Sending my 7D2 back due to high ISO noise

[zlatko]

What's wrong with the image?

The largest version on Flickr (1189 x 1500) just looks like it's out of focus. Not a problem if most of the other shots are in focus. It would be easy to miss focus in that tall grass.

I enjoy full frame and crop cameras. Full frame for better IQ. Crop for smaller size. I'll likely pass on the 7D2 because it doesn't offer any size/weight advantage over the 5D3 or 6D. The 7D2 offers some great features that are not a priority for me right now.

For IQ, the 7D2 should be compared to other APS-C cameras. When compared to full frame, FF will always have a little more detail and a little more high ISO.

 

[Marsu42]

The next is of the same subject with the 7D2 at ISO 800. The detail on the faces is much rougher. The pattern just below the head (the eye is slightly OOF) is less defined than the 5D3 image.

Agreeing with the other comments above, underexposing @iso800 is not what the manufacturer intended, better expose correctly @iso1600 even on crop unless you run into color problems. The ff of course gives you more data to work with, but even with crop and proper postprocessing you can enhance images such as this a lot and get good results for flickr/web sizes.

But maybe a good thread to warn off other potential downgraders ff->crop for more reach: The crop sensors including the 7d2 are for good light only, the very reason why I bought my ff 6d after getting frustrated with the 60d in cloudy weather in autumn and winter.

 

[alexanderferdinand]

(not so) off topic: stunning shots BTW- favorit: eagle with the fish- GREAT!

 

[scyrene]

Clearly it's a problem with the lens. You need a better one...







I, too, am passing on the 7DII. As I've stated previously, the main advantage of APS-C is lower cost. Having owned the 7D, moved to the 5DII for everything but birds/wildlife (because the AF of the 5DII wasn't up to the task), then moved to the 1D X and selling both the 5DII and 7D, I don't see going back to APS-C, except when the constraints of an outing preclude a full sized camera, and in that case the 7DII would also be too big (and that's where my EOS M comes into play).

I agree on the cost thing. But I wonder what you think about this - something I've been considering while idly wondering if I should get this as a backup to my 5DIII. At present I use the 500+2x much of the time for small birds. I use it at f/10 because stopping down seems to improve sharpness a bit. However, this is a bit narrow when light levels are poorer (and the AF is much restricted). If I use the 7DII, I could stay with the 500+1.4 at f/5.6, and the crop reach is 1120mm (which is close enough to 1000mm for me). That extra aperture would allow me to use it in much lower light.

Now, I know the high ISO on a crop sensor isn't as good. So does it all cancel out? I don't think anyone has addressed this, admittedly rather special case. Essentially, are 5DIII @ 1000mm f/10 and 7DII @ 700mm f/5.6 basically the same (and not worth buying a new camera for), or in low light is there a little advantage to the crop here?

I'd welcome anyone else who has thoughts. Thanks

 

[Loswr]

Essentially, are 5DIII @ 1000mm f/10 and 7DII @ 700mm f/5.6 basically the same (and not worth buying a new camera for), or in low light is there a little advantage to the crop here?

I suspect the 5DIII would have an IQ advantage, more so as the ISO rises. The 7DII would have a clear AF advantage in that comparison.

 

[takesome1]

Here are some sample images to illustrate my frustration. Note that these images from an artistic standpoint suck - but were an attempt to take a photo of the same subject under the same realistic lighting conditions with both cameras. Both images are near 100% crops and have absolutely no noise reduction or PP from the RAW images.

The first one is from the 5D3 at ISO 1600. The lighting was very poor at this time, but the details are fair. A similar shot of a more interesting subject could probably be salvaged.
6O6C8809.jpg by CalevPhoto, on Flickr

The next is of the same subject with the 7D2 at ISO 800. The detail on the faces is much rougher. The pattern just below the head (the eye is slightly OOF) is less defined than the 5D3 image.
388A0341.jpg by CalevPhoto, on Flickr

I do have other images at ISO 800 that are better. You are correct that pushed to the right the results are better, but I often do not have this convenience in Seattle.

Here is a mink I took during the same shoot about two minutes after these shots. I took it with my 5D3 at ISO 1600 and it was still too dark so I had to boost up the exposure quite a bit. I then did some PP and cleaned up the noise. The resulting shot still has nice details. There is no way the 7D2 would have done the same.
6O6C8788-Edit.jpg by CalevPhoto, on Flickr

You just haven't developed an understanding of your gear yet.
Yes the 7D II RAW files are noisy, and they can be cleaned up in LR.
ISO 1600 is quite usable once you properly PP.
Do this however, shoot RAW + JPG and compare. In camera JPG conversion is doing a fair job at cleaning up noise.
If your PP isn't exceeding the 7D II the problem is in your PP not the camera.
If both are still bad maybe you have a bad copy.
The copy I have is performing exactly (and sometimes better than) I expected.

