Ugly Hedgehog® - Photography Forum
ETTR followup
If you want to reply, then register here. Registration is free and your account is created instantly, so you can post right away.
Page: 1 2 next>>
Nov 18, 2018 14:21:37   #
gunflint
 
Last week I posted this question and got good feedback: "For my daytime exposures I have been attempting to always expose to the right and watching the histogram as to not clip any whites or blacks. It does seem to help with the dynamic range.

My question is (and I will do some tests myself) does this "rule" apply for night photography such as city lights after dark? Any tips from those with experience at this is appreciated!"

Then I was in the Tetons and tried this for scenery with stars and I am attaching 4 photos (RAW unedited) taken 30-40 minutes before sunrise. I realize they are not the same settings but the histogram is the issue. The darkest one was exposed with the histogram going only half way to the right (in the center), then a little more, and a little more, with the 4th one almost to the right. With the darkest one you can make out some stars but they are gone as it is exposed more to the right. So in this case with stars, I don't see how it is possible to expose the image any where near to the right and still see the stars.

Any feedback? Thank you


(Download)


(Download)


(Download)


(Download)

| Reply
Nov 18, 2018 14:31:30   #
tomcat
 
I don't know what the rest of UHH'ers will say, but my best experience with ETTR has been in bright sunshine conditions and not so much good results with darkened conditions.

| Reply
Nov 18, 2018 14:35:26   #
gunflint
 
tomcat wrote:
I don't know what the rest of UHH'ers will say, but my best experience with ETTR has been in bright sunshine conditions and not so much good results with darkened conditions.


That sure seems to be the case...

| Reply
Nov 18, 2018 14:45:52   #
Linda From Maine (a regular here)
 
I believe the idea is that you will "normalize" the exposure in initial edit of raw file. Edits can also be selective, darkening your sky more than your foreground, for example.

But there is always dynamic range to consider: https://www.bhphotovideo.com/explora/photography/tips-and-solutions/dynamic-range-explained

.

| Reply
Nov 18, 2018 14:48:23   #
BebuLamar (a regular here)
 
The question is at night with street lights and the street lights are a lot brighter than most of the scene do you still want to keep the street lights from clipping and have the rest of the scene very dark?

| Reply
Nov 18, 2018 15:01:39   #
gunflint
 
Linda From Maine wrote:
I believe the idea is that you will "normalize" the exposure in initial edit of raw file. Edits can also be selective, darkening your sky more than your foreground, for example.

But there is always dynamic range to consider: https://www.bhphotovideo.com/explora/photography/tips-and-solutions/dynamic-range-explained

.


Very good article, thanks!

| Reply
Nov 18, 2018 15:35:02   #
Bipod
 
gunflint wrote:
Last week I posted this question and got good feedback: "For my daytime exposures I have been attempting to always expose to the right and watching the histogram as to not clip any whites or blacks. It does seem to help with the dynamic range.

My question is (and I will do some tests myself) does this "rule" apply for night photography such as city lights after dark? Any tips from those with experience at this is appreciated!"

Then I was in the Tetons and tried this for scenery with stars and I am attaching 4 photos (RAW unedited) taken 30-40 minutes before sunrise. I realize they are not the same settings but the histogram is the issue. The darkest one was exposed with the histogram going only half way to the right (in the center), then a little more, and a little more, with the 4th one almost to the right. With the darkest one you can make out some stars but they are gone as it is exposed more to the right. So in this case with stars, I don't see how it is possible to expose the image any where near to the right and still see the stars.

Any feedback? Thank you
Last week I posted this question and got good feed... (show quote)

ETTR buys less noisy shadows at the price of possibly blown highlights. In the sample photos
the OP posted, the hightlights are snow fields, which has almost no detail anyway. This is almost
certain to make ETTR look good.

ETTR is a work-around for noisy digital sensors. Night photography involves long exposures
and/or cranking up the ISO. Either way, you get more noise than normal. So on the face of it,
a bit of overexposure is a good idea for night photography.

In long exposures, digital sensors don't have reciprocity failure like film does, but they accumulate
chroma noise. Also turning up the ISO always increases noise. Over-exposure can help improve the
image-to-noise ratio in shadows, which helps to justify risking a few blown highlights.

The real solution would be to have less noisy sensors. This is achieved in astronomical applications
by supercooling the sensor. That's not practical for everyday photography, but there has been some
improvement, with newer sensors being -- on the whole -- less noisy than old ones. If this trend continues,
we should less and less about ETTR for well lit situations.

Any "rule" for exposure that doesn't take the photographer's intentions (for how he wants the subject to
look) into account is just another form of automatic exposure. It will look how it looks, you may like it,
you may not.

I don't like colored speckles, so I use film and long exposures for night photography (and have to deal with
reciprocity failure). For a given brand/type/speed film, the manufacturer usually provides a curve of that
film's reciprocity failure. This graph can be used to estimate normal exposure. If a mathematical function
is fitted to the curve, then you can even have an exposure calculator for that film. If done correctly, the
resulting negative is no better or worse than one made in normal light.

Photography is about the final image. Camera's can't read minds, so the photographer has to play a role.
Only he knows what the objects in the scene are and how they generally look in real life. Only he knows
the image he is trying to create.

Rules can be helpful, but there are always multiple considerations. The "shoot lots and cull" approach
often amounts to picking the best of a bad lot.

