uz2work wrote:
That is only because the framing is different. If you shoot with the same lens from the same distance and with the same aperture and then you crop the 5D Mark II image to the same field of view as the 7D image, the depth of field of the 7D image and the cropped 5D Mark II image will be identical. On the other hand, if you use the same lens with the same aperture and adjust the distance so that framing will be the same, the 5D Mark II image will have shallower depth of field. In no case, on shots taken with the same lens at the same aperture and with the same field of view (whether cropped to the same field of view or with the same field of view because of a distance adjustment), will the 7D image have shallower depth of field.
For example, using the depth of field calculator that you have linked to, if you shoot with your 400/5.6 and a 7D at 100 feet and with an aperture of f5.6, the depth of field is 4.04 feet. If you shoot with the 5D Mark II and with the same 400/5.6 at f5.6 and adjust the distance to 62 feet to get the same framing, the depth of field is reduced to 2.43 feet.
( I really don't know why I'm bothering to reply.)
Alek Komarnits wrote:
P.S. Yes Les, I knew a bit underexposed ... but I had run out of light and felt 1/6s shutter speed was already pushing it. Appreciate your "turn off the LR sharpening" and I'm going to have to add that step to my processing ... in fact, I think I may just default to ZERO sharpening for the RAW conversion as I typically do in Photoshop as the very last step after any cropping/resizing.
Alex,
My comment was not intended in any way to be critical of any decisions that you made with regard to settings or anything you did with technique. I can't imagine how you could have done any better than you did with conditions that you were faced with. My only point was intended to be that, given the conditions, I think it would have been pretty hard to do any better than you did.
Also, with regard to the beautiful image that you added, while the noise may be visible at 100%, I would not be concerned in the least with trying to clean it up further because I can't image that the noise there would be visible at the image level or detract in any way from the quality of even a moderately large print. It is a great example of how noise can be bothersome only to those who insist on making such judgments by viewing at 100%.
Ian.Dobinson wrote:
or indeed change the focal length of the lens , which in the 5d2 case would be a 640mm
And using a 640 mm lens (if it existed) would make the depth of field even (slightly) smaller than it would be in the 600 mm example that I added in an edit (evidently) while you were posting.
^^^
in fact I just did the above settings but changed the FF to 610mm (couldnt get 640mm) and the circle of confusion is smaller on the FF than it is on the the crop
Ian.Dobinson wrote:
^^^
in fact I just did the above settings but changed the FF to 610mm (couldnt get 640mm) and the circle of confusion is smaller on the FF than it is on the the crop
The circle of confusion is, I believe, dependent only on the sensor size, and it doesn't vary with other factors, such as the lens being used, camera to subject distance, etc.
mco_970 wrote:
This thread is starting to smell like old cheese (which in turn is making me hungry for some cheese).
You are absolutely correct. And I have to admit that those of us who have taken a bite at the bait and responded to drivel that is replete with factual errors have contributed to derailing what had otherwise been a very useful thread that was helping those of us who have been getting good results from using our 7D bodies to get even more from them.
mco_970 wrote:
Huuummm, yep, still looks pretty good for ISO 3200. What are YOU expecting it to look like? ISO 100??
Not only do I agree that the noise doesn't look bad for ISO 3200, if you had to crop the 5D Mark II image down to the same field of view as what you would have with a 7D and the same lens (which is what you would have to do in a focal length-limited situation), the noise in the cropped 5D Mark II image would be virtually identical to what the noise is in the 7D image.
I fully agree with you when you qualify my premise about cropped cameras needing more DOF than full frame. Perhaps I should have been more prescise.
Let us put full frame and cropped cameras side by side 5m away at an oblique angle from a small bird with eyes that we focus on, a beak 3cm long and a some sort of crest 3cm behind. So we need to resolve both the beak and the crest, 6cm deep. The full frame at 5m with a 400mm lens has a DOF at F8 of 7cm, so beak and crest will be resolved.
The cropped camera with a 400mm (or is it 640?) at 5m and F8 has a DOF of 4cm. In order to get the crest and beak in focus, we move the cropped camera back to 6m(more if its 640), DOF 6cm, but we now loose some of the extra reach we had with the cropped camera.
We could stay at 5m and increase the aperture to F11 which has a DOF of 6cm and the beak and crest will be resolved. If we use F11 we will need either to drop the shutter speed, or underexpose, or increase the ISO. That is what I meant, by my original statement and that is why we use DOF tables.
