Is it better to over or to under-expose

[Kat501]

Decades ago, I shot Black & White film. Recently, I have started shooting color with a digital camera that looks like my old 35mm camera. In the old days, if I was uncertain of the best exposure, I would bracket either the time or the f-shop. As a roll of B & W film was limited in the number of shots it would take, I would just pick 2 shots rather than 3. I think it was better to shoot once at the exposure the camera's meter recommended and the other, one stop above, to overexpose. I had a friend who was a professional and I think that he recommended a small overexposure rather than a small underexposure. This was some time ago.

Maybe I should just rephrase this. If one is slightly uncertain of the best approximate exposure should one, over or under, expose.

...ok so... when I saw the title of this thread I was so excited. I told myself....ahhhh....I will finally get the answer to one of my most stringent dilemmas. However, the diversity of responses both “overexposed and underexposed” the “mental picture” I had about this topic (a bit of word play ). Recap for me:

1) Apparently there is not a solid consensus regarding the acceptable definition for “under/over exposure”.

2) A few postings with good links to technical data support mostly the “overexposing approach”. Not much of the kind for the other option.

3) I am puzzled by this: so many experienced photographers (you guys) no doubt got good quality pictures using either of the methods. How is that so? Is there a right and a maybe not so right way or the “waters of truth” are kind of murky somewhere in the middle?

 

[Michael Fryd]

Decades ago, I shot Black & White film. Recently, I have started shooting color with a digital camera that looks like my old 35mm camera. In the old days, if I was uncertain of the best exposure, I would bracket either the time or the f-shop. As a roll of B & W film was limited in the number of shots it would take, I would just pick 2 shots rather than 3. I think it was better to shoot once at the exposure the camera's meter recommended and the other, one stop above, to overexpose. I had a friend who was a professional and I think that he recommended a small overexposure rather than a small underexposure. This was some time ago.

Maybe I should just rephrase this. If one is slightly uncertain of the best approximate exposure should one, over or under, expose.

...ok so... when I saw the title of this thread I was so excited. I told myself....ahhhh....I will finally get the answer to one of my most stringent dilemmas. However, the diversity of responses both “overexposed and underexposed” the “mental picture” I had about this topic (a bit of word play ). Recap for me:

1) Apparently there is not a solid consensus regarding the acceptable definition for “under/over exposure”.

2) A few postings with good links to technical data support mostly the “overexposing approach”. Not much of the kind for the other option.

3) I am puzzled by this: so many experienced photographers (you guys) no doubt got good quality pictures using either of the methods. How is that so? Is there a right and a maybe not so right way or the “waters of truth” are kind of murky somewhere in the middle?

If you read carefully, it's not "over exposing" they suggest. It's "increasing exposure".

As a general rule, the more light you capture, the higher the quality of the image (up to the point of saturating the sensor). If you increase the captured light without lowering the ISO, then you get a result generally called "over exposure". If you increase the exposure while lowering the ISO, you get the benefits of the increased exposure, without the "over exposure".

Now, if you are already shooting at base ISO, then you do need to worry whether increased exposure will blow out your highlights. However, if you are in this situation, you likely have enough light that you can lower exposure a little bit without worrying about getting unacceptable noise.

Most photographers are not used to thinking this way, as common shooting wisdom is mostly from the days of film. In the days of film variable ISO, and Auto-ISO were not realistic options.

 

[tony field]

"Exposure" is the term we use to refer to the amount of light hitting the sensor/film. Usually, this is measured in terms of light per unit area. Exposure does not refer to whether the image is too light or too dark. An over-exposed negative can be printed as a dark, light, or normal print.

Continuing on the subject of confusing terminology for beginners, I think that a beginner would find the distinction between exposure and brightness, particularly on a jpeg , to be more confusing than the meaning of crop Factor as probably Illustrated in this thread.

