Reading mode: Light Dark

Sony A7III posterization banding - blue skies

Go to last post
English Rich said:
First of all - thanks to all of you for the responses. It's much appreciated.

I'm running a Dell IPS monitor that has been calibrated with a Spyder Pro. I've just disabled the calibration profile (to return the monitor to factory defaults) and guess what - the bands have practically disappeared! So much for running with a calibrated monitor.
Click to expand...
these banding claims get posted out here every few months, and they always resolve out to be issues other than the camera.

in my case it was the stock camera profile that dxo supplied, that was limiting how hard I could push a raw file.

that is a bit unusual... but it might be worth disabling the camera profile in whatever editing software you are using, and test a generic profile with the monitor calibration enabled.
 
English Rich said:
Recently I've noticed that many shots taken with my A7III (particularly using the 16-35 F4) contain very severe posterization banding in the sky. I'm shooting with silent shutter disabled, and to uncompressed raw files - the effect is clearly visible in both the raw files and the JPEGs. I initially noticed it in a shot that was underexposed by about a stop, but it's also visible in my daytime shots that are correctly exposed (although less obvious).

Has anybody else seen this? Any help would be much appreciated.

Here's a sample - uncompressed raw files present the same issues. Note the concentric circles in the sky.

https://flic.kr/p/2bV9RvE
Click to expand...
No banding on my monitor.

Try to open the file in Photoshop and use the eyedropper tool to check the actual color values on both sides of the "border".
 
When you see this behavior, try adding just a touch of Gaussian noise to the image, the dithering effect will get rid of the banding. Posterization is definitely something that can occur in gentle gradients like this when noise is very low as the range of available tones is not enough to fully describe the transition without enough dithering.
 
Last edited:
English Rich said:
First of all - thanks to all of you for the responses. It's much appreciated.

I'm running a Dell IPS monitor that has been calibrated with a Spyder Pro. I've just disabled the calibration profile (to return the monitor to factory defaults) and guess what - the bands have practically disappeared! So much for running with a calibrated monitor.
Click to expand...
Mine is calibrated( i1Display) and I see no banding, I would recommend that you double check your settings, both in your computer settings and in the Spider software when you run the profile sampling, it's possible the program is overriding the bit depth or something.

https://www.cambridgeincolour.com/tutorials/monitor-calibration.htm

Seems Dell and Spyder have some issues and that Dell doesn't support Spyder, I would recommend researching your monitor on google and the site below to see if there is a solution.

https://www.dell.com/community/Moni...yder5Elite/m-p/4665573/highlight/true#M102866

You should check with Spyder5Elite for support of their product. We only support the usage of our Dell X-rite calibration software and the X-Rite i1Display Pro Colorimeter.
English Rich said:
The strange thing is that the bands were visible even when viewing in monochrome with the calibration profile enabled. With it disabled they've disappeared from colour and monochrome versions.

The only bands that remain are of a different type - they are faint horizontal bands that run just above the horizon (in the orange gradient). It's entirely possible that these are down to the monitor too. There's no way that you could see them in the low res Flickr image.

Thanks again everybody.
Click to expand...
 
Last edited:
MILC man said:
English Rich said:
First of all - thanks to all of you for the responses. It's much appreciated.

I'm running a Dell IPS monitor that has been calibrated with a Spyder Pro. I've just disabled the calibration profile (to return the monitor to factory defaults) and guess what - the bands have practically disappeared! So much for running with a calibrated monitor.
Click to expand...
these banding claims get posted out here every few months, and they always resolve out to be issues other than the camera.

in my case it was the stock camera profile that dxo supplied, that was limiting how hard I could push a raw file.

that is a bit unusual... but it might be worth disabling the camera profile in whatever editing software you are using, and test a generic profile with the monitor calibration enabled.
Click to expand...
Yes and its bad timing with the banding issues being criticised on the EOS R and Z7.

For what its worth I have never seen banding in any shot with my A7r3.

But these are complex cameras with lots of settings and softwares so a particular combo could cause an issue.

I also could not see any concentric circles in the first sky image.

Greg.
 
Trollmannx said:
English Rich said:
Are you serious? It's as clear as day to my eyes. Concentric layers of arcs passing through the sky.
Click to expand...
Might be your screen - one of my previous screens (a cheapie) was notorious in this regard.

Did not notice any concentric layers on the screen used here...
Click to expand...
No rings on my high rez laptop monitor. It looks perfect. As far as the horizontal bands on the orange/yellow horizon shot its hard to say from a single example that can possibly be explained by - that's what the sky actually looked like when you took the image. Not invalidating the OPs claim but the sample image is not really compelling evidence.

Greg.
 
