Ugly Hedgehog - Photography Forum
Home Active Topics Newest Pictures Search Login Register
Photo Analysis
JPEG Camera settings for Optimum Dynamic Range
Page <prev 2 of 4 next> last>>
Dec 21, 2022 09:22:05   #
rwm283main Loc: Terryville, CT
 
Wallen wrote:
The bit size and compression limits the dynamic range of a jpeg image. If the scene has a narrow dynamic range, this would not be a problem. If the scene has a wider dynamic range, you need to bracket the scene and merge it at post to capture as much as possible. Movement limit the scenes you can bracket. Anything that has movement can be tricky if worth the try.


Thanks for commenting Wallen. It's mostly the landscape scenes where there is a very wide dynamic range, (woodland and sky for example), and the jpeg images aren't to what I would consider keepers. I was hoping that there was an in-camera setting or two that would allow for more detail in the highlight areas. Trying to pull it out of a jpeg image afterwards in post processing leaves a bit to be desired. In these case I'll download the RAW images along with the jpegs since I'm using two SD cards. Simple enough to do.

Rich

Reply
Dec 21, 2022 09:23:19   #
Ysarex Loc: St. Louis
 
Wallen wrote:
The bit size and compression limits the dynamic range of a jpeg image.

JPEG is a final output format and it is print/output sufficient. Bit depth is not a limiting factor here. Cameras do not capture to JPEG they process to JPEG. Cameras can capture as much DR as their sensors permit. The JPEG bit depth places no limit on that.
Wallen wrote:
If the scene has a narrow dynamic range, this would not be a problem. If the scene has a wider dynamic range, you need to bracket the scene and merge it at post to capture as much as possible.

And then save that final merged image as an 8 bit JPEG in which case bit depth is not a problem. The two images I posted above are both 8 bit JPEGs: https://www.uglyhedgehog.com/t-760595-1.html#13626667 -- bit depth and compression are appropriate to the final output task.
Wallen wrote:
Movement limit the scenes you can bracket. Anything that has movement can be tricky if worth the try.

Reply
Dec 21, 2022 10:05:36   #
BurghByrd Loc: Pittsburgh
 
Ysarex wrote:
You can certainly make the effort to learn the DR function in your X-T2. It will be successful for less extreme cases. The problem is learning to make that assessment and so the effort at the camera always being sure to also capture a raw file -- the camera can only do so much. Here's an extreme example which shares a trait with your example posted above.

It's just a snapshot in the park -- first image the SOOC JPEG (resized) and then my processed image. No camera functions could achieve my processed image and that includes DR and automated HDR functions. The trait this image shares with your photo above is two different white balances. The sky in your photo and mine is daylight WB while the foreground is open shade WB -- simple thing for Capture One to handle but impossible for any camera that I know.
You can certainly make the effort to learn the DR ... (show quote)


Nice job Ysarex.

Reply
 
 
Dec 21, 2022 11:31:11   #
kymarto Loc: Portland OR and Milan Italy
 
rwm283main wrote:
Well, that answers my question. Thank You!
I suspect there's just too much jpeg compression. Just a thought.
Rich


It's not so much compression but the fact that the JPG has already thrown away a ton of information. I suggest exposing for the lighter tones as you would like them or slightly lighter, but then you will have to adjust brightness and shadows. SOOC you cannot get both highlights and shadows

Reply
Dec 21, 2022 12:38:49   #
imagemeister Loc: mid east Florida
 
kymarto wrote:
It's not so much compression but the fact that the JPG has already thrown away a ton of information. I suggest exposing for the lighter tones as you would like them or slightly lighter, but then you will have to adjust brightness and shadows. SOOC you cannot get both highlights and shadows


Yes, using a full frame camera with lower MP, using low ISO and exposing for the highlights and bringing up the shadows in post would be the best case scenario for max DR. on JPEG.

Reply
Dec 21, 2022 13:03:06   #
rwm283main Loc: Terryville, CT
 
imagemeister wrote:
Yes, using a full frame camera with lower MP, using low ISO and exposing for the highlights and bringing up the shadows in post would be the best case scenario for max DR. on JPEG.


