Fixing high-contrast lighting

I have several disappointing photos which I'd like to try fixing, but cannot find any practical references for how to go about that. Specifically, do I need to adjust levels, brightness/contrast, exposure, shadows/highlights, gradients, etc.? I don't want to know how to use a photo editor's tools (there are plenty of tutorials), but how to apply the tools to my needs.

 

I tried uploading a single sample, but I can't see it in the preview. I hope it gets attached; otherwise I'll try with a follow-up post. The lighting on the desk in the center is just right, but I'd like a little brighter yard through the blinds on the left, and if possible, I want to bring out a little more of the china cabinet on the right.

 

I have additional images with similar problems. Are there any guidelines from this example that I can apply to other work going forward?

 

Here's the EXIF data. My photo editor is Affinity Photo version 1.10.

 

It is capable of HDR merges, but I didn't like the results of any of the combinations of the four RAW images I have.

 

 

If you shot raw, we can't even start discussing exposure without access to the raw and viewing a raw Histogram in something like RawDigger.

If you shot JPEG, that's the start of the issue! Ideally, you want to stick with the raw data. The JPEG engine that processes the raw massively clips and compresses highlights. We often don't when editing the raw. This compression can clump midtones as much as 1 stop while compressing shadow details! People incorrectly state that raw has more highlight data but the fact is, the DR captured is an attribute of the capture system; it's all there in the raw but maybe not in a camera proceed JPEG.

 

A raw capture that's 10 or 11 stops of dynamic range can be compressed to 7 stops from this JPEG processing which is a significant amount of data and tonal loss! So when we hear people state that a raw has more DR than a JPEG, it's due to the poor rendering or handling of the data to create that JPEG. The rendering of this data and the reduction of dynamic range is from the JPEG engine that isn't handling the DR data that does exists as well as we can from the raw! Another reason to capture and render the raw data, assuming you care about how the image is rendered!

I am definitely shooting in raw. I'll take a look at RawDigger.

 

Affinity Photo has a histogram graphic that I can capture for all channels, or any of the individual red/green/blue channels.

 

There is also something called scope, which has views of intensity waveform, RGB waveform, RGB parade, power spectral density and vectorscope. I would guess that some of these are specific to this software, but I can provide those views as well if they would tell you something they don't tell me.

Affinity Photo's Histogram isn't a raw Histogram. Like so many (Adobe's too), it represents the rendered image in a color space. That isn't what RD is doing. As such, it is one of the very, very rare tools that shows us the actual effect of exposure on the raw data. RD shows us the actual raw data in differing ways, such as:

http://www.digitaldog.net/files/ThisIsRaw.jpg

There are a number of ways of lightening or darkening areas of an image. These are the methods I use, others may disagree or have different approaches they are comfortable with. I don't use Lightroom, these are how I use Photoshop CS5, I think Affinity has the same tools, perhaps under different names.

 

1) Use the RAW processor to create a lighter version of the picture. Put the original and lighter versions on different layers. Use a layer mask or the erase tool to make the lighter version show through where required.

 

2) Select the area to be lightened and feather the edge. Adjust the selected area using shadow / highlight, levels or curves. It's probably best to use shadow / highlight as it's a relatively gentle adjustment, using levels or curves can cause obvious edges.

 

3) Make a new overlay layer filled with 50% grey, and paint over with white to lighten, black to darken. This is how I do darkroom style dodging and burning, using a large, fuzzy brush at low opacity. It may not be the best way to adjust sharply defined areas.

 

By the way are you sure you posted the correct image? I don't see the china cabinet referred to.

Edited April 24, 2022 by John Seaman

I don't use AffinityPhoto but it looks very similar to Photoshop which I do use. I rarely look at histograms other than when shown in an adjustment layer.

 

You should be able to adjust levels, brightness/contrast, exposure, shadows/highlights (and probably gradients) using just a few AffinityPhoto tools:

- add adjustment layer: levels, curves, exposure; initially, this will apply to the whole image

- add a mask to the adjustment layer

- modify the mask (black/white) to apply adjustments only to selected areas - see 'selections' below

 

So in your example, my preference would be to add 2 curves adjustment layers, each with a mask. In one layer, I would drag the curve downwards slightly and call it 'darken' and in the other, I would drag the curve upwards and call it 'brighten'. By default, masks are set to all-white so that they apply to the whole photo. If you select the mask and then select "layer", then "invert" then the mask turns black. A keyboard shortcut is CNTRL + "i". Using the paintbrush tool with the color white, you can then 'paint on" the mask to select the areas of a photo that you want the curve (darken/brighten) to apply. You can further adjust either the curve or (more usually) the opacity of the adjustment layer to the level you want the adjustment to apply.

