7D noise at ISO100? (very large screen cap to show the problem!)

Ben I have to say that its true from my perspective, noise is apparent at low iso's on the 7d. However you will need to pixel peep to see them. I print on a hp 1120 plotter at work and I dont see noise in my 24x36 prints =) so its up to you if it bothers you or not I guess.

J_TULLAR wrote in post #10465109
Ben I have to say that its true from my perspective, noise is apparent at low iso's on the 7d. However you will need to pixel peep to see them. I print on a hp 1120 plotter at work and I dont see noise in my 24x36 prints =) so its up to you if it bothers you or not I guess.

now mr. ben, this answers your question.

Have noticed it when zoomed to 100% on large areas of solid colors such as skies - doesn't bother me at all and its never been apparent in any large print I have made using the 7D.

The same on 5D II for sure, I've posted long that issue long time ago.

monk3y wrote in post #10464342
...he was just merely comparing the photos at ISO 100...

No. What original poster is doing, is viewing a higher pixel resolution image at GREATER ENLARGEMENT than another image on screen when both are at 100%, because the 7D image is larger. The greater the enlargement of any image, the greater the visibility of image flaws like motion blur, noise, out-of-focus subject, etc.
When conditions are equalized, non-focal length limited, the D700 would be expected to have superior sharper image quality, at any ISO from higher signal to noise ratio of larger 35mm sensor with larger pixels.
But, when you are shooting focal length limited, like airplanes, birds, wildlife, then you have to make the comparison proper, or you make yourself nuts pixel peeping.
Alan Stankevitz, at birdphotographers did a some 7D comparisons, but I can't find the .pdf with the images. He ultimately did a bunch of 7D 1.6 FOVC sensor comparisons with 1D Mk III, 1D Mk IV, and 5D Mk II 35mm.
If you Google his name, the reports should come up somewhere, e.g.,
http://www.birdphotogr​aphers.net/forums/show​thread.php?t=53846

http://webcache.google​usercontent.com …nk&gl=us&client​=firefox-a

J Rabin wrote in post #10465319
No. What original poster is doing, is viewing a higher pixel resolution image at GREATER ENLARGEMENT than another image on screen when both are at 100%, because the 7D image is larger. The greater the enlargement of any image, the greater the visibility of image flaws like motion blur, noise, out-of-focus subject, etc.
When conditions are equalized, non-focal length limited, the D700 would be expected to have superior sharper image quality, at any ISO from higher signal to noise ratio of larger 35mm sensor with larger pixels.
But, when you are shooting focal length limited, like airplanes, birds, wildlife, then you have to make the comparison proper, or you make yourself nuts pixel peeping.
Alan Stankevitz, at birdphotographers did a some 7D comparisons, but I can't find the .pdf with the images. He ultimately did a bunch of 7D 1.6 FOVC sensor comparisons with 1D Mk III, 1D Mk IV, and 5D Mk II 35mm.
If you Google his name, the reports should come up somewhere, e.g.,
http://www.birdphotogr​aphers.net/forums/show​thread.php?t=53846

http://webcache.google​usercontent.com …nk&gl=us&client​=firefox-a

You mad? If I look at my 5dII (21mp) at 100% I dont see much noise at 100iso, but when I look at my 7d (18mp) at 100% @ 100iso I see noise... I play with nikons at work and a I dont see much noise on our d3 at 100 iso nor do I see much noise with our d3x at 100 iso all at 100%. Rationalize all you want the camera has noise at low iso. If you wanna disregard a test then back it up with your own flawless test for everyone to see ... but as I said before its still a great camera and I dont see noise on my prints so it doesnt bother me much.

to everyone saying it's my exposure, or the lenses or anything else, that's NOT the case. I have RAW files shot side by side at "matching" exposures (the Nikon was shot at 200 and f/16 and the canon 100 and f/11 to use both camera's bases ISOs AND limit diffraction differences). It's there just as bad as it is here, it's just not a "real world" shot. These shots are shots I actually put on my website and I noticed the noise when processing them, so I'm showing them.

I realize the 7D shot has a much longer EFL and is more prone to camera shake.
I realize they're not identical shots/angles/exposures​.
I realize the 7D has 50% more pixels and therefor viewing both at 100% isn't "fair".

I actually pointed most of that out in the OP if you guys read it... But it's still a valid question, and of all the shots I've taken with a 7D so far, if there's a smooth gradient region in the shot at low ISOs you see this weird noise, end of story. If you have details in the entire image it's hidden. But my question was DOES IT MATTER IN PRINTS OR WEB RES IMAGES? IE what's the final result look like? I pixel peep to compare gear to make educated decisions. Once I've decided I just shoot the camera and enjoy the shots, but I don't want to blindly switch to a 7D and realize this is a major issue in prints in a month or so.

Remember, I shot a 5Dii for 9 months before this really started to bother me. but it's an issue on that sensor as well.