Thanks for the advice. I’ll give that try. I’m headed out with my camera in a few minutes, it’s a bright sunny day with a few clouds. Rich

Reply
Dec 21, 2022 15:31:26   #
profbowman Loc: Harrisonburg, VA, USA
 
rwm283main wrote:
I'm shooting with Fujifilm cameras, I have an X-T2 and an X-S10. I'm trying to get the jpeg images SOOC close to what I can get from adjusting the equivalent RAW image. I'm especially concerned with the sky and how little range I have when attempting to get the jpeg image to look more like the adjusted RAW image. I'm using Capture 1 for my post processing. I use LR at times for post processing the jpeg images and I find that I am able to get the sky to look closer to what I end up with in my RAW image. Bottom line, is it possible to get the sky to look good straight out of the camera?
I'm shooting with Fujifilm cameras, I have an X-T2... (show quote)

Well, here is my attempt at making the sky have more pop. This may not be what you wanted, but I have not appreciated the photos in which the sky has been replaced with a strongly blue one. What you have here is so much closer to what we see most days when we look up. l I do like your composition. It has a lot of interest in a winter scene.

What I did to generate my version was to adjust the picture as a whole to what I thought the bottom section could use--a bit more gamma (just a bit darker. Then a too another copy of the image and concentrated on making the sky a bit darker with more contrast and then I noticed that green had become too dominate, so it was reducted. Then I cut out the foreground and pasted the sky back into the photo in which I had adjusted the foreground. Voilà!

What does anyone think? --Riichard



Reply
 
 
Dec 21, 2022 19:44:11   #
burkphoto Loc: High Point, NC
 
rwm283main wrote:
Well, that answers my question. Thank You!
I suspect there's just too much jpeg compression. Just a thought.
Rich


It isn't JPEG compression that is the problem. It's dynamic range compression that occurs when cramming a 12-14 stop dynamic range into the 5.5 or 6-stop dynamic range an 8-bit file like a JPEG can "contain."

Experiment with the available "Picture Styles" or "Picture Profiles" that the camera offers.

Try lowering the contrast in the camera menu.

Try turning on any sort of automatic dynamic range compensation the camera offers.

The problem you are running into is that NATURE can have a 20-stop dynamic range. The camera can capture 12 to 15 stops of that (depending on sensor and processor capability). The JPEG processor has to choose what "section" of that range to digitize as 8-bits. It "cuts out" the middle of a 12-14 bit range.

When you post-process with a parametric editor such as Lightroom Classic, you can lower whites and highlights, raise or lower exposure, and raise black value and shadows to "fit" the tonal values from the raw data into a better representation than the JPEG processor's range of menu controls will let you do in "pre-processing" (pre-exposure menu setting).

Reply
Dec 21, 2022 23:06:07   #
Jersey guy Loc: New Joisey
 
Ysarex wrote:
Computer algorithms AI or not can't second guess our intentions. In the end they have to apply rules based on averages. If you try and take an "average" photo they can handle it. Stray too far from average and the ball goes to your court.


No offense here, either, but that processed image has all the feeling of HDR work. Which causes me to ask a question of the more technically, scientifically inclined Hogs...how does HDR ultimately differ in capability than using the RAW path?

Reply
Dec 21, 2022 23:48:42   #
Grahame Loc: Fiji
 
rwm283main wrote:
I'm shooting with Fujifilm cameras, I have an X-T2 and an X-S10. I'm trying to get the jpeg images SOOC close to what I can get from adjusting the equivalent RAW image. I'm especially concerned with the sky and how little range I have when attempting to get the jpeg image to look more like the adjusted RAW image. I'm using Capture 1 for my post processing. I use LR at times for post processing the jpeg images and I find that I am able to get the sky to look closer to what I end up with in my RAW image. Bottom line, is it possible to get the sky to look good straight out of the camera?
I'm shooting with Fujifilm cameras, I have an X-T2... (show quote)

I will answer the question you specifically ask, which is, "Bottom line, is it possible to get the sky to look good straight out of the camera?"

Yes, of course it's possible to get the 'sky' to look good straight out of camera. But, there's a few things to consider.

a) You may be able to get the 'sky' looking good but do the in camera available adjustments allow you to get the rest of the image to your satisfaction at the same time?

b) What is 'good'?

c) When you make adjustments in camera to achieve your best possible for one scene are these same adjustments going to be suitable for your next image, and if not are you going to find the time to 're-adjust' a hindrance?

Reply
Dec 22, 2022 00:47:03   #
Wallen Loc: Middle Earth
 
Ysarex wrote:
JPEG is a final output format and it is print/output sufficient. Bit depth is not a limiting factor here. Cameras do not capture to JPEG they process to JPEG. Cameras can capture as much DR as their sensors permit. The JPEG bit depth places no limit on that.