 

Selections

Often, applying adjustments to selected areas by "painting in" white areas on a black mask works fine. But sometimes more precise selections are needed for areas (such as windows, etc.) that have edges. My go-to selection tool for this is the "polygonal lasso". Richt-click on the lasso tool and the polygonal lasso option is shown at the top. This option allows you select areas bounded by edges by clicking on points (for example corners).

 

I googled a couple of AffinityPhoto topics just to make sure that they were available. Generally, I found the videos at affinity.serif.com to be the best ones: short ca. 5 minutes, clear and to-the-point.

My go-to selection tool for this is the "polygonal lasso".

 

I find the "Quick Selection Tool" absolutely invaluable for this task. It's the small paintbush just below the polygonal lasso in the toolbar (CS5). As long as there are reasonably well defined boundaries, moving the tool's circle over the area to be selected almost magically (at least to me) selects the area up to the edges. If it overshoots, just use the same tool while holding down "Alt" to adjust the edge.

Edited April 24, 2022 by John Seaman

Let's go back to basics.

 

You wrote:

 

I don't want to know how to use a photo editor's tools (there are plenty of tutorials), but how to apply the tools to my needs.

 

That's a bit like saying :"I don't want to learn how to fly a plane, I just want to fly it from A to B." If you want to adjust tonality, you need to learn the tools that do that. The suggestions that John and Mike offered are just a few of the many ways you can use those tools to change tonality. However, there is no recipe. You need to diagnose what specific changes you want to make and then decide which of the many tools will work for you.

 

As Dog pointed out, we can't tell from the JPEG whether your raw file has detail in the shadows. If not, then there is no way to retrieve them, and you would have to retake the image using bracking and HDR or exposure blending. if there is detail in the shadows, there are several tools you could use. Personally, I'd probably use luminosity masks, which you can now emulate pretty easily in Lightroom. Alternatively, you could use selection tools, as in Photoshop or many other programs.

digitaldog:

I'm unable to attach the CSV file from RawDigger because it's not an image. Is there a screenshot I can grab for you? RawDigger has a sale that only runs for another few hours, so if the data I provided is useful to you or anyone else, please let me know so I can purchase it at the discounted price. I'll be happy to learn it in depth, but not if it only provides limited value.

 

mikemorrell/John/paddler4:

I have experimented with Affinity's quick selection tool (and earlier the flood selection tool in Paint_Dot_Net) in other images to isolate certain areas and play with all of the various settings, so I know how to use them and what they do. However, after placing selections on new layers and making adjustments, they don't seem to blend in well with the original image. They result in a photo that shouts "...and I made this part lighter".

 

I agree that Affinity's video tutorials are very well done; brief and clear. But they don't help me with my original post. Since I am an amateur trying to get up to speed on both the original image and post-processing, there are still things I don't know. In this thread's example, which tools should I use to fix certain problems. Going back to the airplane analogy, now that I can get the plane up in the air and land it again, and I can turn and change altitude, what do I do in case of bad weather; fly above it, fly around it, or increase speed to plow through and get out more quickly?

 

The JPEG i posted was to save space, but if there is a 1GB limit, I could upload the RAW file, which is 28MB. (I'm used to producing JPEG's that are 5-8 MB). You can't see the china cabinet on the right because it is underexposed, but I can see it if I bump exposure by 2 full stops. (That's not the effect I want for the overall picture, I'm just saying it's possible.)

 

mikemorrell/John:

I'll try each of your suggestions; those are what I'm looking for.

 

I appreciate everyone's patience; thanks.

 

Edited April 24, 2022 by alan_varga

RawDigger is a great tool, I'm not sure you must run out and buy it (of course if you do, it is great for learning exposure for your sensors).