Rai33 wrote in post #10465139
Have noticed it when zoomed to 100% on large areas of solid colors such as skies - doesn't bother me at all and its never been apparent in any large print I have made using the 7D.

J_TULLAR wrote in post #10465109
Ben I have to say that its true from my perspective, noise is apparent at low iso's on the 7d. However you will need to pixel peep to see them. I print on a hp 1120 plotter at work and I dont see noise in my 24x36 prints =) so its up to you if it bothers you or not I guess.

thank you, that's exactly what I'm looking for w/o having to print something.
I think I'll be running some 4x6 crops from these tonight to see how they look.

this is how the shot looks at web res:

I had very noisy skies on an ISO 200 shot due to CPL darkening the sky quite a bit causing underexposure... However, my 24x16 print or whatever I have showed no noise, what-so-ever. It really is a non-issue, unless you peep at it all day on your monitor.

Invertalon wrote in post #10465659
However, my 24x16 print or whatever I have showed no noise, what-so-ever. It really is a non-issue, unless you peep at it all day on your monitor.

thanks.

jacobsen1 wrote in post #10465681
thanks.

I don't know if you have a CVS, Walgreens, or even more preferably a Sams membership, but if you do, try printing a large print. Sams Club has great prices, you would get a good large print to give away, hang, or sell, and you would know what to expect in the future.

For example, a 20x24 is $8, a 20x30 is $9...

Ben, after I posted my reply, I realized that I did not get to the real question about printing. I apologize. The answer depends on the size of the print and extent of post processing, I suppose, but I print up to 13 X 19 and do not have a problem. For 13 X 19 prints that do not require "fixing" the exposure in post processing, I would say there is no visible noise. My printer is a Canon i9900. I most often print on letter size paper and the images do not show any noise at all as long as I can scale the print to about 180 ppi or better.

I frequently use NeatImage, but it is more for the sharpening that I can get with minimal halos.

I did read your original post and all of the others, but I admit to being drowsy from staying up too late last night. I am still curious if you had auto lighting optimizer enabled. And to quote the late Bruce Fraser when asked, "how can I know what an image will look like when I print it", he replied, "Print it and then look at it". Terse, but true.

jacobsen1 wrote in post #10463816
What's the deal with this:
http://gear.benjacobse​nphoto.com …-07-01-at-11-00-15-pm.png

That's one 7D shot and one D700 shot, both at 100% on my screen. I know it's well above the "rules" for maximum resolutions, but bear with me, it's to make a point and the only image I'll hotlink in the thread...

anyway, it's obvious canon has gone after the high ISO end with their recent bodies (5Dii and 7D). I saw this with the 5Dii and it bothered me a bit, and now I'm considering a 7D and I'm seeing it again. Those 2 shots are pretty much straight out of the camera (WB tweaks to try to get them close) JPEGs. Yes JPEGs. The 7D clearly has more resolution, and it's shot with a cropper at full 300mm where the D700 is FF and shot around 230mm, so you could say the distance to the subject is helping the D700 resolve more detail (and it's obviously different lenses even though they're both 70-300s with stabilization).

BUT!!!! What's the deal with the noise? Because the 7D has ~50% more resolution, will they print pretty similarly at the same size? The D700 is NOT perfect either, it's just MUCH harder to see at 100%. Yes I know prints are what matters, but I'm considering buying a 7D so I'm trying to see what the current canon camp's thoughts are on the sensor in the 7D? Is this an issue? Does it hide well in prints? Does running NR on ISO 100 shots cure it? Or should I stick with a Nikon?

The best I can tell you is that I've occasionally noticed "blockiness" in blue skies at ISO 100 or 200 on the 7D when viewed at 100%. I think that processing 7D RAW files in LR3/CS5 will probably eliminate a lot/all of this, as they handle 7D RAW files better than any earlier converters I'm aware of. I can't comment on what to do if shooting in-camera JPEGs (since I shoot only RAW), but assume using a little more aggressive in-camera noise reduction (as Nikon apparently does by default) would take care of the issue.

TeamSpeed wrote in post #10465737
I don't know if you have a CVS, Walgreens, or even more preferably a Sams membership, but if you do, try printing a large print. Sams Club has great prices, you would get a good large print to give away, hang, or sell, and you would know what to expect in the future.

For example, a 20x24 is $8, a 20x30 is $9...

good point on checking with CVS for a cheap large print... I have one up the street so I'll see what their prices are... What I typically do is set the file up in PS so it's the size I'd generally print (12x18, sometimes larger) then just crop a 4x6 section out at that resolution and print those at home. I was too tired last night to bother, so I screen capped it and started a thread instead.

I WILL be printing these this weekend though to see.

Bill Boehme wrote in post #10465954
Ben, after I posted my reply, I realized that I did not get to the real question about printing. I apologize. The answer depends on the size of the print and extent of post processing, I suppose, but I print up to 13 X 19 and do not have a problem. For 13 X 19 prints that do not require "fixing" the exposure in post processing, I would say there is no visible noise. My printer is a Canon i9900. I most often print on letter size paper and the images do not show any noise at all as long as I can scale the print to about 180 ppi or better.

