PDA

View Full Version : 50D Noise - what is normal



mikew09
28-05-2011, 9:48pm
Hi Team,
Noise - seems to be something I struggle to deal with. Been a lot of talk about the 50D and noise and it does raise itself as an issue now and then in my photos. I know I shoul dbe shooting with filters for sunset / sunrise photos and I am off to purchase a Cokin Grad filter kit on Monday. However, I am just wondering if what I am seeing in my 50D noise wise would be consiered normal

I admit it only seems to be an issue low light and not neccessarily long exposures but I keep second guessing myself if the level of noise I experience in my camera is normal or do I have a sensor issue.

Below photo with 100% crop was taken on AV (Aperture Priority) with my Tamron SP 17-50 @ 23 mm, ISO 100, f10, WB auto, evalutive metering.

Would other 50D owners and Canon owners alike consider the noise in the crop normal. Shot in RAW and as the lens flare would indicate no post processing apart from coversion to jpeg.

http://farm4.static.flickr.com/3577/5767941764_1df5b1aa76_b.jpg

http://farm4.static.flickr.com/3028/5767397855_f2af9a216d_b.jpg

mikew09
28-05-2011, 10:26pm
Oh, forgot to mention, no noise reduction set in the the camera settings either.

Scotty72
28-05-2011, 10:48pm
50d is often consider less capable of dealing with noise than its predecessor (40d) or the 7d but, it is still not bad.

The pics you posted, well, I'm guessing the major issue is bad exposure. Try to take a photo of a dark (but evenly lit) scene at 1600 with proper exposure.

mikew09
29-05-2011, 9:16am
50d is often consider less capable of dealing with noise than its predecessor (40d) or the 7d but, it is still not bad.

The pics you posted, well, I'm guessing the major issue is bad exposure. Try to take a photo of a dark (but evenly lit) scene at 1600 with proper exposure.

Thanks for the reply Scotty72. I will post a shot at ISO 1600 - for comment. Interesting comment about exposure, the camera was set at AV so are you saying that AV is not the best option for for sunset / sunrise type conditions. I see quite a number of sunrise / set shots that have a much better balance of light and still displaying a magnificent sunrise / set. I am assuming with the aid of filters but I still do not see evidence of high levels of noise. I have been struggling with this for a while and though frustrating, hope it is my technique rather than an issue with the camera.

Thanks,

Mike

mikew09
29-05-2011, 9:55am
Took a quick snap last night as was suggested by Scotty72 - interested to hear comment of other 50D owners also. Took the shot purely for 1600 ISO example and nothing else.

Shot at 1600 ISO, samer lens ( Tamron SP 17-50 f2.8), AV Mode, f7.1 @ 1/5sec - 23mm with evaluative mettering.

Excuse the mess on the bench, just cooked dinner for the family :-). No where near sharp but more interested in noise characteristics and no suprise considering the shtter speed. Just a quick snap to get it up.

http://farm4.static.flickr.com/3132/5769843550_d26ebf6cfa_b.jpg

Crop:
http://farm4.static.flickr.com/3563/5769304551_2c8c49dc5e_b.jpg

Would this be considered normal at 1600 ISO on a 50D?

Thanks in Advance,

Mike

terry.langham
29-05-2011, 10:13am
Do you get the same amount of noise from other lenses? I have noticed I get more noise on my 11-18 Tamron then my 70-200 f/4L or Sigma 24-60. Haven't yet been able to find an explaination, not that I have really looked, just put it down to cheap glass.

Terry

mikew09
29-05-2011, 10:23am
Another example with the crop pretty much centre of frame. Just took this last night hand held to play with settings, skill building activity. Just seems to me to be a lot of noise for ISO 100. No PP aprt from conversion to jpeg.

39 mm (Tamron SP 17-50 f2.8), 1/50 sec @ f7.1, ISO 100

http://farm4.static.flickr.com/3364/5769886394_d74bccc3c4_b.jpg

crop centre of frame:
http://farm4.static.flickr.com/3450/5769886764_b701932518_b.jpg

mikew09
29-05-2011, 10:35am
Do you get the same amount of noise from other lenses? I have noticed I get more noise on my 11-18 Tamron then my 70-200 f/4L or Sigma 24-60. Haven't yet been able to find an explaination, not that I have really looked, just put it down to cheap glass.