Is the raw file available somewhere?

vpilot
 
Astrophotographer 10 said:
Trollmannx said:
English Rich said:
Are you serious? It's as clear as day to my eyes. Concentric layers of arcs passing through the sky.
Click to expand...
Might be your screen - one of my previous screens (a cheapie) was notorious in this regard.

Did not notice any concentric layers on the screen used here...
Click to expand...
No rings on my high rez laptop monitor. It looks perfect. As far as the horizontal bands on the orange/yellow horizon shot its hard to say from a single example that can possibly be explained by - that's what the sky actually looked like when you took the image. Not invalidating the OPs claim but the sample image is not really compelling evidence.

Greg.
Click to expand...
The horizon image looks more like a mild inversion layer - typical here. Clear sky and some dust and humidity below the inversion.
 
I looked at them with two different monitors. With my high resolution monitor I saw less of an effect.

As one poster suggested; turn off in-camera adjustments. I like to do my adjustments in post.

When I see problems like this I walk through my workflow. Some image are more susceptible to problems with my ham-fisted approach. I back off my adjustments.
 
MILC man said:
English Rich said:
First of all - thanks to all of you for the responses. It's much appreciated.

I'm running a Dell IPS monitor that has been calibrated with a Spyder Pro. I've just disabled the calibration profile (to return the monitor to factory defaults) and guess what - the bands have practically disappeared! So much for running with a calibrated monitor.
Click to expand...
these banding claims get posted out here every few months, and they always resolve out to be issues other than the camera.

in my case it was the stock camera profile that dxo supplied, that was limiting how hard I could push a raw file.

that is a bit unusual... but it might be worth disabling the camera profile in whatever editing software you are using, and test a generic profile with the monitor calibration enabled.
Click to expand...
Exactly. When people complain about banding, esp. in the sky, it is necessarily user issues. Yes hardware limitations will not currently present the infinite range of shades that we see in nature,

but the issue is with DISPLAY technology, not with the imager device.The myth that continues to circulate on youtube over and over on youtube reviews, etc. is either ignorant or intentionally misleading exploitation of dynamic range limits of current tech, but such is not brand or model specific, and again it is a matter of compression and display technology not imaging devices. The presumption that there is tech that does it well is a fallacy, it is only a matter of correctly exposing and balancing, packaging the information well, and correctly color correcting grading giving the limitations of the display device.

"Banding" artifacts that are issues with lighting and electricity cycles/refresh rates and shutter speeds/ esp. slower electronic shutter is a completely different matter and is not seen in gradients of sky.Of course it is also solved by choosing the appropriate shutter speed, go figure.

It would be great if the photography channels made effort to educate on the matter.

Instead it is just a way to garner more clicks on a video as if it is a clever expose on a camera's flaws.

Don't believe the hype.

It's a poor craftsman that blames his tools........
 
Hi Rich,

Yes, I can see the posterizing issue in first of your picture with big part blue sky. Exactly in the right corner of the picture. I post here the right corner 100 % wihtout any change from original. The second picture is am minimal change with Photoshop. Lifting in the shadows/highlight box the 'shadow amount' 1% to show: it is not a problem with your monitor! The problem is in the picture itself. Once you discovered the posterizing in the photshop manipulated crop, you see it in the 100% crop original too.



4422e467b2fa47b1824535d6526d508e.jpg




33bed2ec7f144c729b97cfea36f70006.jpg


Kindly regards
 
German Guy said:
Hi Rich,

Yes, I can see the posterizing issue in first of your picture with big part blue sky. Exactly in the right corner of the picture. I post here the right corner 100 % wihtout any change from original. The second picture is am minimal change with Photoshop. Lifting in the shadows/highlight box the 'shadow amount' 1% to show: it is not a problem with your monitor! The problem is in the picture itself. Once you discovered the posterizing in the photshop manipulated crop, you see it in the 100% crop original too.

4422e467b2fa47b1824535d6526d508e.jpg


33bed2ec7f144c729b97cfea36f70006.jpg


Kindly regards
Click to expand...
No visible trace of the effect you describe in these images.

--
Ed Form
 
edform said:
German Guy said:
Hi Rich,

Yes, I can see the posterizing issue in first of your picture with big part blue sky. Exactly in the right corner of the picture. I post here the right corner 100 % wihtout any change from original. The second picture is am minimal change with Photoshop. Lifting in the shadows/highlight box the 'shadow amount' 1% to show: it is not a problem with your monitor! The problem is in the picture itself. Once you discovered the posterizing in the photshop manipulated crop, you see it in the 100% crop original too.

4422e467b2fa47b1824535d6526d508e.jpg


33bed2ec7f144c729b97cfea36f70006.jpg


Kindly regards
Click to expand...
No visible trace of the effect you describe in these images.
Click to expand...
It's probably somehow monitor, browser, and/or calibration dependent. I can see it on three different calibrated monitors using chrome.
 