I frequently use NeatImage, but it is more for the sharpening that I can get with minimal halos.

I did read your original post and all of the others, but I admit to being drowsy from staying up too late last night. I am still curious if you had auto lighting optimizer enabled. And to quote the late Bruce Fraser when asked, "how can I know what an image will look like when I print it", he replied, "Print it and then look at it". Terse, but true.

in this shot, I think I did have ALO on. But I have other shots RAW to RAW (so NOTHING is on for either camera) that show the same issue, so it's not JUST that, but I do agree it doesn't always help. The blue sky isn't from that though, but the shadow noise could be.

I agree 100% about printing to see for yourself though. But that's why I'm asking. If people already know it's an issue why waste the print, but it sounds like it's very subjective and I'll be taking it to that step this weekend.

stsva wrote in post #10466032
The best I can tell you is that I've occasionally noticed "blockiness" in blue skies at ISO 100 or 200 on the 7D when viewed at 100%. I think that processing 7D RAW files in LR3/CS5 will probably eliminate a lot/all of this, as they handle 7D RAW files better than any earlier converters I'm aware of. I can't comment on what to do if shooting in-camera JPEGs (since I shoot only RAW), but assume using a little more aggressive in-camera noise reduction (as Nikon apparently does by default) would take care of the issue.

that screen cap is in LR3, so it's getting all the help it can there. I'll admit what drew my attention to the issue was a bullied (exposure/shadow wise) JPEG which isn't fair, and it's much less noticeable w/o any processing in LR. But I shot some tests yesterday RAW to RAW and still see the same thing, so it's not just that.

Interesting point on turning on/off noise reduction in camera though, and seeing what it does. Typically whenever my ISOs get to 800 and up I'm ALWAYS shooting RAW to handle the noise myself, but I DO shoot JPEGs at lower ISOs. I'll have to setup some sort of test for low ISO JPEGs with noise reduction on/off and all the different levels (including ALO) just to see how it looks. I'll obviously shoot RAW + JPEG for further testing now that it's a question.


FYI, this airshow shots were my first shots with a 7D ever. I borrowed it from a friend Friday and didn't touch it until Saturday at the show. I set it to my usual JPEG setting from my 5Dii day (landscape with some tweaks) and figured it'd be good enough. It is pretty much, except for the noise which COULD be due to that. But just for fun at an airshow I wasn't going to do RAWs or RAWs + JPEGs. But having seen it, he's letting me play with the camera some more to figure it out so I've been doing more tests this week and will some more this weekend hopefully.


Thanks again for all those offering suggestions and first hand impressions on prints they've done though, that's the point of the thread.

J_TULLAR wrote in post #10465404
You mad? If I look at my 5dII (21mp) at 100% I dont see much noise at 100iso, but when I look at my 7d (18mp) at 100% @ 100iso I see noise... I play with nikons at work and a I dont see much noise on our d3 at 100 iso nor do I see much noise with our d3x at 100 iso all at 100%. Rationalize all you want the camera has noise at low iso. If you wanna disregard a test then back it up with your own flawless test for everyone to see ... but as I said before its still a great camera and I dont see noise on my prints so it doesnt bother me much.

Mad, as in crazy? Hope not. One demands a more costly 35mm sensor, costing thousands $$ more, like Nikon 3 series, 5DII or Canon 1 series, to have significantly higher signal to noise ratio performance than a smaller sensor, at any ISO, on areas of images prone to revealing defects (shadow areas, neutral flat tones like skies, etc.).

One can also make yourself crazy comparing 100% screen enlargements. I doubt most of this shows in normal size prints on well processed images.
But, my point is, IF the poster IS going to compare 100% enlargements between different sensors, then at least data needs to be normalized for image resolution enlargement.

In properly conducted up-rez noise tests I've viewed, like 7D vs. 40D noise from Bob Atkins, and the 7D vs. Nikon D300s tests, 7D out-resolved, had lower noise, less digital pattern noise, and made better prints (when they did that step).

Me? I shoot 1D IV and 40D (soon traded for 7D). 1D IV makes the best prints of my life, particularly blacks and dark tones, up to 13x19, all the way to ISO 5000, that I've ever experienced. Ever. Canon sure got the 1D IV quality right. I do mostly macro. In focal length limited birds/airplanes/wildli​fe, the 7D should yield good comparative images, when the comparisons are conducted properly.
Not sure how this makes me mad, as in loony? Jack

jacobsen1 wrote in post #10463816
Blah blah blah...

That's one 7D shot and one D700 shot, both at 100% on my screen.

Do you have any slightest idea what "Pixel Density" is? Sorry don't mean to be rude, but take a look at this:

• FF vs Crop (Which means 60% extra reach)
  
• 18mp vs 12mp (which means 50% extra resolution)