Terry

Can't say I tried the same conditions with my 70-200 - only have the Tammy (though is supposed to be the professional range SP) and my 70-200 L f4 now, my 50mm has since had a heart condition :-). Saving towards a 24-70 L 2.8 or 105 L at the moment, not exactly the lens for landscapes and such but with my limitation on cash will give me more value.

I see some magnficent photos with the 50D and realise the need for good glass, but will be very disappointed to a degree if it is the Tamron that is at fault. I find the Tamron an excellent lens in normal conditions with adequate lighting and it is sold as one of there pro series lenses.

example : http://www.ausphotography.net.au/forum/showthread.php?83061-Brisbane-River-Lowood

The old saying may apply here, you get what you pay for :-)

arthurking83
29-05-2011, 11:17am
Hey Mike.

just some observations from me and not specifically related to the 50D in any way either.

Noise is a strange thing in digital photography, and from your crops(remember without any experience on the 50D from my end!!) I see no real noise in those crops.

FWIW(and I don't know your post processing routine) but sometimes sharpening artefacts can look like noise.
If you add any sharpening routine, whatever it may be, have you compared an unsharpened image to a sharpened one?

Also, there is no way a lens can produce more noise in an image as a result of the lens itself causing more noise. The impact of a lens on the noise in an image is zero. Will never have any impact(in real terms).. BUT!! in saying that there is a visual illusion of what can appear to be noise.
If you delve deeply into the image, you can notice that a lot of images may have what appears to be noise in the unsharp areas, or in the areas where no detail exists. The blur in the image creates this optical illusion that noise appears there.
The noise levels are exactly the same, make no mistake about that. It's a weird concept to expect a sensor to produce less noise in some parts of the scene(say the detail areas) and more noise in other parts fo the same single shot scene(say the blur areas). it's just the way the brain works. Because it sees less detail, we expect to see more detail, so our brains compromise by looking for detail in other ways. That'd be grain detail.
So while a lens won't actually add to the noise numbers in a scientifically measurable manner(SNR value), it can(and they all can cause this optical illusion) where we seem to think there is more noise.
A lens can't do this. The same lens defocused slightly can also give this illusion as well.

If the image is not totally sharp there is this counter productive effort by many, where they subsequently add massive amounts of USM(or high pass) to try to process in a level of detail that's simply not really there.
Only add USM to images on the areas of the image where there is detail evident. Noise is not as apparent. If anything should be done to the unsharp areas, it should be NR, but usually it's good practise to leave them alone.
(or course I say this, and do the opposite myself, as I'm as lazy as the next chap when it comes to processing my images.. and prefer the easier one step processing method as well).
For images where I think it really counts tho, I do take the time to open it in a proper PP software, and use selective USM, where I would paint in some USM only on the areas that need it.

As an example, in your image #3 of the sunset, there should be no sharpening applied to any of the cloud area if you want the highest quality image!!
Because you have a lot more low frequency image detail over the entire scene, the best way to USM this image would be to paint a layer of editing across the horizon, and in that edit step use USM at a very low level.
Selective sharpening! :th3:
Alternate method to painting USM in, is to paint it out as well. Same principle, opposite application. You add USM to the entire image and use the - brush tool to erase the USM over the parts of the image with no detail.

As I see it, in your ISO1600 image, there is little to zero noise levels there. It looks as tho some noise reduction has been applied to the image tho! as there is some smearing of the detail areas compounded by a slow shutter speed and an OK hand holding technique(not quite excellent, but I reckon still better than my average hand holding technique! :p)

Also on a side note. Don't confuse the terminology of exposure and metering/metering mode.
Exposure is something you can control in most cases, and it's independent of what metering mode or shooting mode you use.

Av is perfectly acceptable for this scene(and I say probably the most useful shooting mode for landscapes too), but evaluative mode for metering may have not been the best choice.
You had two important variables to consider here, the dark areas and the blown out highlights.
A spot readingof the important areas you wanted to capture would have been a better method of metering.
That is, this is still an evaluative metering mode, but done by the operator, not the camera!
Take some readings using spot metering, commit them to memory and then working out an appropriate exposure value to use.