Also if you think the noise out of the 7D II should be as clean as the 5d IIIthen the problem is with your expectations.
I have never seen this clai anywhere. You should have expected at least a stop difference.

 

[scyrene]

I do have other images at ISO 800 that are better. You are correct that pushed to the right the results are better, but I often do not have this convenience in Seattle.

Agree that the 7D2 is not the ideal camera for the heavy cloud cover of the Pacific Northwest. For you, full frame and big glass may be the best option. The 7d2 doesn't do well at all with underexposed images which is what i see in the duck photos. The mink photo, which is very nice, also could have used an extra stop with the 5D3 I imagine. My guess is you would have gotten better images at higher ISO with both bodies.

It's important to always keep in mind that the 7D2 has a 60% magnification imbedded in its images. That magnifies your subject, but also exaggerates the flaws. The 7D2 isn't going to be the best choice for everybody but It can be an excellent choice for some when used as it's designers intended.

I cured myself of my chronic under exposure problem, a remnant of shooting film, by switching to auto ISO. I try not to let ISO get completely out of control, but ETR almost always gives the best result with the current camera bodies. It seems counterintuitive, especially to old school film shooters like me, but the underlying science for it is quite sound.

I certainly found that when I stopped holding back on exposing to the right I got much better results overall. I wouldn't have dreamed of going over ISO 3200 with the 5DIII but when I started using the technique properly, I found up to ~ISO 8000 could yield usable results. I imagine the 7DII could do fine at ISO 3200-4000, especially as you'd be cropping less in pp.

 

[scyrene]

Essentially, are 5DIII @ 1000mm f/10 and 7DII @ 700mm f/5.6 basically the same (and not worth buying a new camera for), or in low light is there a little advantage to the crop here?

I suspect the 5DIII would have an IQ advantage, more so as the ISO rises. The 7DII would have a clear AF advantage in that comparison.

Oh sure, AF would be better but we know that (there are other advantages to 700mm, such as less chromatic aberration, which only seems to appear with the 2x extender). I'd still love to do some side-by-side comparisons. I guess any differences will be subtle in any case. Maybe I'm just talking myself into wanting this camera :/

 

[kirispupis]

The 7D2 should have an advantage with the crop, but in several cases I am not seeing that. For example in the coyote shot from earlier I compared it with a similarly posed coyote taken in worse light with my 5D3 but same rough ISO (400) and found that the detail is roughly the same despite the smaller size of the coyote on the 5D3.

http://www.calevphoto.com/p1067692827#h3ad922d8

In terms of the underexposed images that is what the light was at the time. With my 5D3 normally I would have shot these at ISO 3200, but at the time I was trying to get a good comparison between them. The point of illustration was the noise and loss of detail - not the quality of the shots.

I looked in LR just now to see the average ISO for my shots and can see:
ISO 3200 and above - 20%
ISO 1600 to 2500 - 25%
ISO 800 to 1250 - 20%

This is of all my shots, so most of the shots at lower ISOs were tripod based landscapes while most of the shots at higher ISOs are of wildlife.

 

[Marsu42]

This is of all my shots, so most of the shots at lower ISOs were tripod based landscapes while most of the shots at higher ISOs are of wildlife.

Wildlife as in "bif" doesn't equal wildlife as in "animal standing in the scenery and looking at you". For the latter, you can get away with a much lower iso setting esp. with an IS lens and if you snap a couple of frames and chose the best one. For actual movement in lower light, crop simply no can do.

 

[takesome1]

The 7D2 should have an advantage with the crop, but in several cases I am not seeing that. For example in the coyote shot from earlier I compared it with a similarly posed coyote taken in worse light with my 5D3 but same rough ISO (400) and found that the detail is roughly the same despite the smaller size of the coyote on the 5D3.

http://www.calevphoto.com/p1067692827#h3ad922d8

In terms of the underexposed images that is what the light was at the time. With my 5D3 normally I would have shot these at ISO 3200, but at the time I was trying to get a good comparison between them. The point of illustration was the noise and loss of detail - not the quality of the shots.

I looked in LR just now to see the average ISO for my shots and can see:
ISO 3200 and above - 20%
ISO 1600 to 2500 - 25%
ISO 800 to 1250 - 20%

This is of all my shots, so most of the shots at lower ISOs were tripod based landscapes while most of the shots at higher ISOs are of wildlife.

If you want to shoot at ISO 3200 and above the 7D II is not the right tool.
The half hour around sunset and just before is a curse for wildlife photographers. Animals start moving, the light is to low. I feel the pain.
I wouldn't push the 7D II past ISO 1600 and expect great results shooting wildlife. At sunrise the 7D II starts work later than the 5D III and will end its day before the 5D III.