Now I have to say that the ISO 3200 image looks pretty good, but the NR degrades it too much for my taste. The picture has the advantage of being in nice contrasty sunshine, which brings out the texture. Perhaps I should do the same test on a dull day? Don't worry I won't.
I have made my point about cropped cameras and ISO, and perhaps some will take note.
I fully agree with you when you qualify my premise about cropped cameras needing more DOF than full frame. Perhaps I should have been more prescise.
Let us put full frame and cropped cameras side by side 5m away at an oblique angle from a small bird with eyes that we focus on, a beak 3cm long and a some sort of crest 3cm behind. So we need to resolve both the beak and the crest, 6cm deep. The full frame at 5m with a 400mm lens has a DOF at F8 of 7cm, so beak and crest will be resolved.
The cropped camera with a 400mm (or is it 640?) at 5m and F8 has a DOF of 4cm. In order to get the crest and beak in focus, we move the cropped camera back to 6m(more if its 640), DOF 6cm, but we now loose some of the extra reach we had with the cropped camera.
We could stay at 5m and increase the aperture to F11 which has a DOF of 6cm and the beak and crest will be resolved. If we use F11 we will need either to drop the shutter speed, or underexpose, or increase the ISO. That is what I meant, by my original statement and that is why we use DOF tables.
The entire discussion has been about how the 7D has advantages in focal length-limited situations, and you have rejected or ignored all of the facts that show that it does. Now, you are changing the discussion to a situation that is not focal length-limited, and no one has said that the 5D Mark II does not have advantages over the 7D in such a non-focal length-limited situation.
mikeengles wrote:
Now I have to say that the ISO 3200 image looks pretty good, but the NR degrades it too much for my taste. The picture has the advantage of being in nice contrasty sunshine, which brings out the texture. Perhaps I should do the same test on a dull day? Don't worry I won't.
Mike Engles
But there is absolutely no reason why noise reduction needs to be applied universally to the entire image. Noise reduction can easily be applied selectively only to the areas of the image where noise is at an objectionable level. Areas where detail needs to be preserved don't need to have noise reduction applied to them because, in those areas, noise will be invisible unless viewed at a meaningless level of magnification.
Finally, as I said above, your posts in this thread continue to serve no purpose except to derail what has otherwise been a very helpful and positive thread for those who have an open mind and who are looking to learn things that will help them to get even more out of their 7D bodies. If your purpose has been to kill off such a helpful thread, you may well have succeeded.
I would ask that you have the courtesy, if you wish to persist, to start a new thread of your own that would allow you to discuss whatever points you feel are important.
Royraiden wrote:
Very noisy if you ask me, the files coming out of the 7D are really different looking, I dont think my previous 40D looked so "weird" even at high ISO'S.Comparing it to the 1D3 is out of the question for me, even at 6400 the files have decent detail on them.I shot a friendly football match 2 weeks ago with my 7D as my main body because my 1D3 was on its way to Canon for repair, I honestly was disappointing with the 7D's performance at ISO 800.Its a great camera with lots of great features that gets the job done but I have always said that 18mp on that sensor was just too much....Show more →
I have run a 40D and 7D back to back at ISO 3200, the 7D retained more detail and was cleaner when you resized it to the 40D size. Most likely what is happening is that the 7D has a new metering system, and it is quite possible your exposures are coming out a bit to the left more than the 40D, and you have to pull it back up, which will accentuate the noise. Without side by sides, though, it is hard to say.
40D then 7D, 100% crops, just one sample from the head to head...
TeamSpeed wrote:
I have run a 40D and 7D back to back at ISO 3200, the 7D retained more detail and was cleaner when you resized it to the 40D size. Most likely what is happening is that the 7D has a new metering system, and it is quite possible your exposures are coming out a bit to the left more than the 40D, and you have to pull it back up, which will accentuate the noise. Without side by sides, though, it is hard to say.
Canon's "mistake" in making the 7D was not in putting an 18 MP sensor in it that outperforms all previous Canon APS-C bodies in every way. Instead, their mistake was in not anticipating that people were going to insist on comparing files from the 7D to those of 8 MP and 10 MP cameras in a totally invalid manner by comparing the 7D and the 8 or 10 MP images to each other both at 100%.
n0b0 wrote:
.... I still wanna see what's wrong with Conrad's RAW for that ISO400 white bird on page 4. The one that he claimed to be "really really bad light".
Yeah. Me too. I wish Conrad puts up the RAW of that shot.
uz2work wrote:
Canon's "mistake" in making the 7D was not in putting an 18 MP sensor in it that outperforms all previous Canon APS-C bodies in every way. Instead, their mistake was in not anticipating that people were going to insist on comparing files from the 7D to those of 8 MP and 10 MP cameras in a totally invalid manner by comparing the 7D and the 8 or 10 MP images to each other both at 100%.
