In all my years of printing and teaching printing workshops, the single most important thing…
Most of us at one point or another have heard or read the maxim in digital photography commonly titled “expose to the right”. But do you really do this on a regular basis, and do you know how? More importantly do you know why? I’ll try to answer these questions as simply as I can and also provide some tips to put this practice into use every time you go out and shoot. Afterall, I don’t know of a single photographer that doesn’t want to come home with the highest quality images possible.
Many of the students I work with seem to be unclear about exposing to the right, and I think part of this can be attributed to 2 main reasons:
- Not having a full understanding of the reading and use of the histogram
- Depending on the LCD preview on the back of the camera as a way to aesthetically judge proper exposure, color, and contrast.
First, Let’s Talk About Camera Sensors
All digital camera sensors capture light in a linear fashion, starting from black to white (left to right on the histogram). This means they capture fewer levels of information in the shadows, and the maximum number of levels in the highlights, just before clipping. In other words, the sensor is much more sensitive to brighter levels of light that darker ones.
This is the main reason why we see digital noise in the shadows and not the highlights. The “signal to noise” ratio is much higher in the highlights, therefore noise can only rear its ugly head in the shadows where this ratio is much lower.
Because sensors capture more light levels in the highlights, or towards the right, there is more information that can be utilized by the post-processing software when we make our adjustments to the RAW file. This gives us more data to work with, preventing potential problems such as banding, loss of detail, lack of tonal separation in the shadows, and noisy images. We gain much better shadow depth, clarity, and improve our overall image quality, especially when printing, or making aggressive adjustments in post-processing.
How to Properly Read the Histogram
The histogram is your guide. If your camera supports the separate RGB channels of the histogram, use it, otherwise the simple white histogram is fine as well. Your goal is to make an exposure where the majority of the information is as close to the right edge of the histogram WITHOUT a spike on the extreme right hand side. A spike means the sensor has received more light than it can record, and rendered that part of the image pure white without any detail.
Be aware that in some cases this may be ok depending on the image and your creative goals – perhaps you want to purposely blow out part of the image, or a bright part of the scene has very little detail to begin with ie. an overcast sky.
The important thing to remember is that the size and /or shape of the histogram is irrelevant, all that matters is whether there are spikes on the extreme left or right edges of the display. Either means that detail has been lost. If you have spikes on both ends, that means detail has been lost in the highlights and shadows, and more importantly the scene’s dynamic range is greater than what your camera can capture. For the purposes of this article, I’m assuming that is not the case, and you can move the histogram to the right or left.
Exposing to the Right
So you’ve made an exposure, and the histogram shows there is a space between the levels you captured and maximum exposure or 100% (the extreme right). This means you’ve left potential image quality on the table. To expose to the right, you’d simply apply plus exposure compensation if you’re using one of the semi-automatic modes, or adjust your shutter speed or aperture accordingly if using manual mode. You want to let in more light in order to move the histogram further to the right just before it clips in the highlights. For example, in landscape photography, you typically set your aperture for maximum depth of field, so you would lengthen your shutter speed to let in more light.
However, here’s where lots of people get hung up. When you expose to the right properly, the preview on the back of your camera will look awful and very overexposed. You might be tempted to compensate by adjusting your exposure setting until it looks much better, but now you are basically throwing away image quality and not getting the most out of your expensive camera body. Use your LCD preview for composition and focus, not for exposure. Trust the histogram, and the quality of your images will increase noticeably.
Post-Processing
The final part of this process is properly processing your RAW files. I use Lightroom 4, so I’ll focus on my workflow using this app, but the same concepts can be used in other raw processors. When you import your images, assuming you don’t apply any automatic settings, those that have been “exposed to the right” will again look overexposed. But by adjusting the exposure slider you can adjust them to your liking and get them looking great again. The difference is that now you are starting out with the cleanest data, and redistributing it back into the sweet spot of what we like to see visually. Shadow detail is much cleaner and easier to recover, and all subsequent adjustments to the image will deliver better quality and results.
*Real World Example*
Graphic 1: Image #1 on top is exposed to look good in the preview, image #2 on the bottom is exposed to the right. Notice the difference in the previews and histograms.
