alundeb wrote:
When you compare that way, at pixel level for hi/lo res cameras, and same ISO setting for large/small sensors, you get those results, no doubt.
For normalized pixel count, the 7D leads the 40D by a mile again.
For ~equivalent~ images (Same perspective, FOV, DoF and shutter speed), both 500D and 7D give me better print quality than the 5D at 16x24". Not by a mile, but a baby step for the 500D and two small steps for the 7D.
For time and time again, I see comparisons between different sensor sizes at the same ISO setting. That is biased because even if the framing is the same (that is good, for once), the DoF is narrower or the shutter speed is slower for the full frame sensor. For a comparison of apples to apples, adjust FL * 1.6, f-stop* 1.6 and ISO * 2.5.
Like if you show me an image taken with the 5D, at 135 mm, f/8, ISO400, and say it is better than the 7D image taken with an 85 mm at f/8, ISO 400. I will then answer that those images are different, and I can replicate how the 5D image looks by changing the settings on the 7D to f/5 ISO 160. ...Show more →
A couple of issues - one - I'm not using high ISO unless lighting conditions dictate that I do so. And if the situation calls for high-ISO, I am certainly shooting wide-open or close to it. It would be a rare situation for me where I would use higher ISO to allow me to use a smaller aperture. Most of the situations I've seen where I might want to use a small aperture, but the lighting is low, the subject is stationary. In such a situation, I'd just use a tripod and a longer shutter. The result would be cleaner anyway.
Second, if I want to shoot in a poorly-lit situation where I want the shallowest DOF - say, an 85/1.2 wide open on a full-frame body. 85 mm, f/1.2 at ISO 6400 cannot be duplicated on an APS-C body. Yes, I can duplicate the exposure, but no, I cannot duplicate the shallow DOF.
Tom, that's a very good description of when and why you need the largest sensor format possible.
I can easily imagine situations where I need DoF for moving subjects in low light, like groups of people, that's why I often go with the 17-55 2.8 instead of the 35 1.4.
From my experience 24-105 is not a very sharp lens on APS-C. Even with that in mind the 7D/24-105 image @f8 is uncharacteristically soft. Something does not look right there. I would go do some AF microadjustment or try to use a better lens as some have suggested.
You prefer the 7D's viewfinder over the 5D? Could you please elaborate? It was my impression that the 5D is considerably larger than the 7D. The 7D is 100%, and has the fancy AF indicators... is that what you prefer?
I'm a former 5D shooter who sold his 5D for a 40D because the 40D was a much more usable camera (frame rate, dust, LCD, live view, AF button). I really miss the 5D's viewfinder, however, so I'm interested to hear that you prefer the 7D over the 5D!
Frankly I think we are seeing the difference in sharpness of the 24-105 at full wide and at 40mm than the cameras. I think the 24-105 is a fabulous lens but without doubt it is a bit softer full wide than at other focal lengths.
not exactly what the OP asked for, but here is a 1ds2/7d side by side....very unscientific, just me jerkin around. 1ds2 on left 7d on right. Remember the 7d LR profile is not ready, but I made my own profile in DNG editor via a macbeth color card.
400iso, 24-105, f8 500th
Jeff Kingston wrote:
Frankly I think we are seeing the difference in sharpness of the 24-105 at full wide and at 40mm than the cameras. I think the 24-105 is a fabulous lens but without doubt it is a bit softer full wide than at other focal lengths.
Could be, but I know that my copy of the 24-105 is very sharp at the wide end, even wide open and these samples were at f/8!
Jeff Kingston wrote:
Frankly I think we are seeing the difference in sharpness of the 24-105 at full wide and at 40mm than the cameras. I think the 24-105 is a fabulous lens but without doubt it is a bit softer full wide than at other focal lengths.
My copy is quite sharp at 24mm and 40mm. Very similar performance at the two settings. Photozone's tests say the same thing.
alundeb wrote: Dawei Ye wroteThere always seems to be some romantic notion that new cameras have better IQ. Not so. The 5D clearly has better IQ than the 7D. I tried the 7D against the 5D Mark II and 40D. At the pixel level, the 7D was lagging behind by a mileWhen you compare that way, at pixel level for hi/lo res cameras, and same ISO setting for large/small sensors, you get those results, no doubt.
For normalized pixel count, the 7D leads the 40D by a mile again.
For ~equivalent~ images (Same perspective, FOV, DoF and shutter speed), both 500D and 7D give me better print quality than the 5D at 16x24". Not by a mile, but a baby step for the 500D and two small steps for the 7D.
For time and time again, I see comparisons between different sensor sizes at the same ISO setting. That is biased because even if the framing is the same (that is good, for once), the DoF is narrower or the shutter speed is slower for the full frame sensor. For a comparison of apples to apples, adjust FL * 1.6, f-stop* 1.6 and ISO * 2.5.
Like if you show me an image taken with the 5D, at 135 mm, f/8, ISO400, and say it is better than the 7D image taken with an 85 mm at f/8, ISO 400. I will then answer that those images are different, and I can replicate how the 5D image looks by changing the settings on the 7D to f/5 ISO 160.