So if the camera chose 1/50s in the first image at f/10, I'm sure that you could have shot a little bit brighter say to 1/30 to increase the brightness of the darker areas, and still not blow out the detail in the sky across the horizon. The sun itself is always going to blow out. Nothing you can do about that, and it's of no real consequence anyhow.. it's the sun and is supposed to burn a hole in your eyes!
But in using spot metering, you would have had an accurate meter reading of the exposure requirements of the foreground(as an example it may have read 1/5s at f/10). You then make a conscious decision to under expose by a factor of X with the view to recover that area in PP when you get the images onto your PC.

A grad filter would have made no difference in this scene as it would have only served to darken the upper or lower third of the image even more.

fabian628
29-05-2011, 11:45am
these are good points.

I wonder, how does a lens affect noise? I cannot think how, apart from the lens causing the camera to meter slightly differently causing noise due to more underexposure.

Your first ppicture the noise is very low, i would say quite good.
The 2nd is harder to tell becuase the picture is also blurry. The problem with high iso noise is it also kills contrast and detail. If the picture is blurry it is hard to see detail which may be there.
Your 3rd picture looks ok also, but it looks like a big crop? or is it 100%? The tree line looks pretty soft bu the noise level looks ok.
I will dig up some old 50D pics for you and post them so you can see the comparison. :)

ricstew
29-05-2011, 11:50am
I discovered....that I shouldn't look so close :)

fabian628
29-05-2011, 12:00pm
here are two at iso100

http://i1023.photobucket.com/albums/af352/Natural_Defence_Plant2/2a/IMG_7350.jpg

http://i1023.photobucket.com/albums/af352/Natural_Defence_Plant2/2a/IMG_3111.jpg

mikew09
29-05-2011, 12:41pm
@Arthurking83,
Thanks Arthur for the detailed reply. I understandstand what you are saying - I am not a learned person and takes me a while to digest detail so will print this and read a few times on the train so I can apply. I do like my Tamron SP 17-50 2.8 but aspire to eventually have a L series fleet :-). I just can't fault my 70-200 and it pains me at times that it is too long for my lanscape shots.

One day this week with trustie tripod I will have another crack at this with the advise - thanks. For me photography is a very enjoyable but, well for me; a slow and sometimes frustrating skill to master so very much appreciate your in-put.

Thanks again mate,

Mike

@fabian628 - thanks for the reply Fabian, appreciated and thanks for psoting the images. Feeling pretty confident that it is my poor technique rather than the 50D or Tammy SP 17-50. May seem I am a bit paranoid about noise and my expectation is far outside my current skill set. Like most I imagine, I see so many cracker images on the AP and the web for tht manner that I hope to cut at least a few craker ones myself.
Thanks for replying appreciated, has put my mind at largly at ease about my muched loved 50D which took me 12 months to save for :-)

@Ricstew - haha lol, this cracked me up but I have to say, sound advise. Maybe in our own images we do look for something that does not warrant the effort. When looking at our own images the critic with the hardest critiques of all, is probably ourself. Thanks for the comment.

fabian628
29-05-2011, 1:05pm
No worries. This is the double edged sword of internet forums. They are very informative, but also cause you to become higly paranoid and can turn you to viewing your images at 100% looking for possible flaws.

mikew09
29-05-2011, 1:13pm
No worries. This is the double edged sword of internet forums. They are very informative, but also cause you to become higly paranoid and can turn you to viewing your images at 100% looking for possible flaws.

This is true - there is a possiblity that if I spent all this time out shooting instead of analyzing my imagines, I may have mastered a better skill by now :-)

Luna-blu
29-05-2011, 5:57pm
I've noticed that when comparing screen to print The noise is far less noticeable in print,
Could be due to zooming in on the monitor, or the different rendering of the two mediums.
Either way the point is to take heart, it may not be as bad as you think :th3:

mikew09
29-05-2011, 6:05pm
I've noticed that when comparing screen to print The noise is far less noticeable in print,
Could be due to zooming in on the monitor, or the different rendering of the two mediums.
Either way the point is to take heart, it may not be as bad as you think :th3:

Thanks - I was thinking the very same thing today - maybe I should have a couple printed at around A4 and see how they look. Thanks for that - I will give this a try.

Scotty72
29-05-2011, 6:16pm
Hi,

The pics at 1600 are that bad... I would tend to agree that it may be the illusion of noise due to poor focus / exposure as much as anything else.