What I'd suggest is you examine a much easier feedback loop first; the Display overlay for under and overexposure which will show you what is what on your own image, because you the image creator may decide some areas are fine to block up (especially in shadows). Look at my screen capture above, you'll see that option to click on one or both checkboxes and to the right, the over and under stats of the entire image. Between the stats and the overlay, you can get a quick look at the exposure results on that specific raw.

Look at my screen capture above, you'll see that option to click on one or both checkboxes and to the right, the over and under stats of the entire image.

 

For some reason I can't see your screen capture. I've had a little trouble getting my images uploaded.

 

the Display overlay for under and overexposure which will show you what is what on your own image, because you the image creator may decide some areas are fine to block up (especially in shadows).

 

Between the stats and the overlay, you can get a quick look at the exposure results on that specific raw.

 

OK, with RGB render I see that the china cabinet and the window frame between the two windows is "underexposed", and nothing is overexposed.

If I switch to Raw Channel there is somewhat less area flagged as underexposed.

 

Armed with that, how would I proceed (assuming I haven't gotten to trying the other suggestions yet)? Do I go after exposure, brightness/contrast or shadow/highlight? Or combinations?

Yes, at the very first step, you go 'after exposure'; optimal exposure for your raw data. Long before editing because GIGO:Garbage In Garbage Out.

So the info in RD shows you didn't over-expose any pixels and you did under-exposed some so that indicates you could have increased the exposure. The trick is providing enough exposure so you do not clip highlights you wish not to clip and this is where RD and testing come into play.

From the RD manual:

• RawDigger allows the determination of how the exposure meter is calibrated and what raw level corresponds to the midpoint of the in-camera histogram. That is, RawDigger lets you establish the headroom in highlights and obtain optimal exposures.
  
• RawDigger helps determine the raw level at which the overexposure "blinkies" start on the camera LCD, and it helps know how much headroom is still available after the blinkies start showing or the histogram hits the wall.
  
• If the shadows look blotchy or colorless, or details in the shadows are poorly resolved, you will be able to determine how much they are underexposed and set the Underexposure (UE) indicator in RawDigger accordingly.
  
• RawDigger is a useful tool for examining how the relative per-channel underexposure depends on the color of light.
  
• RawDigger is more precise than any exposure meter for the purpose of evaluation of the uniformity of fill light and reproduction light setups (in terms of the evenness of both color balance and luminosity across the background).
  

More here:

OneZone – The Optimum Digital Exposure

ETTR

The Optimum Digital Exposure - Luminous Landscape

http://digitaldog.net/files/ExposeForRaw.pdf

The Unbearable Lightness of Mystic "Exposure" Triangle

Red Flowers Photography: Now It is Easy to See the Real Picture

Exposure for RAW vs. Exposure for JPEG

Beware the Histogram

Establishing the in-camera exposure meter calibration point is the way to extract more dynamic range from your camera

Edited April 24, 2022 by digitaldog

which tools should I use to fix certain problems.

 

Often there are a variethy of options. For example, one of the most basic techniques to change tonality in parts of an image is dodging and burning. There are a variet of ways to do that. I use a method that I think maximizes control, but I'd be willing to bet that some people here who have very good postprocessing skills use a different approach.

 

Another example: in the Adobe world, there are at least three approaches for basic global tonality adjustments: the sliders in Lightroom, the curve tool in both Lightroom and Photoshop, and the levels tool in Photoshop. All of them do the same thing: they change the brightness of the image. However, they do it in different ways, and the expert editors I know often choose to use different ones. The key is learning how each of them affects luminosity and then choosing the one(s) that enable you to make the changes you want. I personally don't fine the LR sliders all that useful in many cases and prefer the control offered by levels and curves, but I know people who feel exactly the opposite. Different strokes.

 

However, after placing selections on new layers and making adjustments, they don't seem to blend in well with the original image. They result in a photo that shouts "...and I made this part lighter"

 

Not entirely suprising if you have selections with hard edges. You could feather your selections to make the borders fade a bit gradually, you could use a brush and vary the hardness, or you could use luminosity masks, which are usually self-feathering. All of these take time to learn, and which one is best depends on both taste and the particular image. In this case, you have an image with very hard edges, so it's going to be difficult to make adjustments that don't have fine edges. This may limit how powerful an adjustment you can make before things begin to look artificial.