--
Charles Darwin: "ignorance more frequently begets confidence than does knowledge."
tony
http://www.tphoto.ca

 

[Bill Ferris]

1) Apparently there is not a solid consensus regarding the acceptable definition for “under/over exposure”.

Consensus is overrated. Whether or not everybody agrees 2+2=4 is irrelevant to the fact 2+2 does equal 4.

Also, I think a closer reading of most of the responses will reveal folks are using their own voices to convey the same or similar information and recommendations. There's a lot more agreement in this thread than disagreement. That's my perspective, anyway.

2) A few postings with good links to technical data support mostly the “overexposing approach”. Not much of the kind for the other option.

Here's a link to a simple but illustrative video by Bill Lawson demonstrating the difference in the ability to recover detail from overexposed and underexposed negative film and digital camera images:

It clearly illustrates negative film's advantage having more latitude to recover detail from an overexposed image and digital's advantage in latitude to recover detail from an underexposed image.

3) I am puzzled by this: so many experienced photographers (you guys) no doubt got good quality pictures using either of the methods. How is that so? Is there a right and a maybe not so right way or the “waters of truth” are kind of murky somewhere in the middle?

What makes an image, good or bad, is determined by a photographer's subjective standards. We all take the same information (e.g. one has more latitude to recover details from an underexposed or overly dark digital image than from an overexposed or overly light digital image) and apply that to a variety of real world shooting situations. How we choose to apply that knowledge varies according to our respective, subjective preferences.

Each photographer not only brings a different aesthetic sensibility, but also a different set of skills and experiences. The photographer who swears by ETTR may have a finely developed instinctive ability to choose settings in almost any lighting situation that capture a scene to his or her liking, while also optimizing exposure without ever blowing out highlights. The photographer who doesn't apply ETTR in every situation plays to his or her aesthetic, experience and skill set. Both photographers can come away from a scene making fantastic - but very different - photographs.

There will likely never be consensus about what is the best photograph that can be made of a given scene. Nor should their be. Photography has room for a wide range of aesthetic sensibilities. So, even if we share the same, accurate understanding of what exposure is and what role ISO plays; and have the same or similar sensibility towards how many stops underexposed or overexposed an image can be and still retain enough information to be usable, there is still a good chance we would choose to make very different photographs of the exact same scene. And both photos could be equally fantastic...or equally awful

--
Bill Ferris Photography
Flagstaff, AZ
http://www.billferris.photoshelter.com

 

[Kat501]

Decades ago, I shot Black & White film. Recently, I have started shooting color with a digital camera that looks like my old 35mm camera. In the old days, if I was uncertain of the best exposure, I would bracket either the time or the f-shop. As a roll of B & W film was limited in the number of shots it would take, I would just pick 2 shots rather than 3. I think it was better to shoot once at the exposure the camera's meter recommended and the other, one stop above, to overexpose. I had a friend who was a professional and I think that he recommended a small overexposure rather than a small underexposure. This was some time ago.

Maybe I should just rephrase this. If one is slightly uncertain of the best approximate exposure should one, over or under, expose.

...ok so... when I saw the title of this thread I was so excited. I told myself....ahhhh....I will finally get the answer to one of my most stringent dilemmas. However, the diversity of responses both “overexposed and underexposed” the “mental picture” I had about this topic (a bit of word play ). Recap for me:

1) Apparently there is not a solid consensus regarding the acceptable definition for “under/over exposure”.

2) A few postings with good links to technical data support mostly the “overexposing approach”. Not much of the kind for the other option.

3) I am puzzled by this: so many experienced photographers (you guys) no doubt got good quality pictures using either of the methods. How is that so? Is there a right and a maybe not so right way or the “waters of truth” are kind of murky somewhere in the middle?

Thank you Bill and Michael for your detailed answers. Watching the YouTube video and reading (a few times...) your both narrative it makes more sense. Also I can see how the answers posted prior can be viewing same issue from two different perspectives and yet be both precise/factual.

