Ugly Hedgehog - Photography Forum
Home Active Topics Newest Pictures Search Login Register
Main Photography Discussion
W/R winterrose The ETTR fallacy
Page <<first <prev 5 of 8 next> last>>
Apr 24, 2016 06:08:26   #
selmslie Loc: Fernandina Beach, FL, USA
 
Graham Smith wrote:
What you don't understand is that the camera manufacturers engineers are relying on this thread for guidance regarding future development of their products.

I realize that you were pulling our leg but they actually do listen to the public. But they probably know better than to come here for information.

Michael Reichmann's two articles (2003 and 2011) might have had some influence on the development of Active-D Lighting and other in-camera techniques that make the JPEG SOOC look better than when the camera's raw processing was less sophisticated. It has also made them adjust the JPEG histogram to more closely depict what is going on. That's probably why we have not heard from him since his last article.

Nevertheless, taking a proper exposure with little or no exposure compensation still leaves plenty of information on the table - in the raw file - for us to work with when we get back to the computer and do our own thoughtful raw conversion.

To deliberately and routinely overexpose an image by a full stop or more forces us to do more work in post processing and takes away the benefit of previewing the image on the camera's LCD.

Reply
Apr 24, 2016 06:25:53   #
lamiaceae Loc: San Luis Obispo County, CA
 
blackest wrote:

Are all photons created equal or do some have more energy than others? Lazers might suggest photons can have different energy levels so if raising the sensitivity of the sensor allows for these weaker photons to be recorded (along with more noise) than ettr still has a small advantage.


Oh boy, photographers need to know some basic physics. The energy of light or a photon is related by its wavelength or color. See Planck's equation below.





Reply
Apr 24, 2016 06:28:24   #
selmslie Loc: Fernandina Beach, FL, USA
 
lamiaceae wrote:
... The confusing part for me still is translating (exposure) stops or EV in to the 255 tone steps that seem to be available in Ps. That only seems to be 7 or 8 stops. And with 32-bit HDR we can create more. ???

Also from what one of my instructors thinks digital only real operates as zones say II to VIII. More like Slide film than Negative. Am I correctly relating the full "Ansel Adams" Zone range to 0 to 1024?

You might want to scan a thread that I posted recently, Stp Wedges (I noticed too late that I misspelled the title).

The information in the raw file will end up getting shoehorned into the 0-255 range in one way or another. The reality is that whether you divide that up into eleven zones or 18, 52, 83 or 256 evenly spaced steps, only the middle portion will contain any detail that is useful to the image.

The Zone System is a convenient way to describe the range:

0 and X are pure black and white
I to IX is the dynamic range of the image or print where you can see tonality
II to VIII is the range where we can see texture
III to VII is the range where we can see detail

These ranges are approximate but they were useful to Adams and still useful today.

Reply
 
 
Apr 24, 2016 06:31:32   #
selmslie Loc: Fernandina Beach, FL, USA
 
lamiaceae wrote:
Oh boy, photographers need to know some basic physics. The energy of light or a photon is related by its wavelength or color. See Planck's equation below.

Nice to know but do we really need to know it? It all still comes out between violet and red and we can see that. Seeing is really enough for a photographer and the viewer.

Reply
Apr 24, 2016 06:44:26   #
lamiaceae Loc: San Luis Obispo County, CA
 
selmslie wrote:
You might want to scan a thread that I posted recently, Stp Wedges (I noticed too late that I misspelled the title).

The information in the raw file will end up getting shoehorned into the 0-255 range in one way or another. The reality is that whether you divide that up into eleven zones or 18, 52, 83 or 256 evenly spaced steps, only the middle portion will contain any detail that is useful to the image.

The Zone System is a convenient way to describe the range:

0 and X are pure black and white
I to IX is the dynamic range of the image or print where you can see tonality
II to VIII is the range where we can see texture
III to VII is the range where we can see detail

These ranges are approximate but they were useful to Adams and still useful today.
You might want to scan a thread that I posted rece... (show quote)


Thanks for the info on

"The information in the raw file will end up getting shoehorned into the 0-255 range in one way or another. The reality is that whether you divide that up into eleven zones or 18, 52, 83 or 256 evenly spaced steps, only the middle portion will contain any detail that is useful to the image."

That is what I needed to know. The Zone System itself is very familiar to me, Ansel Adams, Minor White, Richard Zakia, Peter Lorenz. It is easy to think in steps as full stops, 2, 4, 8, 16, etc. But, yes, you are right, they do not have to be. What if each step were the Square Root of 2, ~ 1.41. Or 0.333.

Reply
Apr 24, 2016 07:03:05   #
selmslie Loc: Fernandina Beach, FL, USA
 
lamiaceae wrote:
... What if each step were the Square Root of 2, ~ 1.41. ...

Yes, you can it them up into any number of ranges. Three is not enough, 21 probably too many, eleven is just about right.

Adams would have probably run out of words or room on the blackboard to describe 21 zones rather than eleven.

