JohanEickmeyer wrote:
Just as a Canon-only shooter for so many years, the concept of lowered bit depth, cooked RAW files, and lossy compression are rather strange to me.
Just found this thread and I am impressed with the research efforts into the cause of this.
My first response was "Looks like insufficient bit depth in gradient".
When we first started working on digital files back in the 90's we were cautioned about the appearance of banding from trying to "stretch" the corrections too far. Solutions did include adding noise to the gradient to smooth it out.
It had the added benefit of making low res files look a bit sharper.
I am having the same banding with my A7R3. I have finally figured out when it happens but not why or how to fix it. It shows up with both my Sony 24-105 and Voightlander 40 mm, but under the same conditions and settings I do not see it with my adapted Zeiss 21 ZE and Zeiss 50 MP ZE using the Sigma adapter.
It shows up with both spot metering and highlight priority metering.
If I expose for the sky and bring up the exposure in post for the landscape, I do not see the banding. If I expose for the landscape and bring down the exposure of the sky in post, I have banding.
I shoot uncompressed RAW and convert with Image Data Converter Ver 5. I have worked up the JPG files in Adobe CS6, DXO and Affinity and get the same banding.
I have tried all permutations of the lens comp. settings and nothing changes.
What exposure setting do most folks use for landscapes?
It seems that your solution is to expose for highlights and bring up shadows.
In various fora I have seen that Sony and others using Sony sensors enjoy the ability to lift shadows dramatically but that highlights are more easily clipped. It would seem that by trying to darken a highlight that is close to clipping in one channel you will see banding more easily.
Canon, OTOH, seems to have configured their sensors to have a lot of highlight headroom but little shadow latitude. Thus we have been taught to "expose to the right".
I have astonished my Nikon friends by regularly "overexposing" my files in their estimation but can pull the detail right back in.
They maintain that if they did that with their Nikons they could not do that. Then again, they can rip open shadows that leave me with a gravelly mess.
Sep 15, 2018 at 12:13 AM
imagesfromobjects Offline • • • • Upload & Sell: Off
JimKasson wrote:
The file is about a stop and a half underexposed, even if you want to retain in the very bright highlight area. See the histogram attached.
If we look at the upper left corner stats, we can see the blue channel average is about 900. The black point is 512, so the net blue average is about 400. That's more than five stops down from full scale. Because of the PDAF patterning, the a7II can get banding in smooth gradients. It is more apparent in the blue channel. It is more apparent when the image is underexposed.
Looks like you hit the trifecta.
Jim
PS. What shape are the bands you're complaining about? If they're arcs, it could be the lens vignetting correction. ...Show more →
Totally O/T, but I wanted to say hello.
Your blog is incredible, and the tests you ran on the a7S were incredibly informative, even if some of the science was over my head. Pretty much sold me on getting that camera. I have some follow-up questions to some of the anomalies you observed, as well as some new odd things I've come across, but those can all wait for another day.
I'm suffering exactly the same problem with uncompressed raw files, and with silent shutter disabled. Heavy banding in blue skies. It's obvious when shots are underexposed by 1-stop, but on looking back I can see it on most shots taken with my 16-35 f4 that include blue sky.
Sample JPEG - see the concentric circles. The raw file exhibits the same problem.