On a side note, I tried yesterday the ETTR method presented in one of the videos linked above and I think I liked the end result.

 

[Michael Fryd]

"Exposure" is the term we use to refer to the amount of light hitting the sensor/film. Usually, this is measured in terms of light per unit area. Exposure does not refer to whether the image is too light or too dark. An over-exposed negative can be printed as a dark, light, or normal print.

Continuing on the subject of confusing terminology for beginners, I think that a beginner would find the distinction between exposure and brightness, particularly on a jpeg , to be more confusing than the meaning of crop Factor as probably Illustrated in this thread.

I have no problem with avoiding the word "exposure".

Beginners tend to already understand the concept of whether a print looks too light or too dark.

In order to teach photography, I talk about the light being captured by the sensor, and how the ISO setting ties the light hitting the sensor to how dark or light an image looks.

After they understand the concept, I can teach them that the word "exposure" is the term for light reaching the sensor per unit area.

The word "exposure" is only confusing if someone else has previously taught them the incorrect meaning (i.e. how dark/light the image looks). By avoiding the word "exposure" until later, we avoid any incorrect baggage associated with the term.

If the student is old enough to have some familiarity with a film camera, it's a bit easier. With film, the exposure determines how thin or dense the negative is. Whether the print is dark or light is determined by how we print that negative. With film it's obvious that "Exposure" and "Image Lightness" are different things.

 

[Michael Fryd]

...

Thank you Bill and Michael for your detailed answers. Watching the YouTube video and reading (a few times...) your both narrative it makes more sense. Also I can see how the answers posted prior can be viewing same issue from two different perspectives and yet be both precise/factual.

On a side note, I tried yesterday the ETTR method presented in one of the videos linked above and I think I liked the end result.

There are two strategies that fall under the name ETTR.

The first is shooting at base ISO, and using the highest exposure that doesn't blow out important highlights. That strategy usually maximizes image quality, and pretty much requires that you shoot in RAW and adjust brightness later.

The argument against this is that if you have enough light to be shooting at base ISO, noise isn't usually a problem. If the noise level is already so low that it isn't visible, there may be little advantage to further reductions. Obviously, there are some situations where ETTR will make a visible difference at base ISO, but there are certainly situations where it won't.

The other "ETTR" is when you are above base ISO. In that case ETTR is simply a matter of increasing the ISO to the highest value that doesn't result in numerical clipping in important highlight data. Raising the ISO can be helpful, but raising the exposure tends to be more helpful. Simply raising the ISO to shift the Histogram over doesn't have a lot of return (but it might help a little as the camera can internally optimize for the lower exposure).

The problem with most of the videos is that they are assuming you are shooting in "Exposure Priority" mode. Most people shoot this way without realizing it. With film that was the only choice.

"Exposure Priority" mode means that you start by picking a target Exposure. With film the target exposure is tied to the ISO value of the film. With digital, you pick a target Exposure by setting a specific ISO.

However, with Digital, there is no reason to pick the Exposure as the first step. You can select aperture and shutter speed, and allow the ISO to float (Auto-ISO) to match the Exposure you happen to get.

If you shoot that way, you can actually do a little bit better than ETTR in low light situations.

Select the largest aperture that yields enough Depth of Field. Select the slowest shutter that doesn't yield unwanted motion blur. That guarantees that you have the highest exposure you can reasonably achieve. If you set a fixed ISO, are in shutter or aperture priority, you might match the above. However if you don't match the above you are either getting more noise that necessary, not enough depth of field, or too much motion blur.

But the bottom line is that there is more than one correct way of looking at this. However, each of the ways tend to steer us towards some choices over others. It helps to understand the various mental models, so you can choose the one that seems to be a good match for your situation at hand.

 

[bobn2]

"Exposure" is the term we use to refer to the amount of light hitting the sensor/film. Usually, this is measured in terms of light per unit area. Exposure does not refer to whether the image is too light or too dark. An over-exposed negative can be printed as a dark, light, or normal print.