The OP was "I'm trying to get the jpeg images SOOC close to what I can get from adjusting the equivalent RAW image".
The issue at hand is not about the sensor limits nor print/output sufficiency. It is getting a SOOC JPEG similar to a Post adjusted RAW. Hence, the camera JPEG capture settings, (compression/algorithm & bit size) against the scenes dynamic range, is what limits the OP in accomplishing the task in the camera.

Ysarex wrote:
And then save that final merged image as an 8 bit JPEG in which case bit depth is not a problem. The two images I posted above are both 8 bit JPEGs: https://www.uglyhedgehog.com/t-760595-1.html#13626667 -- bit depth and compression are appropriate to the final output task.

Almost everything captured at the end becomes a JPEG image. Merging bracketed photos or processing RAW do not increase the dynamic range of a JPEG image. What those does is provide cleaner data, on areas in the image which has little or none (A reason why some HDR photos look weird and unnatural).

Hence i'm going to repeat my conclusion; If the scene has a dynamic range within the capabilities of JPEG, this would not be a problem. Presented with a scene having a wider dynamic range, a SOOC JPEG will not be able to record it all. There is a need to bracket the scene and merge it in-camera(which is still SOOC as what he wanted) or merge the JPEGS at post* to capture as much range as possible.

*This is where shooting RAW is the better solution, but was not his question.

Reply
 
 
Dec 22, 2022 01:00:03   #
Ysarex Loc: St. Louis
 
Jersey guy wrote:
No offense here, either, but that processed image has all the feeling of HDR work.

It does, yes. I try not to stray too far in that direction but you're observation is well taken. I wonder would you have thought the same if you had not seen the camera JPEG. It's an intriguing question; I was there and certainly the camera JPEG is much farther from the reality of what I saw and experienced. Looking at the sky I saw blue with white clouds. Turning my glance toward the pond, foreground and trees they did not appear as if it was twilight and almost dark. It was the middle of the afternoon (without any direct sunlight there were no shadows to speak of). Glancing back and forth between the sky and foreground my perception adjusted in real time so that the HDRish look is more faithful to the actual experience.

Digital is giving us new tools. We have to think about how we use them. Your observation as such is very appropriate.

Jersey guy wrote:
Which causes me to ask a question of the more technically, scientifically inclined Hogs...how does HDR ultimately differ in capability than using the RAW path?

I dabbled a bit with HDR years ago. I stopped considering it useful when I started acquiring cameras that gave me a usable 10 + stops of DR in a raw file. That photo above is from a Nikon Z7 which provides 12 stops of usable data in a raw file. More than enough to already get me in trouble.

Reply
Dec 22, 2022 01:15:20   #
Ysarex Loc: St. Louis
 
Wallen wrote:
The OP was "I'm trying to get the jpeg images SOOC close to what I can get from adjusting the equivalent RAW image".
The issue at hand is not about the sensor limits nor print/output sufficiency. It is getting a SOOC JPEG similar to a Post adjusted RAW. Hence, the camera JPEG capture settings, (compression/algorithm & bit size) against the scenes dynamic range, is what limits the OP in accomplishing the task in the camera.

It's not the bit depth or compression used by the JPEG algorithm that is limiting. The camera's image processor is the limiting factor by design not by necessity. In the example I posted here: https://www.uglyhedgehog.com/t-760595-1.html#13626667 I saved the image I processed from the raw file as a JPEG. There's nothing about JPEG's bit depth or compression that limited my doing that.

Reply
Dec 22, 2022 06:13:27   #
Wallen Loc: Middle Earth
 
Ysarex wrote:
It's not the bit depth or compression used by the JPEG algorithm that is limiting. The camera's image processor is the limiting factor by design not by necessity. In the example I posted here: https://www.uglyhedgehog.com/t-760595-1.html#13626667 I saved the image I processed from the raw file as a JPEG. There's nothing about JPEG's bit depth or compression that limited my doing that.


If you mean that:
1. In camera: RAW capture ------> Jpeg = Jpeg low dynamic range SOOC
2. In Computer : RAW capture ---> jpeg = Jpeg high dynamic range PROCESSED
Hence both Jpeg is the same but camera is limited by processing power.


