SOOC - Disappointing Exposure Issues - Please Help

[Colin Southern]

I think there is only one way that you could rescue this shot. Do 2 developments from the same raw file - one to give the water etc as you would like and then another set up to show detail in the dark areas. Looking at the histogram there is enough there to do this. Then blend the 2 with the windows enfuse gui which you will find freely available on the web. You may have to do this a couple of times to get what you want.

I posted a set of shots in my flickr account to show the general idea.

https://www.flickr.com/photos/489946...7631399852866/

Simple exposure compensation adjustments may give you what you need but the exposure for the dark areas may need a curve adjustment as well or some combination of the 2. One way of looking at the process is that Enfuse picks pixels based on contrast but the lightness of the dark end exposure may brighten the light end exposure. I would also advise using the default settings. These are all weightings for making pixel selections so may not have the anticipated results.

John
-

John,

I think you're grossly over-complicating it. In essence, all it needs is a little range compression using a fill light slider.

[ajohnw]

John,

I think you're grossly over-complicating it. In essence, all it needs is a little range compression using a fill light slider.

You could well be correct Colin but what I suggested is an easy way of getting the maximum out of the raw file. For one thing it gives user control of any needed contrast in the 2 shots. Actually it's often easy to do as well. The biggest problem really is anticipating the impact of bright area may have on the dark areas that are restored but often the software looks after that all on it's own given 2 shots that show things as they need to be.

John
-

[Colin Southern]

You could well be correct Colin but what I suggested is an easy way of getting the maximum out of the raw file. For one thing it gives user control of any needed contrast in the 2 shots. Actually it's often easy to do as well. The biggest problem really is anticipating the impact of bright area may have on the dark areas that are restored but often the software looks after that all on it's own given 2 shots that show things as they need to be.

John
-

Sorry John, but I don't feel that processing 2 difference versions of the same file and then merging them in a different program is "easy" for a beginner - especially compared to just moving a slider.

I think everyone is over-thinking this one a bit in looking for complicated solutions (exposure metering trade-offs, processing multiple versions etc), where in reality it's simply that the shadow detail needs to be boosted to reveal the detail, with perhaps a little mid-tone burning for local contrast enhancement.

[dragon76]

Hello All: I thought this would turn out better than it did, so I was really disappointed in it when I returned home to look at it.

Don't be disappointed as there is no camera in the world (at least not now) that could produce a raw or Jpeg image that looks as good as what was captured by your eyes.

Is this one of those situations where there is simply too much range in light to dark and that there would be no way to achieve detail out of the dark areas without completing blowing the sky and water?

Yes it is one of those situations. Yes it is possible to achieve detail out of dark areas and keep the sky under control but it can only be done effectively with multiple shots at different exposure values. Then you would need Photoshop to blend these shots together or a HDR program to merge and tone map the combined images. This can be quite challenging if you're not familiar with the mentioned software

I do want to visit this place again and would appreciate any tips. I can't change the direction that I would shoot into as the shoreline makes me face to the east. It's a big lake! I can, however, change the time of day that I go there. Also, if anyone has any PP tips, I would be happy to hear them.

For a single exposure shot, there is nothing wrong with your setting and matrix metering is your best bet in this situation. I am sure you have read the tutorial here about Matrix Metering and for Nikon, without going into technical details, before the shutter closes, it compares this scene to its database of some 30,000 images (through a complex algorithm process) and came up with the best possible exposure match for the scene and although it appeared flat on your display, it gives you room to recover details from both highlight and shadow which allow you to later process the image closer to what you saw.


Also, take a few shots in AF-A mode because it works best in MM mode for single shot exposure (apparently the algorithm calculation is more effective when the camera is in this combination). Just watch the focus points and make sure they are on your subject. Otherwise, shoot, shoot and shoot and you'll never know what MM could conjure up. It's digital you can always delete them later. Besides, you'd learn more from the failed shots than the good shots.

Otherwise, read up about HDR or image blending here. CIC has great tutorials regarding those techniques.

