Take your photography to the next level and beyond...

  • NEWS

Why not join for free today?

Join for Free

Your total photography experience starts here

Canon 7D terrible noise levels! Disappointed.

certx 10 415 1 United States
17 Aug 2011 1:05PM
@ChrisV - DD

Yeah, that was my take on things. But I don't know where artifacting would be coming from. I didn't do anything at all to the photo. It's the way it came out of the camera basically.

shot in RAW at: ISO1600, aperture f/5, shutter 1/50.
opened in DPP, with no tweaks
kept the default (camera) settings of:
sharpening 3
no contrast/highlight/shadow/saturation, etc adjustments
default noise reduction of: chroma noise 10, luminance noise 3 (so technically, that's even got slight, mild reduction)
cropped a 100% section of her face
converted straight to JPG

There should be next to no noise in this picture. At 1600, even totally 0 NR, there should not be this much noise in the frame. You'll also notice some very prominent, hard white, spots; A TON of what I think look like hot pixels .... but they aren't because they're not there below ISO400. I THINK hot pixels would show at all ISO's... I think. For example, my 5Dmk2 DOES have 1 hot pixel, and it's there at all ISO's. It's only one so I don't mind it much, it's virtually invisible unless displayed very large, and easily controlled with spot removal tool.

I have not yet emailed Canon with it, because the 7D was my first experience with what is supposed to be a higher end dslr, and thought I was just doomed to be plagued with the noise level. I've since come to realize that the noise level is excessive, but wasn't sure if it was something I was doing or if it's the camera.

Join ePHOTOzine for free and remove these adverts.

ChrisV Plus
10 1.5k 26 United Kingdom
17 Aug 2011 1:31PM
I've taken the photo below on my 550D this lunchtime, at ISO 1600. It's a 100% crop loaded into Aperture in RAW. I've removed all the default processing that Aperture applies [from the profile Canon add to the camera] - which includes the noise reduction and moiré. I then exported as a JPEG.

There's obvious noise there, but it doesn't look anything like [or anywhere near as bad] as the sort of patterning on your shot - which just looks very strange to me - that is a photo of a person [as opposed to a picture of a person] in your shot? It's just it looks almost like moiré you might pick up from a screen or something. I don't think I've ever seen 'noise' quite like that.


certx 10 415 1 United States
17 Aug 2011 1:40PM
@ChrisV yup. it's a photo of 2 people standing in front of me. lol not a photo of a picture of them. While there is "obvious" noise in your photo above, I would have absolutely NO problem with taking that as is and printing it without any additional noise reduction. Mine? I might be willing to print it at 4x6 or something... but I'm not sure I'd print it at 8x10 even with NR. The NR would make it too soft by the time I was satisfied with the noise level.
ChrisV Plus
10 1.5k 26 United Kingdom
17 Aug 2011 1:51PM
AFAIK it's the same sensor - there's no way a 7D should be returning worse results than a 550D.

That's my 5DII in the shot and as I said earlier in the thread, I reckon the advantage in noise levels is only between one and two stops - it's not night and day!

I quite frequently take indoor shots with both cameras at ISO 3200 and the results, as you say, are quite acceptable printed, even at moderately large sizes.
17 Aug 2011 2:10PM
Firstly - I couldn't find you stating whether the camera was set to take in RAW or JPEG; please can you confirm. Both formats will load into DPP and save as JPEG. I'm pretty sure you took them RAW as checking the in-camera JPEGs, the noise is virtually non-existent (but so is the fine detail).

Secondly - and we are going to need some help here....
You have used DPP - whereas Chris and I have used ACR; they are fundamentally different conversion engines.
Looking at the [link=]60D DPReview comparison tests[/link] set to compare against JPEG at the same ISO - you can see similar artefacts to those in your crop, but no where near as pronounced. DPReview used ACR.
I'm wondering if the artefacts are a by product of DPP?
We need someone with DPP to have a go and post their findings.
certx 10 415 1 United States
17 Aug 2011 2:46PM
@DD Yes, it was shot in RAW:

shot in RAW at: ISO1600, aperture f/5, shutter 1/50.
opened in DPP, with no tweaks"

Smile I have opened it in ACR through PSE, and while it is a bit better looking, it's still very noisy. Looking at the ACR comparisons on DPReview, by the way, your link didn't work but I found the comparison. I will admit, it is possible that some of the artifacts are being produced by DPP, but the samples in the DPReview comparisons are WAY better looking than mine using ACR.

Here's the same photo using ACR in PSE. ISO Noise - ACR
17 Aug 2011 3:53PM
Most interesting!
The ACR crop looks WAY better than your DPP crop.
Opening them in separate browser tabs and flicking between them, the difference is remarkable!

Comparing your ISO 1600 crop with my ISO 2500 crop, they don't look a million miles away from each other. Biggest difference is the lighting contrast, my stage lighting is rather harsh; your is far mre natural but looks a dab of extra contrast wouldn't hurt.