My 7D II underexposes a bit over 1/3 of a stop.

 

[Famateur]

http://www.calevphoto.com/p1067692827#h3ad922d8

Great photographs, Joseph. Thanks for sharing. I really like the head-on portrait of the bull elk with one antler growing from near the middle of his forehead -- never seen that before. If he sheds the other first, he'll be a unicorn elk for a while.

 

[kirispupis]

The 7D2 should have an advantage with the crop, but in several cases I am not seeing that. For example in the coyote shot from earlier I compared it with a similarly posed coyote taken in worse light with my 5D3 but same rough ISO (400) and found that the detail is roughly the same despite the smaller size of the coyote on the 5D3.

http://www.calevphoto.com/p1067692827#h3ad922d8

In terms of the underexposed images that is what the light was at the time. With my 5D3 normally I would have shot these at ISO 3200, but at the time I was trying to get a good comparison between them. The point of illustration was the noise and loss of detail - not the quality of the shots.

I looked in LR just now to see the average ISO for my shots and can see:
ISO 3200 and above - 20%
ISO 1600 to 2500 - 25%
ISO 800 to 1250 - 20%

This is of all my shots, so most of the shots at lower ISOs were tripod based landscapes while most of the shots at higher ISOs are of wildlife.

Hey kirispupis, Hope I didn't offend.

I get what you were trying to do. I'm not second guessing either your conclusion or choices about what's best for your work.

My concern is that other folks, who might benefit from a 7D2 for their work, might conclude that it's not possible to get good results with the camera. I'd respectfully disagree with that generalization.

You had to know that if you posted examples that some folks would focus on your methodology rather than the substance of your argument. It's just what they do. My ETR comment is based on my personal experience. I don't have any idea how that reflects on your work. Hope I didn't seem to imply that it did. Any way, good luck with the 5D3/200-400 combo. I don't think that gear is going to be holding you back -Brian

No offense taken. I do believe the 7D2 can be a great tool for those who live in places where the light is better such as Arizona or Florida. Here in the PNW that just isn't the case - especially this time of year.

My disappointment with the 7D2 was that I expected ISO 1600 to be about the same as the 5D3 at 3200, but from experimentation I found this not to be the case. I also was disappointed with the amount of noise I found even at lower ISOs.

The big decision factor was last night when I showed my wife the photos from both cameras and offered her the choice.
a) Keep this camera and sell my 6D. Do not purchase the 1DX when it comes out, but do upgrade my 5D3 to a 5D4.
b) Send this camera back and replace my 6D with a 1DX2 when it is released. Some time after the 5D4 release upgrade my 5D3.

Obviously choice b is considerably more expensive, but after looking at the results she came to the same conclusion I did and (with regret) gave tentative approval for choice b.

 

[Marsu42]

When you look at all of the tech that Canon was willing to put in the $1800 7D2. Imagine what kind of goodies they are holding back for the next 1D or 5D or even 3D if that's what it is.

I don't think they're holding back a lot (even tough they would like to) because of the market pressure. For the 5d4 it's clear what you can expect: rgb+ir metering, dual pixel af, probably 4k video, once again updated phase af.

I imagine after that Canon's old-school dslr line will run out of steam and they need to introduce serious tech changes - otherwise there'll be little reason to upgrade to the 5d5 or whatever when you can get used 5d3/5d4 for a fraction of its price.

 

[dtaylor]

I am no crop camera apologist, but to me the differences are small, couple that with the fact that the 7D MkII shot light is flatter, either a few minutes later or on a more shady part of the water, and if they are both close to 100% then you are enlarging the 7D MkII image nearly twice as much, makes me think you might not be comparing this as evenly as you think.

Yeah...I clicked all the way through to Flickr to make sure I was viewing the original size, flipped between the browser tabs a couple times...and had to come back here to read which was 5D3 and which was 7D2.

I say it all the time even though it offends people and makes me a pariah, but if you expect A to be better then B you will believe it is better and rationalize that it is better even when someone has flipped the labels and you are actually praising B! This is seen in wine tasting, audio equipment testing, photo print comparisons.... Double blind the test and people often can't tell the two apart where they previously thought the difference was huge...JUST HUGE!

That said, I would not normally recommend a 5D3 owner get a 7D2 for extra reach. Too much is made of the difference. You have to be cropping much further then APS-C...and then printing large...to observe the reach advantage of a crop sensor. It comes into play when you're left with 8 MP from the crop sensor, and <3 MP from the FF, and you need to print 16x24. And if you run into that situation it is significant. You can pick it out in a double blind test. But honestly ask yourself how often you run into that situation? If it's very often you might be better served by a new lens any way.

