What's with this 'banding' issue?

PhotoKhan

Forum Pro
Messages
12,028
Solutions
7
Reaction score
5,266
Location
Cascais, PT
Hi!

I decided to exchange my 350D for a 30D (Don't really know when, since the 30D seems to be in short supply around Europe, at the moment...).

Under "Cons", in Phil Askey's review final appraisa,l he states:

"AI Servo (continuous AF) interference banding issue (certain lenses, high sensitivities)"

What are your experiences with this?

Thanks for the inputs.

PK
--
“Loose praise may feed my ego but constructive criticism advances my skills”
************************************************************
-------------------------------------------------
http://www.pbase.com/photokhan
(Pbase Supporter)
 
...I had an idea of what the problem was...What I want to know is how it affects actual users in everyday situations.

I didn't know, however, that it was also present in the 5D...it's a 3000 USD camera, for god's sake...!!

It's funny how Chuck Westfall tries to mix things up when commenting on it:

"There may be cases where random noise or pattern noise [banding] stands out in images taken at high ISO settings. This is because the noise component is also amplified at high ISO settings. Various measures are taken to reduce noise, but the fact is that it cannot be entirely eliminated technologically. We are continuing to work on reducing noise even further"

Noise "per se" its not a problem if it has a random distribution pattern. There are even some people (...myself included) that feel it may cater for a creative twist, under some circumstances...

Noise in any visible pattern, however, is a completely different story as it may render a picture useless...

Now, he knows this as well as any of us...Why is he trying to mix things up, I wonder...;)

PK
--
“Loose praise may feed my ego but constructive criticism advances my skills”
************************************************************
-------------------------------------------------
http://www.pbase.com/photokhan
(Pbase Supporter)
 
The point was the noise that may be introduced with the particular lens listing. And that you can find more with some google searching.

How it translates to the 30D I can't say.

--
...Bob, NYC

http://www.pbase.com/btullis

You'll have to ignore the gallery's collection of bad compositions, improper exposures, and amateurish post processing. ;)

 
...I had an idea of what the problem was...What I want to know is
how it affects actual users in everyday situations.
If you shoot high ISO and use lenses known to exhibit the issue, on an everyday basis. You may have problems. I shoot lots of high ISO (mainly with the 85mm F1.8) and have rarely seen the problem. Like maybe 5 shots out of 20,000+. Or, 4 shots out of maybe 5,000 indoor action shots.

I got one the first week I had the camera. ISO3200 using B&W mode, indoors, with the 16-35mm. As I said, not too many after that.

http://www.pbase.com/slo2k/image/36208723/original.jpg
I didn't know, however, that it was also present in the 5D...it's a
3000 USD camera, for god's sake...!!
I think Chuck explains it well below.
It's funny how Chuck Westfall tries to mix things up when
commenting on it:
I think he did a good job. Yes, noise will be there at higher ISOs. Yes, it is acceptable (maybe even artsy) if randomly distributed. Yes, it is bothersome if there's a pattern to it. Yes, Canon CMOS sensors seem more prone to it than other mfgrs sensors. Other mfgrs sensors have their own issues :-)
"There may be cases where random noise or pattern noise [banding]
stands out in images taken at high ISO settings. This is because
the noise component is also amplified at high ISO settings. Various
measures are taken to reduce noise, but the fact is that it cannot
be entirely eliminated technologically. We are continuing to work
on reducing noise even further"

Noise "per se" its not a problem if it has a random distribution
pattern. There are even some people (...myself included) that feel
it may cater for a creative twist, under some circumstances...

Noise in any visible pattern, however, is a completely different
story as it may render a picture useless...

Now, he knows this as well as any of us...Why is he trying to mix
things up, I wonder...;)

PK
--
“Loose praise may feed my ego but constructive criticism advances
my skills”
************************************************************
-------------------------------------------------
http://www.pbase.com/photokhan
(Pbase Supporter)
-------------------------------------------------
Steve

--
Everyone has a photgraphic memory......some just don't have any film

http://www.pbase.com/slo2k
http://freezeframephotography.smugmug.com
http://www.photobird.com/steve
 
Yes, Canon CMOS sensors seem more prone to it than other mfgrs
sensors. Other mfgrs sensors have their own issues :-)
Banding (both normal readout banding, and AI-servo related banding) is more visible with low-noise Canons, because there is less random noise to hide it. Just about all digital cameras have banding. In Nikon D200 images I've looked at, there is actually stronger (normal) banding than in the Canon 20D, but it is less visible, because the banding is dwarfed by the random noise component.

--
John

 
Show an example of a well exposed shot with no pp exposure corrections.

--
...Bob, NYC

http://www.pbase.com/btullis

You'll have to ignore the gallery's collection of bad compositions, improper exposures, and amateurish post processing. ;)

 
You are right, sorry, I forgot to mention a very important little detail......banding shows in underexposed areas. But it ruins the the dark areas anyway.

I have a photo I took at a very dark place (very dim light, ISO 3200, 10mm, handheld at 1/10 s) , the histogram is to the left, the exposure is right because it looks how I want it to look, it shows the kind of light at the place, and the dark areas of the picture show banding.

I'm not making a big deal out of banding, but why should I have to settle with banding even if the photo is on the dark side? Lets not excuse canon for everything, I don't care about the noise because it's obvious that in an underexposed image we will see noise, I'm even happy about the amount of noise that shows at ISO 3200 it's really low compared to other products, but banding...hmmm, not sure.

I'll post an example later so you can tell me if I'm wrong.
 
...Exactly the combination I ordered today...:(

Add two more lenses I already have from that "problem lenses" list and I maybe hauted by this quite often, when using the described combination of AF mode and high ISO...

Darn...

PK

--
“Loose praise may feed my ego but constructive criticism advances my skills”
************************************************************
-------------------------------------------------
http://www.pbase.com/photokhan
(Pbase Supporter)
 
Lets not excuse canon for everything, I don't care about the noise
because it's obvious that in an underexposed image we will see noise, I'm
even happy about the amount of noise that shows at ISO 3200 it's
really low compared to other products, but banding...hmmm, not sure.
Exactly my feeling, if this turns out to be a recurrent issue...

PK

--
“Loose praise may feed my ego but constructive criticism advances my skills”
************************************************************
-------------------------------------------------
http://www.pbase.com/photokhan
(Pbase Supporter)
 
I think you're always going to get a bit of banding where you have motors and analog sensors.

Some lenses may have poor RF suppression on the motors.
 
You let ACR do the image adjustment, and it over compensated for an under exposed image. Look at it again, with all RAW settings set to 0, and show us what that looks like (unless this makes sense and you agree).

Note the histogram, where there's no data in the middle to upper range. Yet your image shows it increased RAW Exposure to make it 'presentable'.

--
...Bob, NYC

http://www.pbase.com/btullis

You'll have to ignore the gallery's collection of bad compositions, improper exposures, and amateurish post processing. ;)

 

Keyboard shortcuts

Back
Top