Reading mode: Light Dark

Banding and extreme PP

Go to last post
B

Big Ga

Forum Pro
Messages
18,643
Solutions
4
Reaction score
1,960
Location
Carmarthenshire, UK
Hi Folks,

I would like to know what the various mft cameras are like with regards to banding. Especially in the shadows. Mainly when you are doing fairly severe PP.

Now I'm sure there are a a whole bunch of people reading this who's instant reaction is "I don't have any banding ... and I've never seen any banding ... even with my older fourthirds cameras ....". If this is you, then please read no further, because ALL cameras will have some degree of banding (I've even seen it in my D3 cams), its just a question of when it kicks in and that depends on how far you're pushing the files. And if you aren't pushing the files as hard as I like to, then its probably going to be difficult for you to understand where I'm coming from.

I'm especially interested in using the older fourthirds cameras as a baseline reference. When shooting base ISO, the noise in shadows can be unpleasant, but banding can make a shot unusable. This is now my biggest issue with cameras like the L10, E520, E30, E620. The banding simply stops me pushing the files as hard as I'd like.

The G1/GF1 type of sensor doesn't really seem to have improved the noise that much over the E30/620, however I would say that the banding seems MUCH better. I can push the files a lot more.

The GH1 seems to be a LOT cleaner in the shadows in terms of pure noise ... but the banding is still there. Intensely annoying. I can pull shadows far better noise wise, but the banding often screws things up.

The G3 is a strange one. Its still a fairly noisy sensor. But .... they seem to have tamed the chroma noise really well. And ..... touch wood ... I think the banding is minimal as well.

But I prefer the smoothness of the GH1 and also the multi-aspect sensor. So I'm really curious as to what the GH2 is really like. I'm annoyed with panny that they took the ISO100 away (otherwise it would have been even cleaner again at base ISO), but if they've cured the GH1 banding, then it might be ok.

I'm also curious as to what the various Pen cameras are like? If they're like the newer panny cameras, they might be ok. If they're like the majority of the Oly E system cameras, they won't be!

So ... anyone here shot with the E system cameras and is aware of the banding I'm talking about - and then tried some of the MFT stuff, done similar PP to the files and either seen the same banding, or has been pleasantly surprised as to how much better the newer stuff is?
Any comments appreciated.

Cheers
Ga.
 
Your best bet might be to download the RAW files from the DPR studio compare tool and try pushing the thread spools in the shadow box yourself in a RAW converter of your choice. The RAW files are there to be downloaded for every camera and every ISO - it is an excellent resource. Imaging resource has another whole set of RAW files for your analysis as well.

As far as the GH2 ISO goes, keep in mind for RAW files the GH1 "over-exposed" a bit and so the ISO labels for the two cameras are a bit different. According to DxO the "RAW ISO" of the GH1 at 100 was actually 137 while the GH2 at 160 is 167. A bit less than a 1/3 stop difference. So I wouldn't worry about the ISO 100 setting being gone, they effectively have the same base ISO.

I don't have a GH1 to compare to my GH2, but in general I've seen people happier with the banding in the GH2 than the GH1. Many people claim no banding in the GH2, but as you say that's just because they haven't pushed the files insanely.
--
Ken W
See plan in profile for equipment list
 
For as far as I know banding is caused by pushing the shadows too hard... the best way to counter that is to increase expose to the max you can when you shoot the picture. No matter which cam you pick - push the picture far enough and you get banding especially in dark low contrast area's.

a cam with more dynamic range in the shadow area's might help, but I don't know how much you wanna push it... a sample of before & after might help a bit.

Obviously, no way you shoot in JPG, ever.
 
I mean, push the expose as far as you can WITHOUT clipping the highlights, of course.

Anyway if you pick m43, the G3 (or GH2/GX1) will outperform the other m43 camera's because it has a better sensor with better ISO performance and DR.
 
I have seen banding in highlights with little color. When shooting RAW and processing in 16 bit color, it is seldom and issue. This is on the E5.
--
Phil
 
kenw said:
Your best bet might be to download the RAW files from the DPR studio compare tool and try pushing the thread spools in the shadow box yourself in a RAW converter of your choice. The RAW files are there to be downloaded for every camera and every ISO - it is an excellent resource. Imaging resource has another whole set of RAW files for your analysis as well.
Click to expand...
That's a good suggestion and I've just done and downloaded what I can, however unless I'm mistaken, they only have the newer 'spools in the box' scene for the newer cameras - the GH1 isn't there for example.

