but that is one very useable file at 100%.
Here is something I shot the day before, in very similar lighting and with similar exposure settings. This was with the 50D and I pushed the exposure way over to the right, clipping some of the lightest feathers of the bird, so in this example noise is not an issue. However, the chances of pixel perfect panning accuracy on a moving target are slim, when using high pixel density sensors, so this example is looking a little blurry and needs downsizing for acceptable IQ. Because I have pixel level blur the 15MP are of no use to me as they have not captured more useful detail, they have simply captured more blur.
100% view :
So, to be clear, I am not in any way saying that the 50D or 7D are bad cameras. I am simply saying that in less than optimum conditions I question the benefit of the high pixel density. Absolutely the reach advantage is real, with a good lens, good technique, static scene, low ISO, no blur, no shake and perfect focus but I doubt very much that the advantage remains when the action heats up and the ISO needs to be pushed up. That's just my opinion. Feel free to disagree.
What that means for me, in practical terms, is that each camera has its strengths and weaknesses. If I can outline some shooting scenarios here's how I might favour one over the other....
- Macro, still life, product, fine art : 50D > 1D3
- Landscapes and architecture : 50D > 1D3
- Motorsports and other action sports : 1D3 > 50D
- BIF and lively wildlife : 1D3 > 50D
- Static wildlife : 50D > 1D3 below 400 ISO; 50D = 1D3 at 400 ISO; 1D3 > 50D above 400 ISO
- Weddings : 1D3 > 50D indoors and outdoors
- Studio : 50D = 1D3
- Catwalk : 1D3 > 50D
- Indoors without supplementary (controlled) lighting : 1D3 > 50D
- Portraits and pets : 1D3
Sorry for labouring the 50D/1D3 comparison, since this is supposed to be a 7D thread, but my concerns regarding shake/blur when shooting stuff that moves applies even more to the 7D than the 50D. I just don't happen to have any 7D examples personally.
Full frame :
100% crop :
I'm not sure the screen print has done it any favours but that is one very useable file at 100%.
Here is something I shot the day before, in very similar lighting and with similar exposure settings. This was with the 50D and I pushed the exposure way over to the right, clipping some of the lightest feathers of the bird, so in this example noise is not an issue. However, the chances of pixel perfect panning accuracy on a moving target are slim, when using high pixel density sensors, so this example is looking a little blurry and needs downsizing for acceptable IQ. Because I have pixel level blur the 15MP are of no use to me as they have not captured more useful detail, they have simply captured more blur.
100% view :
So, to be clear, I am not in any way saying that the 50D or 7D are bad cameras. I am simply saying that in less than optimum conditions I question the benefit of the high pixel density. Absolutely the reach advantage is real, with a good lens, good technique, static scene, low ISO, no blur, no shake and perfect focus but I doubt very much that the advantage remains when the action heats up and the ISO needs to be pushed up. That's just my opinion. Feel free to disagree.
What that means for me, in practical terms, is that each camera has its strengths and weaknesses. If I can outline some shooting scenarios here's how I might favour one over the other....
- Macro, still life, product, fine art : 50D > 1D3
- Landscapes and architecture : 50D > 1D3
- Motorsports and other action sports : 1D3 > 50D
- BIF and lively wildlife : 1D3 > 50D
- Static wildlife : 50D > 1D3 below 400 ISO; 50D = 1D3 at 400 ISO; 1D3 > 50D above 400 ISO
- Weddings : 1D3 > 50D indoors and outdoors
- Studio : 50D = 1D3
- Catwalk : 1D3 > 50D
- Indoors without supplementary (controlled) lighting : 1D3 > 50D
- Portraits and pets : 1D3
Sorry for labouring the 50D/1D3 comparison, since this is supposed to be a 7D thread, but my concerns regarding shake/blur when shooting stuff that moves applies even more to the 7D than the 50D. I just don't happen to have any 7D examples personally.
CT
TPer Emeritus
- Messages
- 26,617
- Edit My Images
- Yes
Tim,
A few points in answer to your last post...
As we discussed last time - when I say no NR I mean none other than that which is being applied in DPP at minimum default levels. I never alter them, and if there's any appreciable effect on actual subject detail , then I can' t see it. I NEVER run NR on the whole image, I hate the loss of detail that ensues. When I run NR it's ouitside of DPP - in Paint Shop Pro 7 in fact, and only ever on the bg. With 800 pixel web images I rarely need to do that since the 50D and certainly noticeably less with the 70D. I'm truly mystified when you say you don't know how I do it, because I'm doing nothing clever or involved to my images at all. I'd go so far as to say I probably do less post processing than most people. I'm not trying to be a smart arse, and I certainly wouldn't deliberately mislead anyone over this, in fact I've explained time and again on the board how to run selective NR just on the bg.On the occasions I do so I'll always say so.