One more thing, was your WB set for daylight? as it looks a bit cool for 9.00 or 10.00 am?.

Cheers

[Colin Southern]

it can only be done effectively with multiple shots at different exposure values. Then you would need Photoshop to blend these shots together or a HDR program to merge and tone map the combined images. This can be quite challenging if you're not familiar with the mentioned software

Otherwise, read up about HDR or image blending here. CIC has great tutorials regarding those techniques.

Argh - no no no no no.

The issue isn't that the sensor can't handle the dynamic range - it handled it just fine. The issue is that we can't DISPLAY the range that was captured. At best all HDR multiple-exposure techniques would do is give some cleaner shadow detail.

[William W]

. . . it is possible to achieve detail out of dark areas and keep the sky under control but it can only be done effectively with multiple shots at different exposure values. Then you would need Photoshop to blend these shots together or a HDR program to merge and tone map the combined images. This can be quite challenging if you're not familiar with the mentioned software . . .

No.
Absolutely, positively, emphatically: no.

As already explained: the Dynamic Range of the Camera’s Sensor is more than adequate to capture the detail, shadow to highlight typical of that scene.

The problem is that the Dynamic Range of the resultant capture, cannot be adequately displayed on the screen. This fact also has already been articulated.

It is poor quality advice to recommend the use of shooting in HDRI for this type of scene.

WW

[dragon76]

Argh - no no no no no.

The issue isn't that the sensor can't handle the dynamic range - it handled it just fine. The issue is that we can't DISPLAY the range that was captured. At best all HDR multiple-exposure techniques would do is give some cleaner shadow detail.

Colin

I totally agreed with you. However, I never mentioned that modern DSLR sensor could not capture most of the dynamic range.
In the context of this thread though, I merely suggested that the most Effective way (that I know) to replicate closely to what Susan saw with her eyes is via HDR merge. This was suggested because matrix metering mode is heavily biased with the chosen AF point, especially in high contrast scenes. In this case, it was on the dark wood and Matrix metering would’ve behaved just like spot metering and would have blown out the sky to beyond recovery. Given the unreliability of MM mode in a single exposure shot, meter and capture each chosen subject and reconstruct the image in HDR software would be the most preferable choice in my view.
Cheers

[dragon76]

No.
Absolutely, positively, emphatically: no.


As already explained: the Dynamic Range of the Camera’s Sensor is more than adequate to capture the detail, shadow to highlight typical of that scene.

How could you have determined that? The human eye could see details upto 20 stops difference in brightness. Film or DSLRs can only manage 5-9 stops max. Most natural scenes have more tonal contrast than our cameras can sucessfully record. The only way to reproduce a decent image to the eye is to work with or manipulate the contrast through the use of Filters or blended exposures or both.

It is poor quality advice to recommend the use of shooting in HDRI for this type of scene.

Then what is your constructive advice for Susan. As purpose for most of us here is to learn more about photography.

Cheers

[Colin Southern]

Colin

I totally agreed with you. However, I never mentioned that modern DSLR sensor could not capture most of the dynamic range.
In the context of this thread though, I merely suggested that the most Effective way (that I know) to replicate closely to what Susan saw with her eyes is via HDR merge. This was suggested because matrix metering mode is heavily biased with the chosen AF point, especially in high contrast scenes. In this case, it was on the dark wood and Matrix metering would’ve behaved just like spot metering and would have blown out the sky to beyond recovery. Given the unreliability of MM mode in a single exposure shot, meter and capture each chosen subject and reconstruct the image in HDR software would be the most preferable choice in my view.
Cheers

Sorry Dean, but that just doesn't make sense. The purpose of bracketing exposures is to capture a wider dynamic range - but we've already captured a sufficient dynamic range. All HDR processing (ie tone-mapping) is going to do is squish what's been captured into something that can be displayed ... which is all that needs doing with the single frame already captured. So HDR capture and processing not required.

It really doesn't matter how or if the metering was biased in this example because the sky wasn't blown out beyond recovery - it was just fine. Looking at the histogram it's not blown at all.