Then you are missing my point which is below:
1. In Camera: RAW capture ---> Jpeg = capture scenes high dynamic range & save as Jpeg SOOC

1B. In Camera: RAW capture + HDR setting ---> Jpeg = capture scenes high dynamic range, pull highlight low+lift shadow up & save as Jpeg SOOC

1C. In Camera: RAW capture + RAW capture + Merge ---> Jpeg = capture high dynamic range scenes in different exposure bracket, merge them & save as Jpeg SOOC

2. In Computer : RAW capture-> pull highlight low+lift shadow up+etc + etc ---> jpeg = capture scenes high dynamic range, adjust to one's contentment & save as Jpeg PROCESSED


In editing a RAW file outside the camera, the only difference is the capacity to finely tune the extra steps/personal adjustments which modify data before converting & writing the jpeg file. The same Jpeg algorithm applies and as previously mentioned, the dynamic range will all have the same potential range.

We tend to misuse the term high dynamic range because that does not really happen in the output. The potential dynamic range of a Jpeg file do not change. HDR processing is pulling what is out of bounds to be visible. No different from being able to hear a dog whistle or bat clicks by converting it to a lower pitch sound. It does not change our ears capacity, we still hear the same bandwidth.

The OP's problem is not a limitation of processing speed, Todays cameras are generally fast enough to have some form of HDR setting/processing both in single and multiple exposure(bracketing) mode. But that will be automatic with no further control hence it would not always make the photo better. No different from running random images on a batch file in PS. Some will turn out ok some will not.
The OP's problem is the scenes dynamic range in direct relation to what the file is capable of quantifying.

Reply
Dec 22, 2022 09:27:35   #
Ysarex Loc: St. Louis
 
Wallen wrote:
If you mean that:
1. In camera: RAW capture ------> Jpeg = Jpeg low dynamic range SOOC
2. In Computer : RAW capture ---> jpeg = Jpeg high dynamic range PROCESSED
Hence both Jpeg is the same but camera is limited by processing power.


Then you are missing my point which is below:
1. In Camera: RAW capture ---> Jpeg = capture scenes high dynamic range & save as Jpeg SOOC

1B. In Camera: RAW capture + HDR setting ---> Jpeg = capture scenes high dynamic range, pull highlight low+lift shadow up & save as Jpeg SOOC

1C. In Camera: RAW capture + RAW capture + Merge ---> Jpeg = capture high dynamic range scenes in different exposure bracket, merge them & save as Jpeg SOOC

2. In Computer : RAW capture-> pull highlight low+lift shadow up+etc + etc ---> jpeg = capture scenes high dynamic range, adjust to one's contentment & save as Jpeg PROCESSED


In editing a RAW file outside the camera, the only difference is the capacity to finely tune the extra steps/personal adjustments which modify data before converting & writing the jpeg file. The same Jpeg algorithm applies and as previously mentioned, the dynamic range will all have the same potential range.

We tend to misuse the term high dynamic range because that does not really happen in the output. The potential dynamic range of a Jpeg file do not change. HDR processing is pulling what is out of bounds to be visible.
If you mean that: br 1. In camera: RAW capture --... (show quote)

And then storing that result in a JPEG which works fine. The HDR "pulling" isn't done because of limitations imposed by the JPEG format. JPEG's bit depth and compression are not limiting factors. JPEG's bit depth and compression are not the reason we do the HDR "pulling" in the first place. Your original post; "The bit size and compression limits the dynamic range of a jpeg image." implies that JPEG is the limiting factor. It is not.
Wallen wrote:
No different from being able to hear a dog whistle or bat clicks by converting it to a lower pitch sound. It does not change our ears capacity, we still hear the same bandwidth.

The OP's problem is not a limitation of processing speed, Todays cameras are generally fast enough to have some form of HDR setting/processing both in single and multiple exposure(bracketing) mode. But that will be automatic with no further control hence it would not always make the photo better.

In many cases it will make the photo better and in some cases solve the OP's problem. The final output will be appropriately saved as a JPEG. The JPEG file structure is not then a problem.
Wallen wrote:
No different from running random images on a batch file in PS. Some will turn out ok some will not.
The OP's problem is the scenes dynamic range in direct relation to what the file is capable of quantifying.

No, the bit depth and compression used by the JPEG algorithm is not the limiting factor. You just said as much; Today's cameras often provide some form of HDR processing that may make the photo better. That better photo is then saved as a final JPEG. JPEG's file structure is sufficient to save that photo.

Consider: two of my cameras provide an option to save TIFF files instead of JPEG files. Changing the format from JPEG to TIFF in no way extends or alters what those cameras can do when capturing high dynamic range scenes.

Reply
Page <prev 2 of 4 next> last>>
If you want to reply, then register here. Registration is free and your account is created instantly, so you can post right away.
Photo Analysis
UglyHedgehog.com - Forum
Copyright 2011-2024 Ugly Hedgehog, Inc.