A FRIEND recently sent me a link to a You-Tube video by a popular wildlife photographer/presenter. At first glance, the viedeo appears to question whether the popular concept of ETTR was still valid in today's digital world. In the context of wildlife (and probably most other forms of action photography), that might just be right. Or . . . not. I am going to play "devil's advocate" here, and suggest that ETTR is alive and well. In all types of photography. But - like all the tools and techniques we use - it is just that. A tool. and knowing what it does (and what it doesn't) do is going to make all the difference in its effectiveness. What I do think the video underscores is the over-reliance on any one technique, to the detriment of the overall photographic result. ETTR is shorthand for "Expose To The Right," and refers to the actual tool we are talking about here: the histogram.
I am going to play "devil's advocate" here, and suggest that ETTR is alive and well . . . In all types of photography
FIFTEEN YEARS ago, I posted on this topic in "Expose Right To Expose Correctly." At that time, although I was familiar with the histogram tool, the ETTR concept was somewhat new to me. I had been using my old tried and true exposure techniques and relying heavily on the very good Nikon metering system. I could see the exposures immediately after recording them on the LCD screen on the camera back. I could also see the histogram of my just-shot image, which was a proverbial "game changer." With that tool, you can see if you have blown out your highlights or blocked up the shadows. Well, sort of (the histogram on the camera LCD is only a jpg rendition of raw images - which I have always shot, and therefore not entirely accurate). I also knew that I had a bit of latitude to adjust exposures in post-processing. Consequently, I was still in the habit of bracketing exposures when I wasn't sure. Learning the ETTR method freed me from that (maybe only making room for other bad habits 😀).
IN THE post referenced above, I explained in my simplistic layperson's terms, what the ETTR concept was all about:
"Now, here is why to expose to the right. Consider the graphical 5-stop diagram below. Note that as we progress from 128 to 2048, the area under consideration continues to double. What this means in very simple terms is that the last step contains 50% of all the pixels captured! Note that that is also the highlight. So we want to capture as much of that as possible. We accomplish that by shifting our histogram as far right as we can without blowing any highlights. Another way to say this is that if you do not fill the right side of the histogram you are effectively potentially wasting up to 50% of the available information that your camera is capable of capturing."
2009 SEEMS like eons ago when it comes to digital technology. I was shooting with a 10 megapixel, "APS-C" sensor camera. Today my current primary camera is a 60 megapixel, "full frame," "dual-gain" sensor camera (a technological mouthful, for sure). The noise handling capability of newer sensors is much advanced over the older sensors. Sharpness and overall image quality is many times better. We are using much higher ISO settings today than we felt comfortable with back then.
THE YOU-TUBE video throws a curveball at many of us, who have always assumed that higher ISO = higher noise. But, the video teaches us that ISO really isn't the critical factor for noise. Rather, noise is a characteristic of individual camera sensor designs (smaller sensors still produce more noise), and on a given sensor, is really effected only by aperture and shutter speed (the only two real adjustments to the amount of light the sensor is exposed to). ISO is a measure of light sensitivity - not amount (and amount is what effects more or less noise).
ISO isn't really the critical factor for noise
I AM with the author to this point. I am not sure, though, that I agree on the conclusions he seems to draw about ETTR. I think his assertion confuses two different ideas: namely ETTR and ISO. The ETTR construct is really more directed at the fundamental suggested above: that noise (or not) is about the amount of light that reaches the sensor for a given exposure. In the ETTR explanation above (orginally brought to light by Canadian photographer, Michael Reichman, perhaps best known for his "Luminous Landscape" internet site), what we are talking about is really the amount of light, and there is more in the top 2048 pixels than in the bottom 128 pixels. Noise is traditionally presented in the lower (blacks) level of the image. None of this explanation has really touched on ISO.
ONE OF the things that causes confusion is the ability to (and in many cases advisability of) setting modern cameras to "auto-ISO" (letting the camera choose what it thinks is the appropriate ISO). With the much higher quality performance of modern camera sensors, it makes a lot of sense to do this in certain conditions. With any kind of "action" shooting, things change so quickly that letting the camera decide ISO gives us more flexibility as shooters. It is one less parameter we have to worry about. Instead, we can adjust aperture and shutter speed to meet the conditions, knowing the auto-ISO is going to adjust accordingly.