[Colin Southern]

How could you have determined that? The human eye could see details upto 20 stops difference in brightness.

Film or DSLRs can only manage 5-9 stops max.

You're not comparing apples with apples; if the human eye is allowed to vary the size of the pupil to accommodate different levels of brightness (the 20 stops quoted is overall range, not instantaneous range) then you need to add that range of stops that the camera can accommodate by virtue of it's ability to vary both shutter speed and aperture.

Where did you come up with the "5-9 stops max" for a digital camera? Most are in the 11 to 12 and more range, at base ISO

Most natural scenes have more tonal contrast than our cameras can sucessfully record.

Sorry, but that's just not true.

The only way to reproduce a decent image to the eye is to work with or manipulate the contrast through the use of Filters or blended exposures or both.

Again, this just isn't accurate. GND filters aren't going to help in situation like this, and a blended exposure isn't going to help either because the original information captured was already fine. Again, this ISN'T a captured dynamic range issue - so no need to complicate things (especially considering the water movement in the image) with bracketed exposures; it's display dynamic range issue ie a processing problem. And to fix that problem - in essence - all that's needed is some dynamic range compression by virtue of some fill light ... as demonstrated in my quick makeover (from a jpeg screenshot!).

[dragon76]

Sorry Dean, but that just doesn't make sense. The purpose of bracketing exposures is to capture a wider dynamic range - but we've already captured a sufficient dynamic range. All HDR processing (ie tone-mapping) is going to do is squish what's been captured into something that can be displayed ... which is all that needs doing with the single frame already captured. So HDR capture and processing not required.

If this is the case, would there be any difference between taking multiple exposures of the scene and then merge them vs taking a single exposure and then bring it 2 stops up and down and merge them.

[Colin Southern]

If this is the case, would there be any difference between taking multiple exposures of the scene and then merge them vs taking a single exposure and then bring it 2 stops up and down and merge them.

There's really 3 scenarios here:

1. Multiple exposures: This is problematic because (a) there is movement in the scene, and (b) the energy in the reflected light coming off the water is similar to what's in the sky.

2. Merging 2 developments from a single exposure: This is just a waste of time because all the work required to split and then combine the images just takes you right back to the same dynamic range in the image that you had in the first place. Kinda like digging a hole and then filling it back up again; it doesn't create any data that's not already present.

3. Compress the range that you've already captured so that it's displayable. This is by far the easiest solution. In essence, "move fill light slider, job done".

#2 is a technique I do use from time to time; sometimes it's just easier than some of the techniques required to push/pull data from a single file, but usually that's a "no no" if there's movement in the transition zone.

[William W]

How could you have determined that?

Experience, about 40 years shooting professionally.
It's a typical backlit scene just after dawn or just before dusk.
A modern DSLR will capture that DR easily.

***

The human eye could see details up to 20 stops difference in brightness.

When the pupil of the eye changes size, yes, perhaps. See Colin's remarks.

***

Film or DSLRs can only manage 5-9 stops max.

That's just simple an incorrect statement.
Modern DLSRs manage 10 to 13 stops at their base ISO.
We are not discussing film and never have been, that's just a red herring.

***

Most natural scenes have more tonal contrast than our cameras can sucessfully record.

That is simply wrong information.
Most natural scenes are well within in 7 to 9 Stops compass often much fewer stops.

***

The only way to reproduce a decent image to the eye is to work with or manipulate the contrast through the use of Filters or blended exposures or both.

That's also simply wrong.
And the proof is in many of the images made by professional landscape photographers: for example Colin Southern.

***

Then what is your constructive advice for Susan. As purpose for most of us here is to learn more about photography.

My advice doesn’t deviate much from what has been mentioned in post #12, so my advice for Susan is to follow that advice: but it was not necessary to repeat that advice until now as an answer to your direct question.

The capture file appears OK. Perhaps maybe a smidge more exposure (that’s been mentioned by a few commentators) then after that it is a Post Production matter as Colin has explained.