 

[kirispupis]

I am no crop camera apologist, but to me the differences are small, couple that with the fact that the 7D MkII shot light is flatter, either a few minutes later or on a more shady part of the water, and if they are both close to 100% then you are enlarging the 7D MkII image nearly twice as much, makes me think you might not be comparing this as evenly as you think.

Yeah...I clicked all the way through to Flickr to make sure I was viewing the original size, flipped between the browser tabs a couple times...and had to come back here to read which was 5D3 and which was 7D2.

I say it all the time even though it offends people and makes me a pariah, but if you expect A to be better then B you will believe it is better and rationalize that it is better even when someone has flipped the labels and you are actually praising B! This is seen in wine tasting, audio equipment testing, photo print comparisons.... Double blind the test and people often can't tell the two apart where they previously thought the difference was huge...JUST HUGE!

That said, I would not normally recommend a 5D3 owner get a 7D2 for extra reach. Too much is made of the difference. You have to be cropping much further then APS-C...and then printing large...to observe the reach advantage of a crop sensor. It comes into play when you're left with 8 MP from the crop sensor, and <3 MP from the FF, and you need to print 16x24. And if you run into that situation it is significant. You can pick it out in a double blind test. But honestly ask yourself how often you run into that situation? If it's very often you might be better served by a new lens any way.

Sorry to disagree but it's not just about reach. Setting aside the debate about the smaller sensor and the effect that has on image quality, pro's and con's, my 5D3 feels like a lumbering dinosaur after shooting with the 7D2 for two weeks. It really is that much more responsive. I can see how people get hooked on those 1 series bodies. If you are shooting with a 1DX or 1D4 than it's hard to make the pro 7D2 argument. But the AF and 10/fps are alone worthy of adding this camera to a 6D or 5D3 if you are shooting moving subjects.

In my experience I really did not see that. Yes, the AF of the 7D2 is a bit quicker and the frame rate is higher, but I did not feel it really blew my 5D3 away. The AF of the 5D3 has never really prevented me from getting the shot, but I did find that the 7D2 marginally increased the rate of keepers.

The primary improvement I saw in the 7D2 was the customizability. I really liked being able to set different AF modes to different buttons. I also felt that the silent shutter was even better.

 

[applecider]

Kirispupis where are you in the PNW?

I'm in portland and while the wildlife opportunities are legion this time of year, the light is not.

I shoot with the 5diii and 1dx at between 800 and 3200 and sometimes higher usually in order to get a shutter speed of 1250 for large lumbering birds and 2000 for fast flickering birds.

I think that a climate with consistent snow cover would brighten images and make a crop camera viable. But for me if I can't use iso 1600 to 6400 I might as well paint watercolors. Passing on 7d2, eager to see what the 5D4 and dx2 bring to table, but even there more smaller pixels probably means worse high iso performance, I would think that canon's dilemma here is that adding pixels would push iso performance down, and that is not the way to sell more bodies.

 

[Marsu42]

Although I'm sure there are folks who think the 5D3 has awesome AF. It's all relative I suppose.

Indeed. Try a 6d

 

[dtaylor]

Sorry to disagree but it's not just about reach. Setting aside the debate about the smaller sensor and the effect that has on image quality, pro's and con's, my 5D3 feels like a lumbering dinosaur after shooting with the 7D2 for two weeks. It really is that much more responsive. I can see how people get hooked on those 1 series bodies. If you are shooting with a 1DX or 1D4 than it's hard to make the pro 7D2 argument. But the AF and 10/fps are alone worthy of adding this camera to a 6D or 5D3 if you are shooting moving subjects.

No, it's not just about reach, and I didn't mean to imply it was. What I meant was if a 5D3 owner asked me about buying a 7D2 ONLY for additional reach...i.e. he didn't care about anything else...I couldn't recommend it just for that.

If you want the incredible responsiveness of the AF and 10 fps, by all means, 7D2. Brand new, nothing short of $6,500 is going to get you a more responsive action DSLR.

 

[scyrene]

Kirispupis where are you in the PNW?

I'm in portland and while the wildlife opportunities are legion this time of year, the light is not.

I shoot with the 5diii and 1dx at between 800 and 3200 and sometimes higher usually in order to get a shutter speed of 1250 for large lumbering birds and 2000 for fast flickering birds.

I think that a climate with consistent snow cover would brighten images and make a crop camera viable. But for me if I can't use iso 1600 to 6400 I might as well paint watercolors. Passing on 7d2, eager to see what the 5D4 and dx2 bring to table, but even there more smaller pixels probably means worse high iso performance, I would think that canon's dilemma here is that adding pixels would push iso performance down, and that is not the way to sell more bodies.

1/1250-1/2000 for birds? In flight yeah, but otherwise... that's a surprise. My favourite subjects are birds, usually smallish passerines, and I shoot around 1/320 in poorer light to 1/500 unless they're flying. There's no blur until you get down below 1/200-250 in my experience.