 

I am not trying to discourage you, but it reallly does take time to learn how these tools work and to gain skill in using them. All of the amateurs I know who are good editors have spent countless hours learning how to do this. For example, I've been doing digital editing for many years, but it took me hours of work to gain a basic level of competence with luminosity masks. I still find that selections are often difficult. That's why I responded the way I did to your comment that you don't want to learn how to use the tools. If you don't learn that, you're going to be disappointed in the results.

Many image editors, especially ones that handle RAW, have a preserve shadows slider. Heck, even the image editor on my smartphone does!

That would be my first port of call, and failing that, using the curves tool to lift the shadow end of the tone curve. Failing that, I'd create a much lighter image using the RAW exposure slider and use that as a layer to blend with a highlight-exposed layer. There are many ways to skin that cat.

 

Then there would be controlling the lighting in the first place, to brighten the shadows before or while the shot was taken. A small amount of flash bounced from the ceiling or a wall behind the camera would probably have done the job in your posted example.

Edited April 25, 2022 by rodeo_joe|1

I had another look at your picture, and to make the china cabinet visible I had to brighten it so much that horrible noise was introduced. I think you are expecting too much of your camera and processing software, there's simply too much contrast between the light and dark areas. If you increase the exposure in the camera for the dark areas, the highlights in the window will burn out beyond recovery. As others have said, the answer is to introduce illumination via fill flash or otherwise. Real estate photographers don't carry around cumbersome lighting equipment for nothing. Edited April 25, 2022 by John Seaman

I had another look at your picture, and to make the china cabinet visible I had to brighten it so much that horrible noise was introduced.

 

When you brighten rather than expose ideally, you end up with more noise. That is why the first step is optimal exposure.

First off, it appears that the contrast of the scene exceeds the camera's dynamic range, so there's nothing in terms of exposure that can fix that if your goal is to have properly exposed shadows and highlights.

Simply not so! The image is under-exposed. That's the fact.

The goal of a photographer is to deal with the scene DR if and when it exceeds the DR or the capture device. Under exposing it doesn't do anything at all useful. In fact, optimal exposure will allow the full DR of the capture.

Clearly, you didn't read these facts provided:

Establishing the in-camera exposure meter calibration point is the way to extract more dynamic range from your camera

Brightening the shadows, in this case the interior, with additional light sources will get you closer to an optimal exposure.

The OP is asking about the captured image he has, not how to reshoot it. It wasn't optimally exposed.

Classic misunderstanding of what exposure is! It only takes place at capture. Exposure never happens by brightening the image, impossible. Exposure is the sole attrubute of how much light strikes the sensor (or film) based upon only two attributes: Shutter and Aperture.

Your comments (and a lack of photographic examples or outside peer-reviewed references) again illustrate you should first study such outside references before commenting about exposure (and confusing it with adjusting brightness). The URLs remain above, lots of them, from many experts in this field.

 

"Listen to understand instead of listening to respond." - Barack Obama

Read to understand instead of not reading and responding. ;)

 

I do expect you'll find the need to push back on the multiple peer-reviewed articles on optimal exposure and their effect on DR and noise, and what I've provided here, and you'll get us off-topic. I hope I can refrain from more of the typical CWOBaT (colossal waste of bandwidth and time) that usually follows for the sake of the OP and others.

Edited April 25, 2022 by digitaldog

When you brighten rather than expose ideally, you end up with more noise. That is why the first step is optimal exposure.

First off, it appears that the contrast of the scene exceeds the camera's dynamic range, so there's nothing in terms of exposure that can fix that if your goal is to have properly exposed shadows and highlights. There is no optimal exposure to make that happen.

Brightening the shadows, in this case the interior, with additional light sources will get you closer to an optimal exposure.

First off, it appears that the contrast of the scene exceeds the camera's dynamic range...

Oh, please specify the exact DR of the scene and the exact DR of the capture based on it's exposure.

I've spent the last week exploring of the interesting suggestions that all of you have offered: duplicating the base layer, then masking further adjustments:

• increasing exposure by two stops
  
• using shadow/hightlight
  
• using levels
  
• using curves (both on a single layer, or alternately one light layer and one dark layer)
  
• using brightness/contrast
  
• using a gradient
  
• using a spotlight (maybe specific to Affinity)
  
• using separate layers for dodging, burning and sponging
  

 

I couldn't find a setting to preserve shadows when developing from Raw persona to Photo persona, but that doesn't mean it isn't there.

