My view: The craft of photography may rise to art in the right hands. This craft has four parts: effective composition, good exposure, interesting subject, and perceived intention or purpose.

This practice assumes that photography functions as a wordless medium of human expression. In this realm, a given photograph offers a visual voice. A worthy photograph speaks for itself.

Understanding photography this way skips talk of objectivity and subjectivity, as if one over the other, when both apply.

All said, the medium of photography has room for every expression.

Thank you gentlemen for participating in this thread. RAW vs JPEG will go forever. Do not talk to me about Adobe RGB or Prophoto with outstanding colors, we cannot see them and not even dream about your monitor because it cannot see them either. sRGB is a short color space but if any of you can see the colors, well congratulations.

I increased the number of pixels I was shooting from Medium to Large and then I began to test the new files with Adobe RGB. To my eyes Large Fine are excellent files and they admit somewhat more manipulation than my previous JPEG did. Let me repeat it again, getting the act together in camera is preferable to doing it in post.

I have been shooting only RAW data but I am going to experiment a little more with JPEG, Large Fine. I know my camera is old and I am sure modern JPEGs are superior to what I have. If you shoot strictly RAW data and you are happy with your results so be it. I say the same to those shooting JPEG.
My Olympus Super Fine JPEGs are awesome as my eyes see them. Keep editing from none to low in post and the files should be fine.

Once again, thank you all for participating.

Delderby, thank you for stopping by and for your interesting comments. You are absolutely right, it has been my fault to post images with an old camera when I was supposed to do a better research of new cameras first. I apologize.
Sunsets, like any other subjects in photography are usually enhanced and the enhancement as a rule is done to colors. With JPEG files colors are pretty accurate if the camera was set up right. With my Olympus I just set it to Sunset and the JPEGs of sunrises or sunsets are spectacular and I do not have to edit anything. Using my Nikons I cannot do the same but again, mine are old cameras.

I shot the sunsets for illustration for a couple of reasons, first because sunsets are a favorite and also because of the tendency of JPEG to show banding in the sky. Let me repeat it again, I have acted as a fool using old cameras, the new ones I bet have a different approach to JPEG judging from what I know.
If a dark sky is not shot the JPEG does a great job with the image. Polarizers darken skies, I do not use them often with that purpose but they could bring up the same banding as those I see with my camera at sunset.

As I said, if critical work is to be shot RAW is the way to go.

This is one of the key points of editing JPEGs. Your middle and final images #2 & #3 show these issues clearly, even as a thumbnail embed. Hopefully, all those that don't (refuse to?) understand will look at this image and begin to understand the potential issues of a JPEG-only approach.

I emphasize potential as you have to be shooting & editing these shots to encounter this banding issue / limitation of 8-bit color editing. The less you need to push the 8-bit file, the less risk of color banding in the results.

In a 3-image post, 66% exhibit color banding. That's a pretty poor 'success' rate.

My XT-2 like all of my cameras is set to save RAW only. The camera still creates a JPEG and embeds it in the raw file. And if I want an OOC JPEG I can still have one by re-processing the raw file in the camera -- which I've done here in which case the JPEG was created by the camera software as FINE.

Here's a recent example (walk in the park). The first image below is my processed RAF file -- my photo. The second image below is the OOC JPEG (only resized for forum display), and the third illustration is the RawDigger stats and histogram for the RAF file.

In the OOC JPEG the highlights are blown -- the blue channel is blown. I set a +1.3 exposure comp when I took the photo and that would explain the blown highlights in the OOC JPEG.

The RawDigger stats and histogram indicate that I achieved my goal and captured a full capacity sensor exposure. I can't expose more and I have no reason to expose less -- nailed it.

With a digital sensor more exposure = better IQ with the only downside that if you go too far the error is fatal. It's a game of chicken. When the engineers design our cameras they are well aware of the stakes and they tend to be conservative chicken players especially with today's cameras that are pretty forgiving. So when deciding on the exposure that's going to make a good OOC JPEG they're inclined to engineer in some "highlight protection" which I would say is a very good idea. One way they give the maniac chicken players like me an option around that is to include an extended Low ISO value. The base ISO on my XT-2 is 200 but I do have the option to set it to 100 and that would pretty well negate the built-in "highlight protection." I just use the EC function in P mode to achieve the same thing (feels like I'm using my old Hassy so I'm comfortable with it).

The scene in this case is a good example in that it's high contrast lighting given the larger than average shadows in a sunlit exposure -- there's 10 stops of DR there. The full DR capacity of the XT-2's sensor is used to advantage. If I had exposed less I would have recorded less of that DR range. The only reasons I would have to do that are; I want the OOC JPEG -- I don't, or I'm bad at playing chicken -- I'm not.

It varies with the camera. I have four different camera brands. I treat them all the same in that I have only one exposure goal: place the brightest diffuse highlight at the sensor threshold and click. For every exposure my goal is to expose the sensor in the camera to full capacity. If that's not possible I get as close as I can. (I love digital. Every photo I get to expose the same as every other photo, it's so bleepin' easy!)