Last edited:
German Guy said:
Hi Rich,

Yes, I can see the posterizing issue in first of your picture with big part blue sky. Exactly in the right corner of the picture. I post here the right corner 100 % wihtout any change from original. The second picture is am minimal change with Photoshop. Lifting in the shadows/highlight box the 'shadow amount' 1% to show: it is not a problem with your monitor! The problem is in the picture itself. Once you discovered the posterizing in the photshop manipulated crop, you see it in the 100% crop original too.

4422e467b2fa47b1824535d6526d508e.jpg


33bed2ec7f144c729b97cfea36f70006.jpg


Kindly regards
Click to expand...
It is visible if you look for it - and will be worse once you edit the image. It is consistent with what I see in my test images. Shades are sometimes not rendered smoothly but outside a cricle that almost fills the image frame in bands of color (magenta and cyan). How strong it is depends on the lens. a7r2 w/o AA-filter shows the effect a little stronger than the a7M3. I don't see this on my a6500.

Two images with FE28. First w/o adjusments, second adjusted in LR to make the colors more visible.



8b78d7507923473690a27e73956e9f54.jpg




73630cacbca9405980b8912b3b65dbd6.jpg
 
Last edited:
a7sastro said:
MILC man said:
English Rich said:
First of all - thanks to all of you for the responses. It's much appreciated.

I'm running a Dell IPS monitor that has been calibrated with a Spyder Pro. I've just disabled the calibration profile (to return the monitor to factory defaults) and guess what - the bands have practically disappeared! So much for running with a calibrated monitor.
Click to expand...
these banding claims get posted out here every few months, and they always resolve out to be issues other than the camera.

in my case it was the stock camera profile that dxo supplied, that was limiting how hard I could push a raw file.

that is a bit unusual... but it might be worth disabling the camera profile in whatever editing software you are using, and test a generic profile with the monitor calibration enabled.
Click to expand...
Exactly. When people complain about banding, esp. in the sky, it is necessarily user issues. Yes hardware limitations will not currently present the infinite range of shades that we see in nature,

but the issue is with DISPLAY technology, not with the imager device.The myth that continues to circulate on youtube over and over on youtube reviews, etc. is either ignorant or intentionally misleading exploitation of dynamic range limits of current tech, but such is not brand or model specific, and again it is a matter of compression and display technology not imaging devices. The presumption that there is tech that does it well is a fallacy, it is only a matter of correctly exposing and balancing, packaging the information well, and correctly color correcting grading giving the limitations of the display device.

"Banding" artifacts that are issues with lighting and electricity cycles/refresh rates and shutter speeds/ esp. slower electronic shutter is a completely different matter and is not seen in gradients of sky.Of course it is also solved by choosing the appropriate shutter speed, go figure.

It would be great if the photography channels made effort to educate on the matter.

Instead it is just a way to garner more clicks on a video as if it is a clever expose on a camera's flaws.

Don't believe the hype.

It's a poor craftsman that blames his tools........
Click to expand...
Take the time and look at this thread https://www.dpreview.com/forums/post/61927390 and then tell us again it is the monitor and/or poor craftsmanship that causes the red circle.
 
In-camera adjustments shouldn't affect the raw file (which the JPEG was produced from). That said, I think that I only had CA corrections enabled at the time I took the photo, although I can't be sure.

Since disabling the Spyder calibration the banding is barely visible - although I can still see it slightly on both monitors. That said - the monitors may still be to blame; just because I've removed the most obvious issue doesn't mean that it's solved.

If I see it again I'll try shooting with full mechanical shutter - first curtain and second curtain. I actually hope that it doesn't make a difference - closing the shutter before taking a shot introduces significant shake for landscape photos.

It will be interesting to see if the bands appear in test prints - I'll be ordering some prints soon, and I'll add this one to the order.
 
English Rich said:
In-camera adjustments shouldn't affect the raw file (which the JPEG was produced from). That said, I think that I only had CA corrections enabled at the time I took the photo, although I can't be sure.

Since disabling the Spyder calibration the banding is barely visible - although I can still see it slightly on both monitors. That said - the monitors may still be to blame; just because I've removed the most obvious issue doesn't mean that it's solved.

If I see it again I'll try shooting with full mechanical shutter - first curtain and second curtain. I actually hope that it doesn't make a difference - closing the shutter before taking a shot introduces significant shake for landscape photos.

It will be interesting to see if the bands appear in test prints - I'll be ordering some prints soon, and I'll add this one to the order.
Click to expand...
What program are you using to view the images? Some PC and Mac viewing programs employ a lot of "processing" as they write the image to the display and can sometimes give terrible results - Windows Photos is one such.
 
You must log in or register to reply here.

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