 

One other option was saving an overexposed version as a JPEG, then trying an HDR merge with the original. The result was somewhere in-between.

A variant of this was deleting the dark areas on the light version and trying an HDR merge, but with that image having mostly white, that wasn't a great idea.

 

I tried working in the "Wide Gamut RGB Liner" color space, which I expected would give me a wider tonal range to work with (it's supposed to be 32-bit?). My "best" result was adding a Levels layer, setting the white level at 52 (0-100), and carefully masking the right edge of the drapes to set off the area with the china cabinet.

 

I must agree with everyone who said there is too much contrast to salvage this image. Another way to say it is that there isn't enough data in the lower light range to work with and get an acceptable result. Too much light and the shadows are lost, too little light and the highlights blow out. Overcompensating with extreme adjustments just makes for unacceptable noise.

 

This is a similar problem to audio file editing. If there isn't enough data, the low frequencies can't be boosted enough or the high frequencies sharpened enough that background noise doesn't ruin the result. Vinyl records are easier to restore than audio cassettes.

 

Thanks for your input everybody. Although I couldn't fix this image, I will be looking at how to calibrate the exposure meter on my camera, and reading all of the research material you provided. I got a couple of new ideas out of this thread and improved my masking skills; this week was quite a journey.

 

I don't understand: "preserve shadows when developing from Raw persona to Photo persona"

 

A wider gamut color space in a linear or gamma corrected state will have no bearing here. Color Gamut defines a range of colors. We can go into color gamut and gamma encoding, but it isn't part of this set of issues.

 

Yes, there isn't enough (tonal) data. Your sensor has a max DR; optimal exposure will provide the most DR it can provide. Lower exposure reduces DR and increases noise. Step one is optimal exposure. Adding more light if possible, while under-exposing is counterproductive.

 

If the DR or the scene exceeds the DR of the film or sensor, you decide what you can produce within that range you wish to express photographically in ONE capture or you control the lighting (more fill, etc) again ideally exposing OR you start looking into HDR (image blending of multiple captures, a deal-breaker with anything that moves).

 

I provided a number of URLs on exposure and a few DO discuss how to calibrate the exposure meter on your camera. But it is easily fooled, it isn't that 'smart' but you, the photographer can be. Any reflective meter of a white dog on snow, a black cat on coal will be fooled. You can/should compensate when you do 'calibrate' the meter. While understanding what it believes/assumes to be the gray to correct by optimally exposing (for raw or otherwise):

 

Edited April 30, 2022 by digitaldog

I will be looking at how to calibrate the exposure meter on my camera,...

That should be no problem. Just read up on the exposure technique called "Expose To The Right" - ETTR for short.

This entails using the camera histogram to place the brightest highlights at the extreme right of the histogram. Basically the opposite of the old film advice to "Expose for the shadows and let the highlights take care of themselves"

ETTR then, theoretically, makes the best use the camera's dynamic range.

 

Incidentally the lens has a big effect on capturable dynamic range, by affecting veiling flare, and hence the minimum brightness of the shadows. So 'calibrating' your camera meter may well only be valid for one lens, and one particular set of subject and lighting circumstances.

 

For example: The distribution of light and dark in a shot of a full moon may be 10% highlight area and 90% pitch blackness. But that 'pitch blackness' at the camera sensor will be contaminated by light diffused and reflected back from the brightly-lit 10% highlight area. Taking the highlights as 100% brightness, a camera and lens with very good flare control might only throw 0.1% of the highlights into shadow areas, and 0.1% of 10% = 0.01%, or 10,000:1 (=13.3 stops) Therefore that will be the 'dynamic range' of the camera+lens in those particular circumstances. But increase the highlight area and that same camera and lens might only return a 2,000:1 dynamic range (about 11 stops).

 

Things are never simple are they?

This entails using the camera histogram to place the brightest highlights at the extreme right of the histogram.

Beware the Histogram

Just get RawDigger to do this "calibration" instead of a graph that lies to the user.

Or, just view the actual Raw Histogram on your PhaseOne camera. :)