Continuing on the subject of confusing terminology for beginners, I think that a beginner would find the distinction between exposure and brightness, particularly on a jpeg , to be more confusing than the meaning of crop Factor as probably Illustrated in this thread.

I think you're completely wrong about that. The people who find the distinction difficult are people who have already learned wrong knowledge - that the two things are the same. The problem is that once you've adopted a bad mental model, it's really hard to rid yourself of it, and even though you can intellectually rid yourself of it it keeps on clouding your thinking.

A real beginner has no such problem. They appreciate that the image they see might be light or dark, and if they need a word to describe that, 'exposure' is not the most natural or useful. It's a jargon word they'll be unfamiliar with, and a misused one at that. 'Brightness' is more natural, but if you want a better jargon word, then the right one is 'lightness', which describes where a tone looks (perceptually) on the range of 'white' to 'black' rather than how much light it emits. Having raised the concept of how light or dark the final image is, it's then easy to talk about the factors which can determine how light or dark it is, which are at the simplest level 'exposure' and 'ISO'.

It's really very simple, as long as the beginner hasn't got confused by thinking that 'exposure' happens both at the input and output.

 

[skyglider]

Digital camera? = ETTR

+1

 

[gollywop]

...

Thank you Bill and Michael for your detailed answers. Watching the YouTube video and reading (a few times...) your both narrative it makes more sense. Also I can see how the answers posted prior can be viewing same issue from two different perspectives and yet be both precise/factual.

On a side note, I tried yesterday the ETTR method presented in one of the videos linked above and I think I liked the end result.

The other "ETTR" is when you are above base ISO.

Good that you put ETTR in quotes, because, by definition, ETTR is necessarily accomplished at base ISO. If you're above base ISO, you cannot ETTR.

 

[Paul P K]

Decades ago, I shot Black & White film. Recently, I have started shooting color with a digital camera that looks like my old 35mm camera. In the old days, if I was uncertain of the best exposure, I would bracket either the time or the f-shop. As a roll of B & W film was limited in the number of shots it would take, I would just pick 2 shots rather than 3. I think it was better to shoot once at the exposure the camera's meter recommended and the other, one stop above, to overexpose. I had a friend who was a professional and I think that he recommended a small overexposure rather than a small underexposure. This was some time ago.

Maybe I should just rephrase this. If one is slightly uncertain of the best approximate exposure should one, over or under, expose.

Back in the film shooting days the rule of thumb to, in case one had to be sure to have a slide or negative which afterwards, even with some tweaking, have a decent quality print, was to over expose positive film, and underexpose slides.

Overexposed film would still allow to 'burn in' the overexposed areas while printing to a 'detailed' white. Under exposed film instead would result in muddy colors when trying to print the highlight bright, not even overexposed, when printed

While underexposed slides would allow to 'clear up' the under exposed area's by giving those area's more light when printing. While with over exposed slides burnt out highlights would be impossible to in any way possible get any detail in afterwards in print

My experience is that Nikon sensors, even with the older models like the D1 (again in my experience), behave much like slide film, and allow quite a bit of under exposure

And afterwards, with basically very little effort or software can deliver quite good quality images, e.g.


Nikon Z6 Left : After postprocessing with (only) Nikon NX D Right : SOOC

Since Nikon and Sony sensors basically share the same manufacturer, I expect Sony sensors to do also

Not a Canon shooter, but from what I saw from colleague shooters ( I used to regularly shoot catwalk professionally, which meant I would often run into the same faces) using Canon, those sensors can take over exposing much better (although they suffer much more from under exposure)

--
all in a day's work
http://www.pbase.com/paul_k/

 

[tony field]

"Exposure" is the term we use to refer to the amount of light hitting the sensor/film. Usually, this is measured in terms of light per unit area. Exposure does not refer to whether the image is too light or too dark. An over-exposed negative can be printed as a dark, light, or normal print.