I learnt pretty quick (but longer than it ought to have taken) that a well focused / well exposed image will elimate as much noise as the best noise filter.

Try the kitchen shot again, only this time - make sure the focus is dead on and up the exposure a bit (use a tripod to ensure no blur).

You'll be amazed.

Scotty

Luna-blu
29-05-2011, 6:23pm
Thanks - I was thinking the very same thing today - maybe I should have a couple printed at around A4 and see how they look. Thanks for that - I will give this a try.

I've seen them blown up fairly big (worked in print lab) but do it for yourself and see.
Agree with others about good exposure and focus, if it's not good in camera you can't do much in post.

Scotty72
29-05-2011, 6:25pm
By way of eg.

Here is a shot I took at our local ice rink: this place has dingy lighting at the best of times but, this day, about 5 or 6 of the overhead (out of about 24) were out - making it very dark.

Where I can normally, at this place, use ISO 3200 - I had to use 6400 and slow the shutter to 320 (very slow for ice hockey).

But, the exposure was enough to keep the noise acceptable.

This has a bit of noise PP but, it still was OK in RAW.
http://i1143.photobucket.com/albums/n639/sratcliff2/Ice%20Hockey/April%2010/_MG_9946.jpg
Scotty

mikew09
29-05-2011, 7:16pm
Hi,

The pics at 1600 are that bad... I would tend to agree that it may be the illusion of noise due to poor focus / exposure as much as anything else.

I learnt pretty quick (but longer than it ought to have taken) that a well focused / well exposed image will elimate as much noise as the best noise filter.

Try the kitchen shot again, only this time - make sure the focus is dead on and up the exposure a bit (use a tripod to ensure no blur).

You'll be amazed.

Scotty

Thanks Scotty, will do. Impressive example too - enforces your point very well.

Thanks

Scotty72
29-05-2011, 8:11pm
Thanks Scotty, will do. Impressive example too - enforces your point very well.

Thanks

The first sentence in my previous post (#18) should have read...

The pics at 1600 are NOT that bad...

:o:o:o

mikew09
30-05-2011, 7:54am
The first sentence in my previous post (#18) should have read...

The pics at 1600 are NOT that bad...

:o:o:o


Yea - I assumed that was what you meany - ta

mikew09
30-05-2011, 6:43pm
Yea - I assumed that was what you meany - ta

Now I am doing it :-) . It should say meant not meany...

Tannin
31-05-2011, 1:33am
Stop stressing out Mike!

The 50D (and many other cameras like it) produces a high resolution image. It is the nature of high resolution images that they include more noise. (Assuming any given lens and sensor area, that is.) If you have twice as many pixels, you have half as much signal per pixel, and something approaching twice as much noise. This is laws of pysics stuff, there is no way to avoid it.

But you CAN minimise it:

1: by having reasonable expectactions
2: by exposing well
3: by using the available noise reduction technology.

Note that (3) is NOT an option with today's high-res cameras. Now you can shoot in one of the dummy modes, in which case the camera will do it for you (and do a pretty decent job), or you can shoot in an "expert" mode and do your own NR processing as a routine part of your workflow. Up to you. But you cannot omit noise reduction if you want decent results. The camera is designed for it and requires it to produce the results it was built for. if it hadn't been designed that way it wouldn't be the very capable, detailed, versatile camera that it is.

Some other cameras, broadly similar in most respects, completely refuse to give you an image without NR processing, even if you convert your own raw file. (Nikons used to do this a few years ago, I'm not sure if they have stopped now.) Be thankful that the 50D does not: it gives you the choice whether you want NR, and if so, whether to use the in-camera system or roll your own.

Short answer: use the in-camera NR, or do it in your raw converter, or (best) use a specialist NR program such as Neat Image. You can get excellent results with any of those three methods. If you roll your own, be prepared to spend a little while learning what particular settings work best for you.

hus
05-06-2011, 10:18pm
Stop stressing out Mike!

The 50D (and many other cameras like it) produces a high resolution image. It is the nature of high resolution images that they include more noise. (Assuming any given lens and sensor area, that is.) If you have twice as many pixels, you have half as much signal per pixel, and something approaching twice as much noise. This is laws of pysics stuff, there is no way to avoid it.

What if I reduce the resolution on the camera to S Raw 1 on my 50D menu, would that reduce the noise or still give me the same noise as the full resolution?