Graphic 2: Image #2 has been adjusted to look similar to image #1 using the exposure slider in Lightroom 4 which now sets overall image brightness. I’ve brought it down a little over a full stop.
Graphic 3: To open the shadows more and balance the exposures, I’ve added +70 to the Shadows slider to both images.They both look identical right?
Graphic 4: Above is image #1 at 2:1 zoom ratio in a dark part of the image with lots of detail. [Click on the image for a larger version.]
Graphic #5: Above is image #2 in the same spot, and you can clearly see how much cleaner and less noisy it is compared to image #1. In a 16 x 20 print, this difference would be easy to see. [Click on the image for a larger version.]
I hope this has been helpful, and clears up some confusion about this important and useful technique for maximizing your image quality. Please leave any comments, feedback, or questions in the comments below. Thanks for reading!
Robert, this has been useful! I know the technique and use it on occasion but I am stubborn when it just seems right, straight out the camera. I mean, who really wants to go back and adjust their photos from a shoot -1 stop or more for each of them but this provides a very clear example. Really simply put and explained very well.
By the way, I now have my drivers license so I’m looking forward to getting up the Hudson more. Hope to finally meet someday soon.
Thanks for the feedback Ed – glad it was helpful. Lightroom makes it very easy to batch apply changes to a group of images using the “Sync” command, so making these types of changes is no big deal. Would be great to meet you sometime as well!
RR
Thanks for sharing this info, Robert. You’re right, it’s clearly possible to get more shadow detail using this method, but I see two main issues with the “expose to the right” philosophy:
1. the histogram that your camera shows you for a RAW file can actually not really be trusted (unless it’s a RAW from a Leica M Monochrome), because it’s always based on the embedded JPG. This in turn has been created based on your picture style settings (e.g. with a much wider histogram due to the picture style’s contrast settings)
2. the more our digital cameras try to mimic the dynamic behavior of film, including some highlight compression, the more non-linear the brightest stops in the histogram become. Please correct me if I’m wrong, but I’m not aware of any software that is capable of converting that non-linear part back into the linear part closer to the middle of the curve.
Based on that, my preference is usually to take the picture as close to the final exposure as I can. This way I am faster in post and I don’t introduce unnecessary non-linearities.
Hi Chris, thanks for the feedback and for your perspective. You are correct I should have mentioned that the histogram is based on the jpeg preview, which is why I always advocate setting your picture style to “neutral” or similar flat setting – all controls at zero. This way you are getting a purer histogram that has not been affected by extra contrast, sat, etc.
As for your second point, yes there are options such as “highlight tone priority” that do try to mimic film roll-off so that also would be something to consider. I usually tend to leave these options off and make all adjustments in post.
A few points to consider:
1) The exposure meters built-in to modern cameras tend to be quite conservative, since manufacturers want the exposure to look “good” (like film), and often may be underexposing by a stop or more, especially when photographing scenes with grey or dark tones. The blinkies tend to come on even when I know I have not overexposed as verified later in Lightroom/Photoshop ACR.
2) Imagine photographing a black bear sitting on coal. The “proper” exposure based on the camera would be very underexposed even if it looked right on the LCD. Pushing it to the right would yield a much cleaner image, less noise, and better shadow depth – even if it looked very bright at first.
3) Modern DSLR’s have as much as 10 stops of dynamic range these days, and medium format digital cameras as much as 12-13, so again maximizing what the sensor is capturing is key to getting the most out of these cameras – no reason to leave IQ on the floor.
I think the focus of my recommendation was directed at fine art/landscape photographers like myself who are trying to make that very special image from all of our time spent in the field, and so extrating every bit of info from the sensor is both worthwhile and important. For sure I would not recommend this practice to a photojournalist or sports photographer for the reasons you mentioned. I also do not think every shot needs it – but be aware of when it can be beneficial.
Finally, practice and experience will be your guide, and so I have a very good idea about how my histogram will translate to the actual image on my Canon 1DS Mk III – based on trial and error 🙂
I’m sure this is a topic that will continue to evolve with future generations of cameras, and we’ll have to keep adapting as well. Again, thanks for the insightful points you raised – much appreciated.