Continuing on the subject of confusing terminology for beginners, I think that a beginner would find the distinction between exposure and brightness, particularly on a jpeg , to be more confusing than the meaning of crop Factor as probably Illustrated in this thread.

I think you're completely wrong about that. The people who find the distinction difficult are people who have already learned wrong knowledge - that the two things are the same. The problem is that once you've adopted a bad mental model, it's really hard to rid yourself of it, and even though you can intellectually rid yourself of it it keeps on clouding your thinking.

A real beginner has no such problem. They appreciate that the image they see might be light or dark, and if they need a word to describe that, 'exposure' is not the most natural or useful. It's a jargon word they'll be unfamiliar with, and a misused one at that. 'Brightness' is more natural, but if you want a better jargon word, then the right one is 'lightness', which describes where a tone looks (perceptually) on the range of 'white' to 'black' rather than how much light it emits. Having raised the concept of how light or dark the final image is, it's then easy to talk about the factors which can determine how light or dark it is, which are at the simplest level 'exposure' and 'ISO'.

It's really very simple, as long as the beginner hasn't got confused by thinking that 'exposure' happens both at the input and output.

I will not argue against your position You obviously have excellent instructional background.

As far as I can see crop factor and exposure are easy to correctly teach providing the beginner does not have a bunch of bad ingrained ideas. I recently helped a grade 10 student with the effects of focal length on different size sensors using a 4 and a 6 in diameter magnifying glass projecting an image on two different sizes of paper representing sensor size. With very little guidance from me he managed to deduce the concept and arithmetic of crop factor by himself. It was slightly more difficult for him to wrap his head around exposure versus brightness since I had to introduce him to some basic Theory on how sensors actually operate. Certainly different teaching techniques could have been used.

 

[lightandaprayer]

If a photographer relies on camera feedback such as the JPEG-based histogram for ETTR, they are going to invariably leave at at least 2 stops of dynamic range on the table. But if they shoot RAW using a calibrated camera, they can consistently obtain the maximum DR possible for the specific sensor. Optimal exposure as well.

How to Use the Full Photographical Dynamic Range of Your Camera

 

[FafBill]

Love the quote. Thanks for passing it on.

 

[24Peter]

Love the quote. Thanks for passing it on.

Bill - your simple question has provoked quite a discussion!

I will defer to my learned colleagues on the points that have already been made (and unmade and made again.)

I would say for me I tend to err on the side of under-exposing an image, although what matters is what part of the scene you are exposing for. Once your brightest tones are clipped (over-exposed), there is no data to recover from the digital file. That's the biggest EVIL of digital photography.

As others have pointed out, RAW files from most modern digital cameras have lots of latitude so what may appear on the back of your camera as over-exposed, may in fact, not be.

I would also add, it depends on what you're worried about over-exposing. The scene as the camera meters it and the scene as seen by the eye of the photographer may be two different things. For instance, certain specular highlights might appear natural "blown out". So me thinks the hard and fast advocates of ETTR may be a bit zealous in their cause.

If you don't feel you have to get it right in camera, you have many more options when post processing an image on your computer. But alas, not everyone likes to spend time in front of their computer!

So... if there's something in the scene you want to be clearly seen by your viewers, then expose for that particular part of the image and protect against oversaturation of the digital photo sites (i.e., pixels). And if you are shooting RAW, rest assured you'll have some wiggle room.

Most importantly - enjoy the ride!

--
www.peteralessandriaphotography.com
Instagram
My DPReview Feature Article

 

[baloo_buc]

From my experience with sensors in Nikon cameras you can use an ISO 3 stops lower than the calculated one or an ISO that is 1 stop more in RAW and get the same photo as result using the same exposure (the same combination of aperture and shutter speed. In Fujifilm the limits are 1 stop under and 1 stop.