Another thing to check - how old is your version of ACR?
There was a major change in the way it worked that was released along with LR3 and Photoshop CS5 - it's far far better at handling noise than olderr versions.

Looks like a big thumbs down for your version of DPP...
User_Removed 13 3.3k 4 United Kingdom
17 Aug 2011 4:43PM
Have you applied any Noise Reduction? High ISO image will probably need some.


certx 10 415 1 United States
17 Aug 2011 5:35PM

Quote:Have you applied any Noise Reduction? High ISO image will probably need some.

Hey CL. I haven't applied any *additional* NR, beyond the what is the default setting in DPP (the camera setting) of 3 for luminance noise. This was left intentionally not reduced further, as it is virtually the same noise that I get at ISO400. Which is what started this whole thread....the fact that my copy of the 7D seems to be problematic. At 1600 with luminance NR of 3 in DPP, it might be expected to still have a little noise, but at acceptable levels. At 400, there should be, in my opinion, virtually NO noise, and certainly with any NR at all, it should be noise free. But on my copy, the noise even at ISO400 is basically the same as it is at 1600.

Anyway, yes, there is slight NR applied in the DPP crop.
ChrisV Plus
10 1.5k 26 United Kingdom
17 Aug 2011 5:40PM

Quote:Most interesting!

Looks like a big thumbs down for your version of DPP...

I agree with Duncan - although there's still probably more noise than I'd expect, it's nowhere near as ugly as the DPP crop.

But I do have DPP on my Mac [v3.8.1] and so I decided to have a look at the same photo and crop via that.

I haven't changed any default settings [because as I don't use it I don't really know how!]

DPP has obviously boosted and readjusted exposure - there does appear to be noticeably more Chroma noise [if you look at the shadow under the camera it's most apparent]. But it still doesn't look anywhere near as bad as cert's example. Something strange seems to be happening...


User_Removed 13 3.3k 4 United Kingdom
17 Aug 2011 5:40PM
I know how little noise I get from my 5D II and also how quickly noise appears when upping the ISO on my 1D MkII. I can't believe you are getting such bad noise from your 7D, surely camera has a fault? Is there something environmental (interference from a mobile or mast, some strong infra-red close by etc.) that might explain the excess noise?
certx 10 415 1 United States
17 Aug 2011 6:09PM
I know exactly what you mean with the 5d2. The bride in my photos on the stairs is the same girl on the same day, that I used for this sample, but the stair shot was taken with my 5d2. That one is at ISO 3200, and has almost no noise in it at all before NR, and none after NR was applied with no notable loss of fine detail.

On the 7D, hmmmm.... I hadn't considered interference. I am never without my cell phone, or bluetooth, I suppose that could cause the problem, but I don't seem to have the problem with it on my 5D or on my 30D before I got the 7D. On the other hand, I also use the Pixel Knight TTTL triggers, but the noise seems to be there whether I have the knights on or not. BUT I certainly can't just dismiss interference out of hand. It's a thought I need to explore I think.
User_Removed 13 3.3k 4 United Kingdom
17 Aug 2011 6:37PM
If it was interference and it manifested itself when the sensor was at it's most sensitive it would affect your other cameras too. You need to pester Canon or a dealer. If you rule out workflow and other factors it has to be the camera.
certx 10 415 1 United States
19 Aug 2011 5:01PM
@CL et al, unfortunately, I agree. I'm afraid it's going to take come Canon/Dealer intervention. *sigh* Thanks all for all the great thought and advice and such. I guess on the bright side, I don't think it was anything *I* was doing. I guess there's that, at least.
certx 10 415 1 United States
4 Oct 2011 1:02PM
Ok, so... it's been quite some time I got back to this thread... so here's what happened. I did call my local Canon authorized center. The guy I talked to was not familiar with any problems like that with the 7D, but had some thoughts on the matter. I would not have thought this, but he did suggest checking my firmware, and updating it if needed before bringing it in. My thought was that it would not help with something like this, but coming from a software development background, I could see it as having possibilities. So I checked my firmware, I don't remember what the version number was, but it was VERY old. I updated the firmware, I have to admit, I was wrong in my thought that it wouldn't help. It is still noisy (of course, now I'm used to my 5DmkII), but the improvement on the noise was like night and day. I'm wondering if the version of firmware that was on my 7D was like a very first firmware version, or maybe an early prototype/development/beta stage firmware. Because, in truth, though it still looks like it's noisy to me, it is fathoms better. In fact, I wouldn't hesitate to use it, and now have, in a low light, indoor, wedding reception and the photos are all quite useable. I can only assume that there was something in the in camera processing/sensor data handling/sensor control that was faulty in the firmware it had.

I'm much pleased. Imagine. I simple firmware update. Don't I feel like a fool. lol

Thanks to all who responded and offered their thoughts, suggestions, and ideas.

Sign In

You must be a member to leave a comment.

ePHOTOzine, the web's friendliest photography community.

Join For Free

Upload photos, chat with photographers, win prizes and much more.