Plus, while the DPR scene is sort of useful for trying to gauge sharpness and resolution, I'm not so convinced on its noise usefulness in extreme situations since the scene is so well illuminated. I have a feeling if one was to shoot a dawn/evening landscape with something like a 10 sec exposure, the noise would look a tad worse than the DPR studio shots ;-)

Cheers
 
always use 16bit files

either always use LAB colour

or

use RGB and set the adjsutment layers to either luminosity or colour or hue depending on what you want to acheive. Otherwise a curves correction (for example) will adjust colour AND luminosity and screw your images.

LAB is best.

always.
 
weesam said:
always use 16bit files

either always use LAB colour

or

use RGB and set the adjsutment layers to either luminosity or colour or hue depending on what you want to acheive. Otherwise a curves correction (for example) will adjust colour AND luminosity and screw your images.

LAB is best.

always.
Click to expand...
so how Lab (is it Lab or upLab btw ?) will help you w/ banding (take RPP raw converter, push m43 raw files many stops, see the result).
 
papillon_65 said:
http://forums.dpreview.com/forums/readflat.asp?forum=1041&thread=40267711&page=1

I pushed and pulled the GH2 by two stops at base ISO and didn't see any banding and that's further than you want to go on this camera.
Click to expand...
OK, I've just looked at those comparisons. Interesting. This is the kind of side by side test I'd want to see (especially against the 16MP sony sensor which is my gold standard benchmark at the moment).

Those results do give me some hope, however that particular test is rather flawed in terms of what I'm looking for.

Firstly, because the main brown bark of the tree which is still brown after being pulled back, is inherently of such a random nature, that its very difficult to spot noise. and secondly, the dark interior of the inside of the hollow trunk is clipped at absolute black, so there isn't anything to pull back.

The types of scene where banding becomes obvious and unpleasant are areas of solid colour - especially things like dark blue skies or still water.

The classic type of test that you really want to do if you're interested in seeing how much you can pull back (and this would show the banding as well) is to shoot an interior room (with plain walls!), expose for the bright windows, then pull up the walls by say three stops to equalise the exposure using the fill light slider in LR/ACR. That'll separate the men from the boys ;-)
papillon_65 said:
I haven't seen any banding on my GH2 yet.
Click to expand...
Well I've had a nice cup of tea, and I do have hope that the GH2 is indeed fairly band free. Its just that I've had years of watching people post the most hideously banding riddled low light concert shots over on 1022 where 90% of the commentators over there just gush with "wow .. fantastic picture there ....." and if anyone dares comment on the quite obvious and horrible banding , its a case of "banding ... what banding ?" or the shooter goes off in a prima donna huff even though its the gear and not his photography being criticised ...
papillon_65 said:
You do also know that Nik Define and some other noise reduction software will remove banding quite easily?
Click to expand...
I have banding removing NR software, but while it can indeed get rid of some occurrences, it can't quite remove the larger type of yellow banding I can get on the GH1 - not without leaving a colour cast.
 
papillon_65 said:
http://forums.dpreview.com/forums/readflat.asp?forum=1041&thread=40267711&page=1

I pushed and pulled the GH2 by two stops at base ISO and didn't see any banding and that's further than you want to go on this camera. I haven't seen any banding on my GH2 yet. You do also know that Nik Define and some other noise reduction software will remove banding quite easily?
--
It's a known fact that where there's tea there's hope.
Tony
http://the-random-photographer.blogspot.com/
Click to expand...
From my testing, dfine banding reduction will only work some of the time, and then not necessarily remove it 100%.
 
weesam said:
always use 16bit files

either always use LAB colour
Click to expand...
These won't help the OP at all because he's asking about fixed pattern noise banding. You are giving recommendations to prevent posterization which is a completely different issue. Both are sometimes referred to as banding, so your confusion is completely understandable!

--
Ken W
See plan in profile for equipment list
 
Big Ga said:
I have a feeling if one was to shoot a dawn/evening landscape with something like a 10 sec exposure, the noise would look a tad worse than the DPR studio shots ;-)
Click to expand...
Ah, yes indeed. In fact things get so different at long exposures (dark current limited) that it isn't unusual for two cameras being compared to result in a different one coming out "on top" depending on short vs. long exposure. I've heard people complain about the dark current noise on the Sony sensors, but I haven't heard "compared to what" or seen any specific examples.

In theory, with good dark frame subtraction, the banding shouldn't get much worse at long exposures though.