I will give an example. If the shutter speed must be 1/160 s to freeze slow motion and f/5.6 to get desired DOF I would get the same photo if I use ISO 400 to ISO 6400 on Nikon (ISO 3200 you don't need to push shadows or pull highlights in processing of RAW). With Fujifilm I can use ISO 1600 to ISO 6400.

With Olympus cameras I can have 2 stops under and 1 stop over (ISO 800 - ISO 6400).

So usually it is better to have a darker initial image than brighter as you have more latitude to push shadows.

 

[Iliah Borg]

Decades ago, I shot Black & White film. Recently, I have started shooting color with a digital camera that looks like my old 35mm camera. In the old days, if I was uncertain of the best exposure, I would bracket either the time or the f-shop. As a roll of B & W film was limited in the number of shots it would take, I would just pick 2 shots rather than 3. I think it was better to shoot once at the exposure the camera's meter recommended and the other, one stop above, to overexpose. I had a friend who was a professional and I think that he recommended a small overexposure rather than a small underexposure. This was some time ago.

Maybe I should just rephrase this. If one is slightly uncertain of the best approximate exposure should one, over or under, expose.

Depends on what you are trying to achieve. If you want to preserve more of the highlights, decrease the exposure; if you want to preserve details in shadows, increase it.

I use spot-meter mostly, and set appropriate EC, depending on what I put under the metering spot.

P.S. Film is such a different media (and mostly poorly known one, for example, pretty much every single photographer I've asked believes that push processing increases the sensitivity, while it only increases ISO rating) that I wouldn't bring it up at all, but if you are interested, Kodak study agrees with your friend - always err on the side of overexposure.

--

About LibRaw | LibRaw

www.libraw.org

 

[Son Of Waldo]

Decades ago, I shot Black & White film. Recently, I have started shooting color with a digital camera that looks like my old 35mm camera. In the old days, if I was uncertain of the best exposure, I would bracket either the time or the f-shop. As a roll of B & W film was limited in the number of shots it would take, I would just pick 2 shots rather than 3. I think it was better to shoot once at the exposure the camera's meter recommended and the other, one stop above, to overexpose. I had a friend who was a professional and I think that he recommended a small overexposure rather than a small underexposure. This was some time ago.

Maybe I should just rephrase this. If one is slightly uncertain of the best approximate exposure should one, over or under, expose.

Depends on what you are trying to achieve. If you want to preserve more of the highlights, decrease the exposure; if you want to preserve details in shadows, increase it.

I use spot-meter mostly, and set appropriate EC, depending on what I put under the metering spot.

P.S. Film is such a different media (and mostly poorly known one, for example, pretty much every single photographer I've asked believes that push processing increases the sensitivity, while it only increases ISO rating) that I wouldn't bring it up at all, but if you are interested, Kodak study agrees with your friend - always err on the side of overexposure.

that would be Iliah Borg, folks. Oh, and I almost forgot, in my humble opinion

By the way, thanks for the great price on the RawDigger / Fast Raw Viewer combo Iliah. I've done almost nothing with them thus far and don't really need fast culling on raw files (currently), but I do I plan on using them as learning tools.

I thank you for your (many) contributions to the photography community and as always, all the best to you.

 

[bobn2]

It was slightly more difficult for him to wrap his head around exposure versus brightness since I had to introduce him to some basic Theory on how sensors actually operate. Certainly different teaching techniques could have been used.

I don't think that you need any sensor theory to distinguish exposure and lightness, in fact it often leads astray. The main thing, with respect to understand what 'lightness' is, is to think about how the visual system works. One example I like to help people think about it is viewing a projector image on a white wall. Suddenly some of the white wall becomes black. So now one thinks about why that would be the case, and it gives you an introduction to what 'white' and 'black' mean.

 

[Marek M]

Generally it is easier to compensate in PP for underexposure than overexposure.

Do keep in mind that people like you and me and generally photographers worldwide will understand the above but here some will be able to twist the accepted meanning of words to the point that neither you or I will have a clue of what they mean.

Very well said and beautifully proven further in the thread.