My workflow...
Invariably a RAW file and opened initially in DPP.
I very rarely do anything at all to the images in DPP, they mostly look great staright out of the tin. I never need to adjust WB. I sometimes may need to do slight highlight or shadow recovery, but it's quite unusual.
I then output a 16 bit TIFF file to PSP 7, (If I intend to use it immediately),where I resize and make any minor levels adjustments, sharpen for the downsize and save. I'd ony run NR on the BG and only if it was intrusive in the bg, which I'm increasingly finding it isn't, and I shoot at 800 ISO most of the time, and 1600 ISO quite often. A bird which has intrusive noise in the feather detail woiuld be a bin job, unless it was a Golden Oriel, when I might make an exception.
With regard to your two sample crops, I know you're trying to make a point, come on though - you just wouldn't crop either of those images that tightly surely, but yes- the noise is bad. Here's an edit with NR run just on the bg, to show you could improve the BG noise at least .I wouldn't even try to run NR on the bird itself,you'd just lose feather detail.
This is trying to address noise which is really a bridge too far in this instance though, and it isn't the root cause of your problem anyway - It's pretty obvoius, you need to either...
(a) Get closer.( I hate how glib that sounds, because I know all too well, it's often just not possible), OR
(b) Invest in some longer glass if you're seriously interested in bird photography, and I imagine you are.
Seriously mate ,the last thing you need at the moment to pursue bird photography is yet another camera body, especially one which is only going to compound your problems at that range. You need to invest in some longer glass if you're serious about bird photography - all birders eventually do! If it's just a mild interest and you don't want to invest in the glass, then fair enough, but there are lots of people on TPF posting some pretty stunning images with the 50D- you'd have to be blind not to see them. Yes - the 50D and cameras like it are extreme, and you need the lenses to extract the best from them, but decrying the camera when you're seriously under-gunned for the job you're trying to do is unfair and misleading to people considering buying the camera. The Robin shot I posted was with the 300mm 2.8 btw, so it wasn't extreme long glass,although the fast max aperture obviously helped in that light.
I feel your frustration and if there's anything I can do to help (short of parting with my 500mm) then I'm only too happy to do so, but I'm definitely not practising any sort of Photoshop black magic - far from it!
A few points in answer to your last post...
As we discussed last time - when I say no NR I mean none other than that which is being applied in DPP at minimum default levels. I never alter them, and if there's any appreciable effect on actual subject detail , then I can' t see it. I NEVER run NR on the whole image, I hate the loss of detail that ensues. When I run NR it's ouitside of DPP - in Paint Shop Pro 7 in fact, and only ever on the bg. With 800 pixel web images I rarely need to do that since the 50D and certainly noticeably less with the 70D. I'm truly mystified when you say you don't know how I do it, because I'm doing nothing clever or involved to my images at all. I'd go so far as to say I probably do less post processing than most people. I'm not trying to be a smart arse, and I certainly wouldn't deliberately mislead anyone over this, in fact I've explained time and again on the board how to run selective NR just on the bg.On the occasions I do so I'll always say so.
My workflow...
Invariably a RAW file and opened initially in DPP.
I very rarely do anything at all to the images in DPP, they mostly look great staright out of the tin. I never need to adjust WB. I sometimes may need to do slight highlight or shadow recovery, but it's quite unusual.
I then output a 16 bit TIFF file to PSP 7, (If I intend to use it immediately),where I resize and make any minor levels adjustments, sharpen for the downsize and save. I'd ony run NR on the BG and only if it was intrusive in the bg, which I'm increasingly finding it isn't, and I shoot at 800 ISO most of the time, and 1600 ISO quite often. A bird which has intrusive noise in the feather detail woiuld be a bin job, unless it was a Golden Oriel, when I might make an exception.
With regard to your two sample crops, I know you're trying to make a point, come on though - you just wouldn't crop either of those images that tightly surely, but yes- the noise is bad. Here's an edit with NR run just on the bg, to show you could improve the BG noise at least .I wouldn't even try to run NR on the bird itself,you'd just lose feather detail.
This is trying to address noise which is really a bridge too far in this instance though, and it isn't the root cause of your problem anyway - It's pretty obvoius, you need to either...
(a) Get closer.( I hate how glib that sounds, because I know all too well, it's often just not possible), OR
(b) Invest in some longer glass if you're seriously interested in bird photography, and I imagine you are.
Seriously mate ,the last thing you need at the moment to pursue bird photography is yet another camera body, especially one which is only going to compound your problems at that range. You need to invest in some longer glass if you're serious about bird photography - all birders eventually do! If it's just a mild interest and you don't want to invest in the glass, then fair enough, but there are lots of people on TPF posting some pretty stunning images with the 50D- you'd have to be blind not to see them. Yes - the 50D and cameras like it are extreme, and you need the lenses to extract the best from them, but decrying the camera when you're seriously under-gunned for the job you're trying to do is unfair and misleading to people considering buying the camera. The Robin shot I posted was with the 300mm 2.8 btw, so it wasn't extreme long glass,although the fast max aperture obviously helped in that light.