So the answer is yes, ETTR is a helpful rule in digital night photography, but it's no substitute for visualization--
and choosing the exposure that produces the image you want. It's just another consideration: when
shooting with a noisy sensor, over-expose a bit.

| Reply
Nov 18, 2018 16:02:28   #
ballsafire
 
Linda From Maine wrote:
I believe the idea is that you will "normalize" the exposure in initial edit of raw file. Edits can also be selective, darkening your sky more than your foreground, for example.

But there is always dynamic range to consider: https://www.bhphotovideo.com/explora/photography/tips-and-solutions/dynamic-range-explained

.


Ah--Yes! I have read the article and enjoyed it very much; only to capture a very simple idea takes SO many words. I wish there were other ways for us to learn..so I'll dream on. By the way, this guy (in the discussion section) comparing HDR to a piano keyboard really cut many words out of the author's original explanation. Thanks for sharing your post on the subject of HDR, avery interesting concept.

| Reply
Nov 18, 2018 16:24:34   #
Linda From Maine (a regular here)
 
ballsafire wrote:
... Thanks for sharing your post on the subject of HDR, avery interesting concept.
When people see the letters "HDR," they often think of the technique wherein you combine several exposures. The HDR Forum on UHH: https://www.uglyhedgehog.com/s-107-1.html

A lot of new photographers aren't aware that a camera can't capture extremes of light and dark within the same image. I like this quote from the article:

"Dynamic range is an often overlooked aspect of photography, mainly due to the fact that it’s not something easy or always possible to control. Often creative decisions are made to negate the desire to have a fully controlled range of tones in order to favor a high- or low-key aesthetic, and on the other hand many photographers are very conscious of the apparent dynamic range and go through great lengths to compress as many stops and as much detail as possible into an image."

| Reply
Nov 19, 2018 06:33:44   #
cameraf4 (a regular here)
 
gunflint wrote:
... So in this case with stars, I don't see how it is possible to expose the image any where near to the right and still see the stars...


Don't really think you can. As the sky gets brighter, you lose the stars just like we do during the day.

| Reply
Nov 19, 2018 17:23:36   #
rmalarz (a regular here)
 
ETTR works at night as well as day. However, keep in mind two things. Your camera histogram is showing values from the in-camera jpg, not the RAW capture, which can very well capture even more highlight information. Secondly, the ETTR concept needs to be coupled with your processing techniques to gain full advantage of using this technique.
--Bob

gunflint wrote:
Last week I posted this question and got good feedback: "For my daytime exposures I have been attempting to always expose to the right and watching the histogram as to not clip any whites or blacks. It does seem to help with the dynamic range.

My question is (and I will do some tests myself) does this "rule" apply for night photography such as city lights after dark? Any tips from those with experience at this is appreciated!"

Then I was in the Tetons and tried this for scenery with stars and I am attaching 4 photos (RAW unedited) taken 30-40 minutes before sunrise. I realize they are not the same settings but the histogram is the issue. The darkest one was exposed with the histogram going only half way to the right (in the center), then a little more, and a little more, with the 4th one almost to the right. With the darkest one you can make out some stars but they are gone as it is exposed more to the right. So in this case with stars, I don't see how it is possible to expose the image any where near to the right and still see the stars.

Any feedback? Thank you
Last week I posted this question and got good feed... (show quote)

| Reply
Nov 19, 2018 17:27:13   #
williejoha
 
When ever I confront this issue I fall back on HDR. Very simple. It gives me three or more exposures to work with in PP. Nothing says that the final product has to be a HDR rendering.
WJH

| Reply
Nov 20, 2018 04:33:08   #
selmslie (a regular here)
 
gunflint wrote:
Last week I posted this question and got good feedback: "For my daytime exposures I have been attempting to always expose to the right and watching the histogram as to not clip any whites or blacks. It does seem to help with the dynamic range.

My question is (and I will do some tests myself) does this "rule" apply for night photography such as city lights after dark? Any tips from those with experience at this is appreciated!" ...

The problem doesn't have anything to do with ETTR.

In the fourth image the stars are not enough brighter than the sky to show any contrast.

As you record the sky darker, the stars begin to show up in #3 and a little more with each darker image.

The sky nearer the horizon is visibly lighter than overhead.

For the same reason, you can just make out the moon on a clear blue day but if there is a little haze you can't.

| Reply
Nov 20, 2018 08:13:15   #
gunflint
 
selmslie wrote:
The problem doesn't have anything to do with ETTR.

In the fourth image the stars are not enough brighter than the sky to show any contrast.

As you record the sky darker, the stars begin to show up in #3 and a little more with each darker image.

The sky nearer the horizon is visibly lighter than overhead.

For the same reason, you can just make out the moon on a clear blue day but if there is a little haze you can't.



The sky was not lighter or darker in any of the images, it was the exposure that was changed. They were all taken within a few minutes of each other.

| Reply
Nov 20, 2018 08:29:55   #
selmslie (a regular here)
 
gunflint wrote:
The sky was not lighter or darker in any of the images, it was the exposure that was changed. They were all taken within a few minutes of each other.

We know that. However, it was rendered lighter in the image by the additional exposure.

If you had post-processed all four images to achieve the same sky brightness as in the fourth image, the stars would have been equally indiscernible.

Only the brightest "star" (probably a planet) reaches an RGB value of 255,255,255 in the darkest image. None of the others get above middle gray.

And the longer the exposure, the longer the star trail. The star light gets spread out.

As they say in Vermont, "You can't get there from here."

| Reply
Page: 1 2 next>>
If you want to reply, then register here. Registration is free and your account is created instantly, so you can post right away.
UglyHedgehog.com - Forum
Copyright 2011-2019 Ugly Hedgehog, Inc.