Tannin
05-06-2011, 10:35pm
Well, yes, that will work for sure .... and I could lose weight for sure if I just cut a couple of my legs off at the thigh!

By reducing the resolution you are combining pixels and in doing that you average out the noise. But it's a terrible way to fight noise! As bad as cutting a leg off to lose some kilograms, and just as difficult to reverse if you change your mind.

A much, much better plan is to use software. This ain't cheating - the 50D was designed to be used in conjunction with noise reduction software - any of
(a) the camera's built-in NR, or
(b) the Canon raw converter software that came with the camera (which actually does exactly the same thing as the in-camera software, but being post processing is more flexible and can be tweaked, tuned, turned off, whatever you want
(c) third-party raw converter software such as Adobe Camera Raw
(d) third-party specialist noise reduction software such as Neat Image and Noise Ninja and various others.


You can get pretty much the same noise reduction, probably quite a lot more , and not sacrifice detail. (Depending on how aggressive you get with the noise reduction that is - you are free to select minimal NR for a more natural look (that's what I prefer) or heavy NR for a very clean but possibly rather plastic look, or anywhere in between.

Note that, in reality, you almost never view or print a finished image at 1:1 resolution - nearly always, it is downsized to fit on the screen or page. In other words, your brute-force noise-averaging technique with an sraw file is wasted as you were always going to shrink the file down before you used it anyway. All you have done is sacrifice detail that might (or might not) have come in handy.

Making sense?

hus
05-06-2011, 10:48pm
Just out of curiosity after reading this thread I tried the built in noise reduction, did 2 shots in jpeg nothing technical just a simple test and I was impressed with the amount of noise reduced, sorry forgot to take shot with built in noise reduction turned off but will do it monday night.

mikew09
07-06-2011, 7:40am
Thanks Tannin,
The explanation makes sense and I can take on board your comment to stop stressing. Have spent a little time playing with suggestions and start more seriously at NR software now.

Thanks to all who have assisted in this thead. Really appreciated.

mikew09
07-06-2011, 7:51am
Fair to say, as some had suggested I was fairly paranoid about noise. Thanks to all who helped with comments in this thread - the help has been extremely helpful, more so than trailing the web.
Proof to this great web site.
Oh Just an FYI - printed some images at A4 and no real noise as to what I see on the screen.

Thanks again team

GTP 290
09-06-2011, 11:06pm
A littl bit late but thought I'd add my little bit of info.


Oh, forgot to mention, no noise reduction set in the the camera settings either.

1: I have a 50D and I've got long exposure noise reduction setting in the menu. I'm not sure what shutter speed yiou are using but if it's anything that you can't shoot by hand I find this helps.

2: A really odd thing I have noticed with my 50D is that if I shoot in RAW and convert to JPEG the image has ALOT of noise no matter what ISO so I just shoot JPEG. Maybe you could shoot the same image twice, 1 in JPEG and 1 in RAW then convert and compare the 2 images?

Tannin
10-06-2011, 12:31am
A really odd thing I have noticed with my 50D is that if I shoot in RAW and convert to JPEG the image has ALOT of noise no matter what ISO so I just shoot JPEG.

This is because the JPG has had noise reduction applied to it by the in-camera firmware.

You can apply the exact same noise reduction to the raw by using Canon's DPP to develop it instead of whatever other raw conversion software you have. (You may have to change a setting somewhere in DPP.)

OR you can apply rather similar (but probably slightly inferior) noise reduction is most other raw converters. Adobe Camera Raw (comes with Photoshop and Lightroom), for example, has a noise reduction function. (not as good as DPP in my opinion, but good enough.)

OR you can use third-party specialist noise reduction for the best result of all. http://neatimage.com for my favourite one, Neat Imagge, but there are several other good ones as well.

mikew09
10-06-2011, 7:36pm
Thanks for the comments GTP 290 and Tannin. I have spent a little time practicing with my 50D based on advise in the thread. It has highlighted a lack of attention to detail on behalf and also that noise is un-avoidable at times and noise reduction is a mandatory part of the process when shooting at higher ISO. When I say attention to detail I mean my attention to the camera settings when take a shot and how important the right setting are to the end result.

