[BrianR]

I finally settled the "problem" ... direct to CS5 --- Camera RAW --- TIFF (or a high quality JPEG) = no banding !
As I also told Steve: also works with DCU5.
Cheers!

Excellent

Do try the different compression settings. Depending on your end use, the highest quality jpeg settings may deliver no noticeable improvement over more modest settings but at cost of a much higher file size (web use is definitely a case to dial back the quality, obviously not too far)

[jpzk]

Excellent

Do try the different compression settings. Depending on your end use, the highest quality jpeg settings may deliver no noticeable improvement over more modest settings but at cost of a much higher file size (web use is definitely a case to dial back the quality, obviously not too far)

Good idea.
Thanks!

[stevebrot]

TIFF it shall be.

Edit: I posted this after you contributed your improved result. I think you figured it out.

Probably Not needed. I do primarily landscape shooting and seldom (almost never) have problems with banding. The formula goes like this:

• Capture at a high bit-depth. For current market dSLRs that means a maximum of 14-bit color with capture to either RAW or TIFF. Note that many consumer-model cameras (e.g. K-50) capture at 12-bit maximum in RAW.
• Process at high bit-depth using non-destructive editing
• If you must reduce bit-depth, do so only as a final step when publishing to that file format (e.g. JPEG)
• Be aware that some plug-ins and popular features of some programs (e.g. Adobe Elements) are often 8-bit only. If you must use them, do the reduction in bit depth as a final step and not as a destructive edit.
• If you must apply additional JPEG compression, do so only as a final step when publishing to the compressed file.
• If you must capture as JPEG, take care to avoid radical changes to global contrast or abrupt curve modification. Avoid editing a compressed JPEG, the results are seldom good.
• Be aware that swaths of pure primary colors (red, green, blue) are more prone to banding than other hues

In practice, for most images and using my available tools this means:

• Shoot in RAW with my K-3
• Do most (almost all) edits in Lightroom
• If an external editor is needed, use LR's "edit in" feature to create a 14-bit TIFF or PSD for use in the external editor
• Export as JPEG when publishing to the Web or for e-mails or social media using minimal compression, if any

Steve

[jpzk]

Edit: I posted this after you contributed your improved result. I think you figured it out.

Probably Not needed. I do primarily landscape shooting and seldom (almost never) have problems with banding. The formula goes like this:

• Capture at a high bit-depth. For current market dSLRs that means a maximum of 14-bit color with capture to either RAW or TIFF. Note that many consumer-model cameras (e.g. K-50) capture at 12-bit maximum in RAW.
• Process at high bit-depth using non-destructive editing
• If you must reduce bit-depth, do so only as a final step when publishing to that file format (e.g. JPEG)
• Be aware that some plug-ins and popular features of some programs (e.g. Adobe Elements) are often 8-bit only. If you must use them, do the reduction in bit depth as a final step and not as a destructive edit.
• If you must apply additional JPEG compression, do so only as a final step when publishing to the compressed file.
• If you must capture as JPEG, take care to avoid radical changes to global contrast or abrupt curve modification. Avoid editing a compressed JPEG, the results are seldom good.
• Be aware that swaths of pure primary colors (red, green, blue) are more prone to banding than other hues

In practice, for most images and using my available tools this means:
Shoot in RAW with my K-3

• Do most (almost all) edits in Lightroom
• If an external editor is needed, use LR's "edit in" feature to create a 14-bit TIFF or PSD for use in the external editor
• Export as JPEG when publishing to the Web or for e-mails or social media using minimal compression, if any

Steve

Wow ! Quite a bit to digest (and learn) ...
I'll just hit the hay (common expression here) and get back to this in the morning.
I'm sure I'll be back with more questions !

[arkav]

Those images are actual Print Screen shots (direct from computer monitor) while viewing with FastStone viewer. The saved images were then saved as TIFF files and uploaded to Flickr.

I'm coming a bit late to the party, but I can confirm that by default, FastStone displays the embedded JPEG. Depending on the quality of the embedded JPEG, you may or may not see banding and other artifacts. I say that because my K200D embedded a really high quality JPEG in its RAW files. Banding and such was rarely evident. This is not true for my K30.

I use FastStone extensively - it's my browser of choice.

These days, I mostly shoot RAW+JPEG, so when I'm browsing/culling, I look at the JPEGs. If I want to perform simple edits for quick and dirty crops, I will edit the JPEG directly using FastStone. If it needs more sophisticated processing, I will work on the RAW using one of several Raw converters ( usually PDCU or DxO , then Elements if I need to work with layers ).

The fact that FastStone shows you the embedded JPEG is a big advantage - it makes accessing/displaying files so much faster. Others, with a different workflow may not like this feature.

[ghl]

FastStone also has a setting under the RAW tab in the Settings dialog (F12 on Windows, or Menu>Settings>Settings then RAW tab) where you can select whether FS will show the embedded JPEG or the actual RAW file.

[arkav]

FastStone also has a setting under the RAW tab in the Settings dialog (F12 on Windows, or Menu>Settings>Settings then RAW tab) where you can select whether FS will show the embedded JPEG or the actual RAW file.

But in my experience, that really slows things down, which sort of defeats the purpose. Also, AFAIK, you can't adjust any of the conversion parameters, so the results may end up looking very different from the Camera's JPG output ( none of your custom settings will be applied, no shadow or highlight compensation, etc. ).

What I like about looking at the embedded JPG is that it shows you ( approximately ) what you'd get from PDCU using the camera settings.