So I test my hardware to make sure I'm getting the result I want. Here's an example from my Fuji XT-4 which is one of my newer cameras. My processed version is on your left and the OOC JPEG on your right. Highlights are blown in the OOC JPEG. All the stats are in the RawDigger display below the photos. You can see that I have a +1.3 exposure comp dialed in with the camera at base ISO. The histogram indicates that I achieved my goal and exposed the sensor to capacity. Note the right end of the green channel histogram and the small straight line up -- green channel has begun to clip. .1% green channel clipping according to RawDigger and that's due to specular highlights (reflections off glass and metal) -- so that's a perfect exposure given my goal.

With my other cameras the discrepancy tends to be less with least for my Nikon and most for the Fuji. The Fuji is an APS-C sensor camera and 1.3 stops is significant enough to show a noise difference in a high DR image where I would want to lift the shadows, although arguably slight, as I originally said not a big deal just a little deal.

If you can't see it it ain't there.

You are out to lunch on my photo's. The last time you tried showing how raw would improve my photos you moronically picked a picture that WAS shot as raw, then provided ridiculous opinion on how the color was off.

I never claimed my photo's were anything special, most are shot with cheap equipment, many by my kids with cell phones, and not meant to be entered in some crazy photo contest, nor analyzed by some RAW ZEALOT light physicist with a spectroscope.

There are PLENTY of photo's both better and worse than what I've posted on the hog, and RAW has about NOTHING to do with any of it.

a) It is not about exposure, its about color... exposure bracketing or HDR will not help in this situation
b) This is not a rare as you may tink... it happens constantly.

One practical example: I had to make a bunch of product photos for an industrial magazine. Te product in case was a big outdoor structure made of galvanized steel with equipment attached. The sky was nice and blue... and the damn zinc was reflecting that nice blue so good that it was blue too.

Solution: mask the structure, leave alone the sky, grass and instruments. Correct WB of the structure until it is gray without ruining the sky.

Another practical example: blue hour beach. The sand *is* and *was perceived as white*... but it will have about 20% excess blue on the photo. Correct for the sand and the sky gets sick yellow.

Solution: create a layer named "sand" and correct that, leave the beautiful sky alone.

Botom line: why, in coder's heaven is so important to some people to lose information ?? The camera sensor produces 14 bit output, why throw it away !! I understand it for people that want to share immediately images on Instagram, but ... editing JPGs?? 8-bit VGA bit depth ???? Why ?

And again: we are having this discussion not because JPEGs are good... but because - from an engineering point of view - they are horrible. Had the industry adopted JPEG 2000 instead (only 9 years newer than JPEG which turns 30 years old next month), then JPEGs would have 16 bit color depth instead of 8 and nobody would even touch those RAW's.

Both the "jpeg algorithms" that cram 14 bits into 8 (destroying information) and the need to use RAW (to access that information) are known by the industry as temporary patchy workarounds to an obsolete standard. I hope that it will be solved by 2027 (when the next big conference will take place) and the damned 8-bit 1992's JPEG will be put to rest.

It's the same problem as the IPV4 Internet standard... we need to use zillions of NAT's just because IPV4 has only 32 bit addressing space and we already run out of addresses. We had IPV6 for years which has 128 bit space and solves all those problems, but is not widely adopted.

a) It is not about exposure, its about color... exposure bracketing or HDR will not help in this situation
b) This is not a rare as you may tink... it happens constantly.

One practical example: I had to make a bunch of product photos for an industrial magazine. Te product in case was a big outdoor structure made of galvanized steel with equipment attached. The sky was nice and blue... and the damn zinc was reflecting that nice blue so good that it was blue too.

Solution: mask the structure, leave alone the sky, grass and instruments. Correct WB of the structure until it is gray without ruining the sky.

Another practical example: blue hour beach. The sand *is* and *was perceived as white*... but it will have about 20% excess blue on the photo. Correct for the sand and the sky gets sick yellow.

Solution: create a layer named "sand" and correct that, leave the beautiful sky alone.

Botom line: why, in coder's heaven is so important to some people to lose information ?? The camera sensor produces 14 bit output, why throw it away !! I understand it for people that want to share immediately images on Instagram, but ... editing JPGs?? 8-bit VGA bit depth ???? Why ?

And again: we are having this discussion not because JPEGs are good... but because - from an engineering point of view - they are horrible. Had the industry adopted JPEG 2000 instead (only 9 years newer than JPEG which turns 30 years old next month), then JPEGs would have 16 bit color depth instead of 8 and nobody would even touch those RAW's.

Both the "jpeg algorithms" that cram 14 bits into 8 (destroying information) and the need to use RAW (to access that information) are known by the industry as temporary patchy workarounds to an obsolete standard. I hope that it will be solved by 2027 (when the next big conference will take place) and the damned 8-bit 1992's JPEG will be put to rest.

It's the same problem as the IPV4 Internet standard... we need to use zillions of NAT's just because IPV4 has only 32 bit addressing space and we already run out of addresses. We had IPV6 for years which has 128 bit space and solves all those problems, but is not widely adopted.