Anyway, now I completely understand why the "standard" shots don't answer your question!
--
Ken W
See plan in profile for equipment list
 
exdeejjjaaaa said:
weesam said:
exdeejjjaaaa said:
exdeejjjaaaa said:
so how Lab (is it Lab or upLab btw ?) will help you w/ banding (take RPP raw converter, push m43 raw files many stops, see the result).
Click to expand...
Click to expand...
Click to expand...
Click to expand...
exdeejjjaaaa said:
you didn't...
Click to expand...
exdeejjjaaaa said:
weesam said:
....otherwise a curves correction (for example) will adjust colour AND luminosity and screw your images.
Click to expand...
Click to expand...
okay? comprende?
 
weesam said:
exdeejjjaaaa said:
weesam said:
exdeejjjaaaa said:
weesam said:
so how Lab (is it Lab or upLab btw ?) will help you w/ banding (take RPP raw converter, push m43 raw files many stops, see the result).
Click to expand...
Click to expand...
Click to expand...
Click to expand...
weesam said:
you didn't...
Click to expand...
weesam said:
weesam said:
....otherwise a curves correction (for example) will adjust colour AND luminosity and screw your images.
Click to expand...
Click to expand...
okay? comprende?
Click to expand...
just reread kenw postings above this one of yours and below my post that you quoted - he explains that very nicely for folks like you, dude...
 
Big Ga said:
papillon_65 said:
http://forums.dpreview.com/forums/readflat.asp?forum=1041&thread=40267711&page=1

I pushed and pulled the GH2 by two stops at base ISO and didn't see any banding and that's further than you want to go on this camera.
Click to expand...
OK, I've just looked at those comparisons. Interesting. This is the kind of side by side test I'd want to see (especially against the 16MP sony sensor which is my gold standard benchmark at the moment).

Those results do give me some hope, however that particular test is rather flawed in terms of what I'm looking for.

Firstly, because the main brown bark of the tree which is still brown after being pulled back, is inherently of such a random nature, that its very difficult to spot noise. and secondly, the dark interior of the inside of the hollow trunk is clipped at absolute black, so there isn't anything to pull back.

The types of scene where banding becomes obvious and unpleasant are areas of solid colour - especially things like dark blue skies or still water.

The classic type of test that you really want to do if you're interested in seeing how much you can pull back (and this would show the banding as well) is to shoot an interior room (with plain walls!), expose for the bright windows, then pull up the walls by say three stops to equalise the exposure using the fill light slider in LR/ACR. That'll separate the men from the boys ;-)
Click to expand...
At base ISO you can pull the shadows by at least two stops and you won't see any banding on the GH2. It reminds me of the E-330 in this respect, it's dynamic range is biased towards the shadows. Go up the ISO range and you may well be able to induce banding, I guess it depends what ISO you intend to stress it at.
Big Ga said:
Big Ga said:
I haven't seen any banding on my GH2 yet.
Click to expand...
Well I've had a nice cup of tea, and I do have hope that the GH2 is indeed fairly band free. Its just that I've had years of watching people post the most hideously banding riddled low light concert shots over on 1022 where 90% of the commentators over there just gush with "wow .. fantastic picture there ....." and if anyone dares comment on the quite obvious and horrible banding , its a case of "banding ... what banding ?" or the shooter goes off in a prima donna huff even though its the gear and not his photography being criticised ...
Click to expand...
I've shot in low light at ISO 3200 and didn't see any but then I'm careful with my exposures in that kind of light.
Big Ga said:
Big Ga said:
You do also know that Nik Define and some other noise reduction software will remove banding quite easily?
Click to expand...
I have banding removing NR software, but while it can indeed get rid of some occurrences, it can't quite remove the larger type of yellow banding I can get on the GH1 - not without leaving a colour cast.
Click to expand...
--
It's a known fact that where there's tea there's hope.
Tony
http://the-random-photographer.blogspot.com/
 
You must log in or register to reply here.

Similar threads

V
Dark band along the right border of RAW files (Olympus E-M5 III)
2
Replies
28
Views
3K
John Sheehy
John Sheehy
MJ_Photo38
Olympus E-M10 mark II surgery !
Replies
1
Views
2K
johnCam
johnCam
F
  • Question Question
Olympus OM-D E-M5 II vs Panasonic Lumix DMC-GH4
Replies
6
Views
634
Furqan Altaf
F
TumblingTiger
Olympus 14-42mm F3.5-5.6 vs. Leica 15mm f/1.7 or Lumix G 20mm f1.7
Replies
16
Views
6K
17mmX2
1
D
High ISO comparison between 16mp and 20mp olympus sensors
Replies
4
Views
241
Raist3d
R

Keyboard shortcuts

f
Forum
m
My threads
Mobile site
About
Editorial content
Cameras & Lenses
Community
All content, design, and layout are Copyright © 1998–2025 Digital Photography Review All Rights Reserved.
Reproduction in whole or part in any form or medium without specific written permission is prohibited.
When you use DPReview links to buy products, the site may earn a commission.
©GPS Media - Guides, Products, Services.
Back
Top