ETTR has never espoused clipping the highlights
FOR LANDSCAPE shooting involving "still" subjects, there normally isn't as much benefit to high ISO settings (though there could be some in a few cases). Here, we are wanting to create the best quality images we can. This usually means shooting at the "native" ISO of the camera (usually around 100 for full frame, between 100 and 200 APS-C and 200 M4/3 cameras). It also means shooting from a stable platform (generally a tripod), allowing us to make slow shutter speed exposures in many cases (of course, wind or other movement will limit that). Generally the only reason I would change the ISO on a tripod-mounted landscape image would be for wind movement or if the light is too bright (though there are better methods for addressing the latter). However, there is certainly continued benefit to using ETTR for landscape images. Detail is often where it is at in a good landscape image and ETTR is going to enhance our ability to show that detail.
THE VIDEO suggests that ETTR can actually degrade an image at higher ISO by causing parts of the image to "clip" at the lightest pixels. But that is where I think maybe we are mixing concepts. It is not the ETTR technique that causes clipping. It is more likely a misunderstanding of the purpose of ETTR by the person using it. The histogram is a tool to indicate to us when we have clipped either highlights or blacks in an image (kind of: remember that it is based on a jpg and if you are shooting raw files, you probably still have a slight amount of headroom and you will still have to use some judgment there). ETTR has never espoused clipping the highlights. It only suggests that, when appropriate, moving the histogram to the right as far as possible without clipping the highlights will result in retaining the most digital information in a file. I think this applies at any ISO setting.
It is not the ETTR technique that causes clipping
THE REAL point of the video, in my view, centers on the fact that the histogram, auto-ISO, and ETTR are all parts of the photographers "toolbag." It is still requisite that the photographer understand what the tools are doing for them. An over-reliance on any one tool or technique is quite likely to result in lowered image quality from time to time. The video warns of clipping in the highlight areas of the photograph and implies that ETTR creates a danger of doing so. But it is really not the ETTR that creates that danger. I do think over-reliance on it might do so. But that is not really on ETTR. It is on the shooter not understanding exposure and the histogram, and again - I think that is true at any ISO. A lot of wildlife images have high contrast and an abundance of whites (and often blacks). Those are areas where exposure is touchy. Again, the concept of ETTR is not to indiscriminately jam everything to the right. It is more to gain an understanding of where the most digital date is in an image file. Some images don't have much information in the lightest areas of the histogram. Others, while they do, aren't necessarily pure white (e.g., clouds and bird feathers). The far right side of the histogram represents pure white. Knowing this and knowing that your whites should not be pure white in an image will inform your use of the ETTR technique. At the same time, I do believe the presenter makes a very good point - perhaps the point - in his video. When shooting a moving target like wildlife, often in difficult lighting conditions, and more often than we would like, handheld, any reliance on the ETTR technique in the middle of the action might simply be impractical.
In my view . . . many photographers are obsessed with the issue of "noise"
IT SEEMS to me, though, that the real "worry" here is about noise. In my view too much is made of this and many photographers are obsessed with the issue of "noise." I blogged about this a year back in "What's All The Noise about Noise?" I am repeatedly taken by how often an on-line discussion about a camera model turns to the issue of noise. For years and years we used cellulose films. The "faster" their ISO rating, the more "grain" was present in the photo. While not the same phenomena, "noise" in a digital image creates a very similar effect in images. While there was certainly discussion about grain (or lack thereof) back then, it wasn't the all-consuming issue that today's ability to "pixel peep" on your monitor hath wrought. I have many images the display noise in the shadows. In most cases it is not that pronounced and is not, in my view, unpleasant. With that, today's sensors have significantly decreased digital noise. And, our post-processing software can nearly eliminate it. So if noise really bothers you, there are solutions out there. Frankly, it doesn't usually bother me that much (and when it does, I apply some noise reduction in post-processing).