Had a a koala visit the other day, very poor lighting and not much choice with back ground lighting to a point but I am pretty happy with the end result considering the ISO. Anyway, I will post a couple of shots at high ISO as example.

mikew09
10-06-2011, 7:58pm
I should mention - no NR PP has been done and no in camera NR.

Examples:

ISO 800

http://farm4.static.flickr.com/3572/5817892252_dea683c4f0_b.jpg

ISO 1600

http://farm3.static.flickr.com/2201/5817324373_99652d7b20_b.jpg

fairy bombs
11-06-2011, 7:19pm
No worries. This is the double edged sword of internet forums. They are very informative, but also cause you to become higly paranoid and can turn you to viewing your images at 100% looking for possible flaws.

So true-I think many people get hung up looking for flaws,rather than getting out and enjoying using their gear.

pmack
12-06-2011, 11:40am
Shooting in raw and not applying any noise reduction, you're definetely going to notice noise when viewing at 1:1
i'm a little concerened about this pic here though, i can see your sensor, notice the vertical banding?
I can definetley follow vertical lines for half the height of the photo. If that is an untouched photo, I wouldn't be all that impressed



39 mm (Tamron SP 17-50 f2.8), 1/50 sec @ f7.1, ISO 100

http://farm4.static.flickr.com/3364/5769886394_d74bccc3c4_b.jpg

Roosta
12-06-2011, 2:57pm
Hi Mike, I'll get those spec's/settings for you when I get home, sorry been of line for quite some time. One big thing I found with low light and landscapes particulary, was to find the sweat spot where my lens works best, for my Sigma 18-50mm F2.8 at low light/night, it's F5 and it is spot on, if I close it up a touch, I need to lengthen the shutter duration quite a bit, but in saying that, I haven't really had a need to go any longer than 5 - 10 seconds, except on some of the light painting shot I took in Kings Park, they worked with no NR on and at F5, MF is also a must I believe for getting the right exp, do you use MF or Auto for your lowlight/landscape shots?

I tend to stay around the 200 ISO, seems to be pretty sweat on my 50D, The shots I took whilst down south of the Milky way where upto 800 + ISO, but the exp time was up to 500 Seconds and several stacked images (All be it on LOUSHA'S Telescope tracking a specific star) which helped greatly, then I had long Exp NR on, same length of time to apply NR as shot taken, so it make for a long series of shots. That was the only time i've needed to use that function.

Hope this is of some help.

mikew09
12-06-2011, 3:00pm
Hi pmack and thanks for commenting. I am looking at this shot both on the forum and the raw shot in DPP and I cannot see the vertical banding you mention. Are you maybe using some valution tool. I of course be very concerned about this if I could see it and would have the cmaera in asap but I don't see it in any shots - I am interested if anyone is seeing the banding?

Hmmm - not sure what to make of this now.

mikew09
12-06-2011, 3:02pm
Hi Mike, I'll get those spec's/settings for you when I get home, sorry been of line for quite some time. One big thing I found with low light and landscapes particulary, was to find the sweat spot where my lens works best, for my Sigma 18-50mm F2.8 at low light/night, it's F5 and it is spot on, if I close it up a touch, I need to lengthen the shutter duration quite a bit, but in saying that, I haven't really had a need to go any longer than 5 - 10 seconds, except on some of the light painting shot I took in Kings Park, they worked with no NR on and at F5, MF is also a must I believe for getting the right exp, do you use MF or Auto for your lowlight/landscape shots?

I tend to stay around the 200 ISO, seems to be pretty sweat on my 50D, The shots I took whilst down south of the Milky way where upto 800 + ISO, but the exp time was up to 500 Seconds and several stacked images (All be it on LOUSHA'S Telescope tracking a specific star) which helped greatly, then I had long Exp NR on, same length of time to apply NR as shot taken, so it make for a long series of shots. That was the only time i've needed to use that function.

Hope this is of some help.

Thanks Roosta - all info is helpful.

Roosta
12-06-2011, 3:03pm
Not on either of my work monitors. But these probally not calibrated to be the best/most effective.

mikew09
12-06-2011, 3:11pm
Not on either of my work monitors. But these probally not calibrated to be the best/most effective.

Hmm - I don't see it at all. This monitor was calibrated many moons ago by a friend by I have had a laptop crash since then so I imagine it would not be valid. I have had one of the shots printed for my daughter at 8" x 12" and nothing noticeable but not sure if it was this exact image.
Thanks Roosta, lets see if anyone else notices banding.