The IQ of the 5D shines at lowest ISO, or narrowest DoF, or with wide angle lenses. Those are the cases where the ~equivalent~ on a crop camera is impossible to acheive. ...Show more →
otoh the 7D needs higher shutter speeds to maintain sharpness so that gives back some of your adjustment there
carlk wrote:
From my experience 24-105 is not a very sharp lens on APS-C. Even with that in mind the 7D/24-105 image @f8 is uncharacteristically soft. Something does not look right there. I would go do some AF microadjustment or try to use a better lens as some have suggested.
it does look a bit soft
although it is true that the 24-105 i tried wasn't the sharpest beast in the world at 24mm
and the 7D does have a bit rougher time getting things fully resolved
but for the 18MP sensor to appear to deliver like maybe 7MP equivalent to the 13MP 5D just makes no sense at all, at the very least i would think it would at least deliver a trace more detail!
skibum5 wrote: carlk wroteFrom my experience 24-105 is not a very sharp lens on APS-C. Even with that in mind the 7D/24-105 image @f8 is uncharacteristically soft. Something does not look right there. I would go do some AF microadjustment or try to use a better lens as some have suggested.it does look a bit soft
although it is true that the 24-105 i tried wasn't the sharpest beast in the world at 24mm
and the 7D does have a bit rougher time getting things fully resolved
but for the 18MP sensor to appear to deliver like maybe 7MP equivalent to the 13MP 5D just makes no sense at all, at the very least i would think it would at least deliver a trace more detail!
Exactly. I could understand any noise differences but not the sharpness. It just does not make sense for a down-rez'd higher resolution image to be softer when it was taken with the same (properly calibrated) lens.
Its al the heavy NR Canon put in to counter the noisy sensor. The test he did was fine. Sid eby side, the 5D sensor produced a cleaner, sharper, more detailed image than the 7D sensor did. Thats it.
Ken Campbell. I have a 1d mk2 and had made my mind up to replace it with a 1d mk3. The launch of the 7D has created some doubt in my seventy three year old mind !
I bet that 73 year old mind is telling you right Ken.
I use a 1D3 and I think if you do BIF, you might better keep that 1DII. Canon has not matched the AF performance of that camera to date yet. If you get mixed posts here telling you otherwise, go with you instinct. If you are reading a bunch of posts where people arent getting good focus for BIF with these cameras, chances are you wont either. You know that already though.
You could do this, rent a 1D3 and try it. About $250 will get a week to spend shooting birds to try for yourself. I bet you will send it back. I dont know anyone thats been more happy with the 1D3 AF after owning the 1DII. Know IQ is way above the 1DII when its in focus. You can use ISO1600-3200 and really never think twice about it.
Maybe the 1D4 that should be out in a few weeks(rumored) will be better. I sure hope so. If you live close by, I'd be happy to let you use mine. I have some friends that have both and I tried their 1DIIn's and its better. You can almost just forget about the LCD, cause it will be in focus. The 1D3 on the other hand takes some chimping on the LCD, cause you never know.
If you like, email me and I'll shoot you some examples for you to look thru Ken.
although it is true that the 24-105 i tried wasn't the sharpest beast in the world at 24mm
and the 7D does have a bit rougher time getting things fully resolved
but for the 18MP sensor to appear to deliver like maybe 7MP equivalent to the 13MP 5D just makes no sense at all, at the very least i would think it would at least deliver a trace more detail!
I don't know what this lens is like optically, but the 5D has a pixel pitch almost twice that of the 7D. Unless the lens is improbably good, it's going to look much softer on the 7D on a per-pixel basis. These results aren't far from what I would expect, at least at f/8. Things might improve for the 7D (and the 5D too, to a lesser extent) at f/5.6 if the lens is good enough.
Crop-sensor cameras also demand more from the focussing system, which is another factor that leads to full-frame cameras doing better in the real world.
I'm waiting to see what the next FF offering is from Canon, and if it's another MP monster I'm buying 3 new 5D's and calling it quits for another 4 years. By the way, any comparisons at ISO 100?
It's best to assess sharpness at 50% rather than 100% which is rather harsh on the camera as pixel counts skyrocket. Imaging looking at the 100% images from a 60MP Phase One, this would correspond to a gigantic enlargement.
Specularist wrote: skibum5 wroteit does look a bit soft
although it is true that the 24-105 i tried wasn't the sharpest beast in the world at 24mm
and the 7D does have a bit rougher time getting things fully resolved
but for the 18MP sensor to appear to deliver like maybe 7MP equivalent to the 13MP 5D just makes no sense at all, at the very least i would think it would at least deliver a trace more detail!I don't know what this lens is like optically, but the 5D has a pixel pitch almost twice that of the 7D. Unless the lens is improbably good, it's going to look much softer on the 7D on a per-pixel basis. These results aren't far from what I would expect, at least at f/8. Things might improve for the 7D (and the 5D too, to a lesser extent) at f/5.6 if the lens is good enough.
Crop-sensor cameras also demand more from the focussing system, which is another factor that leads to full-frame cameras doing better in the real world.
It's a good test, but I think lens and aperture could be factors here. The best-case optical match to the 7D sensor is probably the 17-55, and the 7D diffraction limit is something like 6.3, so f/4.5-5.6 would probably be a better aperture for looking at 100% crops.
In my comparisons with the 5D2, it's clear that the 7D is less sharp at the pixel level, so I'm not surprised that the 5D wins on that measure. But in a comparison at equal image magnification, it would be a surprise me a lot if the 7D did not resolve a bit more detail than the 5D, given best-case optics.