Notwithstanding the above my other advice I have given to Susan and that was about using Manual Mode – as it occurs to me, by what she wrote, she does not / did not fully understand the nuances of the Camera Modes.

In addition, apropos the original image and 'advice', the bird would be moving. HDRI though not impossible to pull off, would NOT be "the only solution" simply because of that fact.

If your purpose is to learn about photography then it would bode well for you to revisit the source information that you acquired and based the many statements you consider as fact which you have presented here and have subsequently been called out as being incorrect. Because either your source documentation is wrong or you have incorrectly interpreted it, or both.

WW

[ajohnw]

I love the way you slate software you have never used Colin. It's often just a case of moving the exposure compensation for the 2 shots to merge. Enfuse isn't typical HDR software. It produces results similar to the methods you used on your parents house shot with a lot less work.

This may help the OP

https://documentation.apple.com/en/a...8%26section=14

It doesn't cover remapping the black point to make it brighter in the output. Basically that means lifting the dark end of the curve up and ideally moving the top of the curve to the left to retain the original slope of the line. The other way at that end is to move it down and then adjusting the contrast separately. The manual does give some idea of what adding more points does. The biggest problem most have with curves is making sufficiently tiny adjustments. If the curves panel can be enlarged it really is best to make it very large.

Much easier to do this sort of thing from a dexterity point of view with levels but Aperture's levels control is rather strange - the input range and output placement and range doesn't seem to be adjustable as per the GIMP for instance. If it is it would be a case of selecting the full input range - entire length of the histogram, setting a brighter output dark level, a dimmer bright level to leave some room for later contrast adjustments and then moving the input grey point slider to taste bearing in mind the need to adjust contrast some how later.

An easier option would be to use the free Nikon software that came with the camera and use the d-light slider. This would also enable the various tone curves - landscape etc to be tried in a manner that is much more suitable for a beginner - click and see what it does. Adjusting these sorts of things manually is not easy. This approach is also likely to produce a better quality initial raw conversion than other approaches leaving less need for other adjustments. Nikon pass a lot of information to their own software.

How well will any of these methods work? Without having the actual raw file it isn't really possible to say.

John
-

[dragon76]

There's really 3 scenarios here:

1. Multiple exposures: This is problematic because (a) there is movement in the scene, and (b) the energy in the reflected light coming off the water is similar to what's in the sky.

2. Merging 2 developments from a single exposure: This is just a waste of time because all the work required to split and then combine the images just takes you right back to the same dynamic range in the image that you had in the first place. Kinda like digging a hole and then filling it back up again; it doesn't create any data that's not already present.

3. Compress the range that you've already captured so that it's displayable. This is by far the easiest solution. In essence, "move fill light slider, job done".

#2 is a technique I do use from time to time; sometimes it's just easier than some of the techniques required to push/pull data from a single file, but usually that's a "no no" if there's movement in the transition zone.

Colin

The reason I asked because it was discussed in another forum

.....So in most cases a properly exposed single RAW file can contain as much information as a -3, 0, +3 series of JPEGs.
But some scenes contain even more than the 10-12 stops of dynamic range a single RAW file can record. If we shoot a series of RAW files we can expand the total amount of information from the darkest to the brightest parts of the scene even more. By shooting a -3, 0, +3 series of RAW files, we can add an additional 6 stops of dynamic range for a total of 16-18 stops! That means that the brightest exposure records highlights that are 262,144 times as bright as the shadows in the darkest exposure. The more intermediate steps of exposure we have, the better the software we use to combine the images can do so smoothly.

[Colin Southern]

.....So in most cases a properly exposed single RAW file can contain as much information as a -3, 0, +3 series of JPEGs.

Yes. JPEGs are optimised for small size and thus fast downloading; they discard a lot of information that can't be seen (minus a small safety margin).

But some scenes contain even more than the 10-12 stops of dynamic range a single RAW file can record.

Yes - they most certainly can, but this isn't one of them.

If we shoot a series of RAW files we can expand the total amount of information from the darkest to the brightest parts of the scene even more. By shooting a -3, 0, +3 series of RAW files, we can add an additional 6 stops of dynamic range for a total of 16-18 stops!