I feel your frustration and if there's anything I can do to help (short of parting with my 500mm) then I'm only too happy to do so, but I'm definitely not practising any sort of Photoshop black magic - far from it!
- Messages
- 6,481
- Name
- Tim
- Edit My Images
- No
CT, thanks for posting. I agree that (a) getting closer is the best solution; (b) getting longer glass is the next best solution; (c) getting higher pixel density is not necessarily the solution at all.
Like I said, I'm really not trying to knock the 50D and 7D. Indeed I have many times jumped to the defence of the 50D when people complain about its pixel level noise. All I'm really trying to do is warn people that throwing pixels at a problem may not be the best solution, and they should not expect the 7D to work miracles for them if the other parts of the equation are not in place. The better solution by far is to fill the frame with your subject more, not crop away huge swathes of image. I think we are in agreement
EDIT : Examples from someone else of heavy crops from the 7D. I don't know whether these are 100% or something smaller but they look absolutely bloomin' awful, and the EXIF does not appear to justify the poor IQ....
http://photography-on-the.net/forum/showpost.php?p=8917205&postcount=135
Like I said, I'm really not trying to knock the 50D and 7D. Indeed I have many times jumped to the defence of the 50D when people complain about its pixel level noise. All I'm really trying to do is warn people that throwing pixels at a problem may not be the best solution, and they should not expect the 7D to work miracles for them if the other parts of the equation are not in place. The better solution by far is to fill the frame with your subject more, not crop away huge swathes of image. I think we are in agreement
EDIT : Examples from someone else of heavy crops from the 7D. I don't know whether these are 100% or something smaller but they look absolutely bloomin' awful, and the EXIF does not appear to justify the poor IQ....
http://photography-on-the.net/forum/showpost.php?p=8917205&postcount=135
- Messages
- 5,869
- Name
- Richard
- Edit My Images
- No
I would be interested to see how many owners of 7d's think that the camera isnt as good as they thought it would be. I love mine and as to file sizes most of my work is reproduced at A1 & A2 for business conferences. I bought the 7d as a stand in till 1d mk4 arrived i love the changes canon have made to the 7d af is better than my 1d mk2 i was also very impressed with the lack of noise at higher iso i agree with Tim and CT good glass is a must to get the best out of the sensor. So come on does any owner not like their 7d and why
Regards
Lost
Regards
Lost
- Messages
- 803
- Edit My Images
- No
I like my 7D, but do not love it yet. I was expecting high-iso performance to be better than my outgoing 40D but to be honnest it is pretty similar. On a per-pixel level the 7D has a very slight edge, but the extra 8MP's do a good job of making things seem better than they are. I would say the 7D is usable at iso 3200 vs 1600 for the 40D. AF is where the largest gains have been made and the 7D is much better at tracking small moving objects. The viewfinder is better, but that does not really help take better pictures. Sealing is sopposedly better but I have not dropped it in a puddle yet to test (my 40D survived the puddle test). The LCD is better, 8fps vs 6fps is better (but I never needed 6fps and have not even tried 8fps). Handling is the same. As for HD Video - I have a camcorder for that!
I know the 7D is a good camera, but it just does not have the same wow factor I experienced when I opened the box of my 40D. Perhaps the 40D spoiled me because it was really ahead of the game two years ago. The 7D just adds bells and whistles.
Oh, nearly missed out micro focus adjustment. This is one really neat feauture that has helped me get the best out of my slightly back-focusing EF-S 17-55mm. Setting MFA to +5 has removed the odd mis-focus I was getting with this superb IQ / average build quality lens.
Even though I got my 7D at a bargain £1060, I am not 100% sure it was money well spent. Even with pretty sharp lenses (EF-S 17-55mm F/2.8, 70-200mm F/4L IS) the final results are not much better than on my 40D. Most of this is probably due to poor user ability where even the 40D surpassed my skill.
So IMO the 7D is good, but not as good as the 40D was two years ago, or as good as many people suggest. DSLR's appear to be evolving at a much slower rate now, so maybe I will not feel the urge to upgrade for another 3 or 4 years. Manufacturers really should concentrate on IQ and high-iso performance rather than squeezing more and more megapixels onto a postage stamp. I hate to think how much glass will cost to get the absolute best out of such high density sensors.
- Messages
- 273
- Name
- Paul
- Edit My Images
- Yes
Even though I got my 7D at a bargain £1060
Where from and how please!!!
Where from and how please!!!