Les
[TIC mode on]
Lol. I guess for marketing purposes Canon should add other smaller RAW options aside from mRAW, sRAW, etc. in future bodies; e.g. - 300D-RAW, 20D-RAW, 40D-RAW, and so on.
This will be just adding some more lines of code, but it will allow people to shoot say a 20D-size (8 MP) RAW using a 7D MII with perhaps a 32 MP sensor. If 7DM2 buyers become unhappy when pixel-peeping 32 MP files for noise, they can just shoot at their old camera's RAW size, and be happy with their purchase.
And there will be peace in the internet.
(Albeit photo forums will become boring because everybody, perhaps including Mike, will agree how awesome the 7DM2 will be. What will we have left to argue on? )
In an effort to try to bring the thread back to something more positive, I'll post a sample of a picture that I recently took.
My preference is to be able to freeze the action because I like the feeling of capturing "a single moment in time". Thus, I wanted to keep my shutter speeds up. I would have liked to have shot with even a little faster shutter speed than I did because I still have a bit of motion blur in the wing tips, but to even get a shutter speed of 1/1600, I needed to use ISO 800 because of the combination of early morning light with somewhat overcast skies. I know that isn't extremely high ISO, but there are those who claim that anything over 400 is unusable.
Here is the full image converted in Capture One and before any post processing. Noise reduction settings were at default, and I completely disabled sharpening in the converter.
Here is a 100% crop of that converted image before post processing. Yes, there is some noise visible in the background at 100%, but it disappears at any magnification lower than 100%.
And here is the image after post processing. No further noise reduction was used, and sharpening was done selectively just to the bird. Note that, with the final composition, the image was down to about 7 MP. Had I taken the shot with the same lens and, say, a 5D Mark II, I would have been down to fewer than 3 MP, and use of the image would have been limited to only a very small print. Also note that I have admitted that my processing skills and post processing skills are not nearly at the level of some of the others here who, I'm sure, could have done better than did I.
thw2 wrote:
Those who insist on viewing their images 100% on screen all the time ought to stick to their 10 MP 40D and 12 MP 5Dc for the rest of their lives.
Andrew J wrote:
Then shoot Med RAW and you have 10MP with a 1.6 crop.
I've read a number of threads about the lower resolution settings and the general consensus seems to be that the camera doesn't average or bicubic reduce the image size but sub-samples the sensor. Therefore I wouldn't expect the 7D MRaw image to be as clean at the pixel level as the 40D raw. You would be better off shooting full raw on the 7D and bicubic reduce to 10 MP in post.
Also, while I think the low ISO 40D files are cleaner at the pixel level than the 7D files (yes 100% but bear with me ), my crop cameras are for use with my 100-400 where even in good light I usually need to use ISO 360 or above. The 7D's noise v ISO curve isn't as steep as the 40D. So for normal wildlife photography use I think the 7D IQ is an improvement over the 40D, if perhaps not a 1.8X improvement. I upgraded for the body features which are a larger improvement. It is a higher performance camera which requires its user to put higher effort into optimizing the outcome. The 40D was a very nice camera, I put a 15-85 lens on it and gave to to my wife for when she wants more than her P&S.
@thw2: I find my 5D2 as clean at pixel level as my 5Dc; both good, but this is OT.
Alek Komarnits wrote:
P.S. On the "lighter note" (pun intended), here's a fun little picture from the earlier in the same trip to Kenya - focus isn't dead on as you can see in the actual pixels shown in the 2nd image which is ISO 800, F/2.8, 1/100s ... but this was a fleeting capture. This is also an example of what I consider annoying noise even at ISO 800 ... so I may have to go back and try to clean it up after this informational thread! ;-)
Here's one of mine. ISO 2500, 500mm lens, hand held , f/4.5 (wide open), 1/160. Setting up a tripod was not an option. Caught him right at dark thirty, and he let me grab a pic. I was suprised to get a photo at all in those conditions. Sure IQ could be better, but I was pretty tickled that this worked out at all. I know there is some noise that could be removed, but it doesn't really bug me so whatever.
I have a couple of shots of a more mountain lion colored bob taken is similar circumstances to yours. More motion blur because the cat was in motion, but still I'm happy to have the shots because I wouldn't remember seeing him otherwise. Looking back at the EXIF, I could have chosen more optimal settings, but I'm happy to have what I've got.