Yes - of course we can - that's basic HDRI - BUT - (a) this scene doesn't require it (most incident light sources like this only add about 3 stops to the dynamic range of the scene - the rest of the scene is only around 4 stops - so only about 7 or so needed), and (b) although one can capture 16 to 18 stops (or more) easily, the question then becomes "what do you do with them"? You can only print 4 and display around 6 -- so it becomes a real battle to squish the 16 to 18 into 4 to 6 and have it still look good.

The more intermediate steps of exposure we have, the better the software we use to combine the images can do so smoothly.

No - not really; It's noise driven. The more frames you have the more work the software has to do to try and align everything for no gain. Usually 2 to 3 stop deltas is fine. 3 frames with a 2-Stop delta will get you a cleaner result than 5 frames with a 1 stop delta. Above 4 and you're starting to push your luck though.

[Colin Southern]

I love the way you slate software you have never used Colin. It's often just a case of moving the exposure compensation for the 2 shots to merge. Enfuse isn't typical HDR software. It produces results similar to the methods you used on your parents house shot with a lot less work.

So tell me again John just how much work my parents house shot was for me? (I didn't realise you were watching when I did it).

[ajohnw]

So tell me again John just how much work my parents house shot was for me? (I didn't realise you were watching when I did it).

There's no telling what surveillance goes on these days Colin. Nobody can be sure just who is watching.

Perhaps I hacked Adobe's usage database.

John
-

[dragon76]

That is simply wrong information.
Most natural scenes are well within in 7 to 9 Stops compass often much fewer stops.

That's also simply wrong.

This was written on an online magazine by one of my favorite Landscape Photographers
. Darwin Wiggett
Do you have any references to back up your disagreement

If your purpose is to learn about photography then it would bode well for you to revisit the source information that you acquired and based the many statements you consider as fact which you have presented here and have subsequently been called out as being incorrect. Because either your source documentation is wrong or you have incorrectly interpreted it, or both.

Statements maybe but I never called them facts and I have learned from all my discussions with knowledgeable people in many forums, seminars and webinars and yes that is including most of the good tutorials here.
Maybe it's some people biggest fear but it never fazes me of being called out incorrect because that is when I learn most.
This is a forum and by its definition this is "where ideas and views on a particular issue can be exchanged" and I did just that. But I also understand quite well that every forum has an aggressive, non-constructive critic, or an obnoxious, or a bully or all of the above and that also does not faze me.

[ajohnw]

On Dean's post. The vast majority of decent cameras put 9 stops into jpg's these days and that is in the standard mode. More in other modes such as Nikon's D-Light which is something else as well. All cameras I am aware of now have modes which one way or the other will get more in and do other things at the same time.

To do this they re arrange tone levels just as a raw converter does. Just like a shot from a raw converter the tone levels can be post processed to suit requirements with one draw back - only 8 bit info at the dark end. The problem is that the lower stops go 0,1,2,4,8 etc and in real terms contain little information due to the numbers. Many people wouldn't even notice brightness changes of 0 to 1 to 2 or even up to 4 on their monitor. They would see 4 to 8 but seeing that as a "stop" is a bit dubious. The bright end is different - last stop 127 1/2 to 255, plenty of info to process. Even so it can be rather surprising how much can be recovered easily from a jpg with dark areas. Thing like D-Light use the whole info from the raw file and process the dark end in camera based round results from Nikon's N thousand photograph data base. For more control they provide a slider in their raw conversion software. This is the way all cameras seem to be going. Also adding in camera HDR.

There are some catches shooting jpg's with the view to processing them further. Mainly down to the curves used to produce them but one camera i have for instance allows me to shape it myself on the fly - not that I would. It's easier to add a bit of exposure compensation in the few instances where it's needed or rather a lot in some cases if I want to work from raw or need maximum highlight retention in difficult situations. Adobe? Well the conversion is done with Adobe curves or maybe they offer a pure linear conversion allowing users to do the whole thing themselves.

John
-