Reply
Apr 24, 2016 07:39:20   #
John Howard Loc: SW Florida and Blue Ridge Mountains of NC.
 
Totally confused. I thought shifting to the right, without blowing out the highlights, added detail to the shadows and improved the noise to data ratio. So long as the highlights are not blown, which is critical, seems more detail in the shadows does effecrively enhance the dynamic range. I know some bird shooters that do this and shoot at rediculoisly high ISO hand held and get great results. Can't do the math but if there is a weakness in the discussion it is because you are mostly discussing the right side and Adams in the math it is full. Mostly it is not and the left side is empty.

Reply
 
 
Apr 24, 2016 07:49:07   #
blackest Loc: Ireland
 
Nikon USA wrote:

understanding iso sensitivity

(1)Photography is built on the three pillars of exposure: shutter speed, aperture and sensitivity. Shutter and aperture are controls for adjusting how much light comes into the camera. How much light is needed is determined by the sensitivity of the medium used. That was as true for glass plates as it is for film and now digital sensors. Over the years that sensitivity has been expressed in various ways, most recently as ASA and now ISO.

(2)The "normal" range of ISO is about 200 to 1600. With today's digital cameras you can sometimes go as low as 50 or as high as 204,800. The number chosen has two important qualities associated with it. First, it sets the amount of light needed for a good exposure. The lower the number, the more light required. The more light that's required, the more likely a slow shutter speed will have to be used. That means low ISOs, like 100 or 200, are most often used in bright situations (like sunlight) or when the camera is mounted on a tripod. If you don't have a lot of light, or need a fast shutter speed, you would probably raise the ISO.

(3)Each time you double the ISO (for example, from 200 to 400), the camera needs only half as much light for the same exposure. So if you had a shutter speed of 1/250 at 200 ISO, going to 400 ISO would let you get the same exposure at 1/500 second (providing the aperture remains unchanged). This is why high ISOs are so often used indoors, especially at sporting events. Needing a fast shutter speed to stop action, photographers regularly choose ISO 1600 or above.

(4)The other important quality tied to ISO is the amount of noise in the image. In the days of film, as you used film with higher ISO values (often referred to as ASA then), your images had more visible grain. Film grain is what made up the image, and higher numbers resulted in larger grain, which was more obvious. Most people found visible grain objectionable and so photographers worked to avoid it when possible.
In digital cameras, raising the ISO means a similar decrease in quality, with an increase in what's called "noise." It's the digital equivalent of grain and results in a sort of "chunky" look to the image. Very early digital cameras had objectionable levels of noise at ISOs as low as 800. Today most digital SLRs can make good quality images at ISOs up to 1600 and above. However, several variables affect this.

(5)One important factor affecting the amount of digital noise in an image is the size of the pixels used on the sensor. Large pixels result in less noise than small ones. That's why digital SLRs perform much better at high ISOs than compact cameras. The SLRs have larger sensors and larger pixels.

(6)Another factor is the amount and type of noise reduction being applied in the camera. Because all pixels collect some noise, every digital camera runs processing on every image (although with a NEF, or RAW, file that can be changed later) to minimize that noise. Newer cameras use newer technology to reduce that noise, with the result being less noise at similar ISOs than what earlier cameras could achieve.

(7)All of this means photographers are constantly doing a balancing act. They want to keep their ISO low for high quality images (low noise), but also they need a fast enough shutter speed to get a sharp picture. That's why there's such high value placed on groundbreaking cameras such as Nikon's D3, D700 and D3S that allow photographers to shoot at higher ISOs with less noise than ever before.

(8)Auto ISO was introduced into digital cameras several years ago to help photographers manage that balance. Turning on that feature allows the camera to push the ISO up when it decides the shutter speed is getting too low for a good picture. Even better, newer Nikon cameras have added "ISO Sensitivity Auto Control" to the menu choices. This takes Auto ISO and lets you have some say about what happens. Using it, you set the limit for how high it can go (800? 3200?) and at what shutter speed it should start raising the ISO (1/125? 1/30?). The amount of control this feature allows means more photographers will start taking advantage of it.

(9)A solid understanding of ISO will help you make smart decisions about how to set your camera. And that, in turn, will lead to better pictures.
br url=http://www.nikonusa.com/en/learn-and-expl... (show quote)
paragraph numbers added for reference

What nikon seem to be saying here is that higher iso values are a method for keeping up shutter speeds in order to reduce motion blur and camera shake(7) (8).

"Pixel size" matters, the size of the sensor and its pixels makes a difference to the amount of noise captured. Larger sensors with large pixels beat small sensors with small pixels(5). Can a small sensor be good enough, absolutely, as good as the full frame sensor image of the same scene next to it maybe not.

Noise reduction techniques strongly influence the appearance of noise, this can be done in camera or out of camera.(6) In camera techniques are limited by the image processor and the routines developed at the time.

It is a balancing act (7) (8) The optimal exposure as far as noise is concerned uses the lowest iso. if 1/125th at 5000 is the same as 1/125th at 16,000 then you can only do better in noise terms by increasing the exposure time, provided your subject doesn't move.