RR
Chris more eloquently articulated part of what I was trying to say. I also agree there are different times and circumstances for different approaches. That’s one of the beauties of the Canon cameras custom function buttons on the dial C1-3. You can program each of those with particular settings reflecting the choices you were both talking about above and apply them as needed.
Excellent blog here! Also your website quite a bit up
fast! What host are you the usage of? Can I get your associate link in your host?
I wish my website loaded up as quickly as yours lol
This is also possible if they offer SEO as part of their service.
One should keep in mind that communication lines for live support are kept open for paid accounts, who
are given top priority. The amount of space required by a website should also be considered while choosing a
web host.
Thanks for the interesting post. A couple of things through me a bit though: you state that “When you expose to the right properly, the preview on the back or your camera will look awful and very overexposed.” Apart from the amusing little typo (sure wouldn’t want my camera looking awful :-), I find this a bit surprising. I shoot MFT and appreciate the advantage of the WSIWYG viewfinder, so perhaps this is a DSLR thing. Or perhaps you just forgot to reset the brightness of the preview after working in full sun.
The other–related–thing was a ‘miss-speak’ I think. In the caption for Graphic 1 you say “Image #1 on top is exposed to look good on the histogram…” I’m guessing you mean on the preview. But that aside, the histogram does not look good to me. It is clearly crowding the left side, while leaving unused room on the right–a classic case of under exposure!
I think your choice of scene to illustrate the value of exposing to the right would have been better served by an example having less-than-normal dynamic range. The bear on the coal pile or the snowman on the ski slope would have made the point a little more convincingly perhaps. In all these situations, though, the important thing is to avoid under- or over-exposure. And so I think the best advice would simply be to bracket your exposure and decide which one is best later! That buys you some dirt cheap insurance and the possibility of HDR or exposure blending if the DR proves to have been too much for the sensor.
Still, a good article, and interesting website!
Hi Virgil – thanks for the feedback. I think you may be confusing some of my points. When I say “preview” I mean the actual image (jpeg preview) you see on the camera’s LCD after you’ve taken the image and want to review it to evaluate exposure, composition, etc. So yes if you expose to the right, the jpeg preview will often look overexposed, regardless of whether you’re using a DSLR or mirror-less camera. I have both and this happens on either system. This is why the histogram is so important, because it’s the only factual way to tell if the image is overexposed or if anything is clipped. – loss of detail.
And yes I mean what I said about the first image. IF I were to use the preview that the camera shows me in it’s LCD after making that image, I would assume it was exposed properly because it “looks” more natural and like the actual scene. But the histogram tells a different story – it is, like you state, underexposed. The second image is properly exposed, yet does not “look” as good on the LCD. I could have used an image with lower dynamic range, but I wanted to show how even a small margin can make a different in the shadows with respects to noise. That also has major ramifications for prints as well.
The basic point behind all of this is that I’ve had many, many students who not only do not use the histogram, but don’t know how to use it to evaluate exposure. And for those that do, they are not aware of how a sensor captures information and the discrepancy between the right and left side. They simply use the LCD preview after taking an image for aesthetic and technical evaluation. Hope it’s a little clearer now…
Thanks for the clarification Robert. I don’t disagree with the points you’re making, but when you say in the Graphic 1 caption “Image #1 on top is exposed to look good on the histogram…” you surely meant “…look good in the PREVIEW…”. That completely reverses the meaning. The histogram looks bad, but the preview probably looked okay, as I think you meant to say.
You also seem to have missed the other typo I was trying to point out: “…the preview on the back or your camera will look awful…” The “or” should be “of”, (but then the humour will be lost 🙂
Just trying to clean up a couple little typing errors that interfere with the message.
Yes, thank you Virgil – I completely understand now – errors corrected. Appreciate your patience 🙂
Hi Robert,
Do you think that ETTR still applies to the current crop of cameras on the market today? Also if one is making a portrait or shooting a model would ETTR be a good practice?
BTW I learned about you through your BH Event presentations. Great stuff!??.
Thanks,
Jerome
Hi Jerome, I still think ETTR is a good practice in my experience especially if you print your work. The whole point is to get the most amount of data into the RAW file, and that makes no difference with subject matter. Cleaner files make better images and prints.
Thanks for the info!!!