@pmack, is there a way of testing for sensor banding issues, I regularly take a shot to check the status of dust on the sensor and not seen any irregularities there. I have seen a couple of threads where members have had banding issues on there 7D but the samples I seen were pretty obvious.

William
12-06-2011, 3:20pm
Hmmmm, I cant see any banding either :confused013

mikew09
12-06-2011, 3:25pm
Hmmmm, I cant see any banding either :confused013

Thanks William.

mikew09
12-06-2011, 3:28pm
Hi pmack,
OK, i have been playing with the image and by pushing the shadow, brightness, sharpness and contrast pretty hard, the noise in the dark area is pretty bad (actually really bad) but I would expect that and I think the exposure on this shot was pretty poor actually. However, I can now see the banding you speak of, sort of like lines of noise. I am assuming this is not an abnormly effect or is this an indication of a not so perfect sensor. I will post a shot shortly.

mikew09
12-06-2011, 3:41pm
Sample of shot with exaggerated settings to show noise and banding. Still pretty hard to see I think but they are visible.

http://farm3.static.flickr.com/2242/5823624396_07afff5c6e_b.jpg

http://farm3.static.flickr.com/2626/5823624314_00388ea1f9_b.jpg

mikew09
12-06-2011, 3:45pm
Not sure I should be concerned but will leave that to the many more experienced members on the forum - thanks pmac for the input in this thread, makes for a some interesting discussion.

One thing I do know, at some stage when I am happy with my photography I will be looking at a 5D mkII - or maybe the mkIII will be out by the time I have the cash :-).

pmack
12-06-2011, 5:59pm
no problems, could be my monitor shows the banding more for some reason, as it's quite obvious to my eyes.
there is a very obvious vertical line in the middle when i look at it, go straight up from the tall tree in the middle.

William
12-06-2011, 6:11pm
no problems, could be my monitor shows the banding more for some reason, as it's quite obvious to my eyes.
there is a very obvious vertical line in the middle when i look at it, go straight up from the tall tree in the middle.

Paul , From what I'm told thats not banding, I get those Vertical lines if I'm trying to pull to much out of dark areas, , Banding usually shows up as Big coloured arcs across the sky , I'll try and find an example , But those verticles are'nt banding , I get it if I try to pull the dark corners of Vignetting to even the sky

William
12-06-2011, 6:26pm
Some examples of the banding I was talking about , http://forums.dpreview.com/forums/readflat.asp?forum=1006&message=34171423&changemode=1

mikew09
12-06-2011, 6:44pm
Some examples of the banding I was talking about , http://forums.dpreview.com/forums/readflat.asp?forum=1006&message=34171423&changemode=1

Thanks William - I was under the same impression of banding but thought the term may be the same. Can't say I am concerned though about what we are seeing, I only see it in this shot (which is poorly exposed) and only when I push the PP well past the point of what I would call acceptable processing. What does interest me is why Paul is seeing it so obviously - seems this is not the case for others and certainly not for me, even when printed at 12" x 8".

pmack
12-06-2011, 11:26pm
Some examples of the banding I was talking about , http://forums.dpreview.com/forums/readflat.asp?forum=1006&message=34171423&changemode=1
ahh ok interesting. i've printed out a photo on a4 on a fairly basic canon printer (pixma mp640) and i got that type of effect on a sunrise shot.
i figured it was due to the limit of hue/colour variation of the printer, as i couldn't see it on the computer.
so i wonder what you call the vertical lines, maybe just sensor noise.

mikew09, did you lift this exposure at all or is the shot untouched?

mikew09
13-06-2011, 6:44pm
ahh ok interesting. i've printed out a photo on a4 on a fairly basic canon printer (pixma mp640) and i got that type of effect on a sunrise shot.
i figured it was due to the limit of hue/colour variation of the printer, as i couldn't see it on the computer.
so i wonder what you call the vertical lines, maybe just sensor noise.

mikew09, did you lift this exposure at all or is the shot untouched?

The last sample I have posted included a screen shot of the settings I pushed to reveal the noise lines - pretty extreme and not something I would usually do to a photo, the original (from memory), no changes apart from covert and the 100% crop.