As far as I can tell, ETTR has advantage at base iso, sometimes, since it increases the capture time and the light recorded. Of course that has a cost if the subject moves. Also if overdone highlights can be lost.

Post processing the raw image can improve image quality over in camera techniques by customising the processing to suit the image. This also needs a skilled operator.

Under exposing really is not recording as much signal as was practical at the time and over exposing is recording at too high a level to be recorded within the the limits of the adc.

Some people might define underexposing or overexposing as recording an image so the tonality of the image is shifted up or down using default processing settings. I can't agree if the correct tones are produced cleanly as the final result then the exposure used was reasonable. If highlights are blown or the image is disfigured by noise in the final result it was over / under exposed.

reading

Exposing to the right explained

"If you aren’t shooting at your camera’s base ISO, ETTR is all but useless. For example, you wouldn’t want to shoot a scene at ISO 1600 and then decrease the exposure by one stop in Lightroom — it’s just as good to shoot the scene at ISO 800 in the first place, and that is less likely to blow out the highlights in your image anyway. The added noise from ISO 1600 would cancel out any benefits that come from darkening the photo in post-processing."

So Ettr only really good at base iso and with a cooperative subject.

Reply
Apr 24, 2016 08:00:23   #
blackest Loc: Ireland
 
lamiaceae wrote:
Oh boy, photographers need to know some basic physics. The energy of light or a photon is related by its wavelength or color. See Planck's equation below.


Thanks for that, it was a question you resolved it. So a given sensor has a fixed sensitivity to light and noise and digital iso is a trade off

Reply
Apr 24, 2016 08:15:58   #
PB73 Loc: Fremont, Ohio
 
W/R,

You might find the following (along with Part 2) informative.

Peter

Reply
Apr 24, 2016 08:30:32   #
winterrose Loc: Kyneton, Victoria, Australia
 
PB73 wrote:
W/R,

You might find the following (along with Part 2) informative.

Peter


Errrr.....yes.....waiting......

Reply
 
 
Apr 24, 2016 08:34:18   #
PB73 Loc: Fremont, Ohio
 
Sorry, Too early in the morning!

http://www.dpreview.com/articles/8189925268/what-s-that-noise-shedding-some-light-on-the-sources-of-noise

Reply
Apr 24, 2016 08:49:29   #
winterrose Loc: Kyneton, Victoria, Australia
 


Thanks PB but that is not news to me....

Reply
Apr 24, 2016 09:11:23   #
selmslie Loc: Fernandina Beach, FL, USA
 
blackest, everything that Nikon said is spot on and so is your recap.

Exposing to the Right Explained article wrote:
....(which you quoted) "If you aren’t shooting at your camera’s base ISO, ETTR is all but useless. For example, you wouldn’t want to shoot a scene at ISO 1600 and then decrease the exposure by one stop in Lightroom — it’s just as good to shoot the scene at ISO 800 in the first place, and that is less likely to blow out the highlights in your image anyway. The added noise from ISO 1600 would cancel out any benefits that come from darkening the photo in post-processing."

So Ettr only really good at base iso and with a cooperative subject.
....(which you quoted) "If you aren’t shootin... (show quote)

The Exposing to the Right Explained article sum it up nicely despite getting off to on the wrong foot with a shaky example.

What we are left with is the main premise of this thread, that ETTR is useless until you reach base ISO and then it becomes almost impossible to justify with a real life example. That's why nobody will supply us with a valid demonstration of a real scene. They simply can't do it!

Reply
Apr 24, 2016 10:02:35   #
burkphoto Loc: High Point, NC
 
winterrose wrote:
Others have had no trouble so please either try to offer something intelligent or if you choose not to I'm sure that there are lots of other threads you can join.

Cheers, Rob.


ETTR with JPEG processing in camera does result in overexposure if it is more than "expose just TO the right". EBTR DOES overexpose in-camera JPEGs. However, few camera manufacturers use ALL the available sensitivity of a sensor at a given ISO... And especially not at higher-than-base ISO. The histogram of a raw image often has some headroom --- some potential values under the point of saturation where all highlights are lost (except specular reflections, which should never have tonal values other than white).

The idea behind EBTR is not to *actually* overexpose the raw image, but to take advantage of the unused headroom. You gain detail in the shadows, and enable better compression of the recorded tonal range onto paper or into a post-processed 8-bit image.

All it is is a recognition that the in-camera histogram lies about raw files. It only shows the distribution of tones in the JPEG processed with the camera's current menu settings. When the brightest graph spikes are just touching the right side of the graph, the raw image still is not saturated (meaning it is not yet overexposed to the TRUE burnout point).

How much headroom exists to be utilized depends on many factors, and the practice must be tested with one's own gear and subject matter to learn the boundaries.

Reply
Page <<first <prev 5 of 8 next> last>>
If you want to reply, then register here. Registration is free and your account is created instantly, so you can post right away.
Main Photography Discussion
UglyHedgehog.com - Forum
Copyright 2011-2024 Ugly Hedgehog, Inc.