Ugly Hedgehog - Photography Forum
Home Active Topics Newest Pictures Search Login Register
Landscape Photography
ETTR in landscape photography
Page <<first <prev 5 of 7 next> last>>
Dec 6, 2018 15:07:21   #
R.G. Loc: Scotland
 
selmslie wrote:
Absolutely not! That would only be the case when shooting at base ISO or lower.


OK - you're still saying that at base ISO the raw file maxes out at the sensor's limits.

selmslie wrote:
........the sensor still uses the full physical capacity of the sensor......


Others have suggested that the only way to use the sensor's full capacity is to deliberately overexpose. It has also been suggested that the conditions that get the raw file to max out aren't the same conditions that get the sensor to max out (even if base ISO is used throughout).

Reply
Dec 6, 2018 15:11:54   #
gunflint Loc: Rocky Mountain High, Colorado
 
Beautiful! And thanks for showing us the original raw file.

Reply
Dec 6, 2018 15:27:17   #
dsmeltz Loc: Philadelphia
 
selmslie wrote:
I agree!

Dave and I should have this conversation offline but he has me on his Ignore list.

As an alternative, he can start a new thread in the Photo Gallery like I did or in the Main Photography Discussion and move the discussion of ETTR, etc., out of the Landscape Photography forum.

Then he can post a RawDigger analysis of both of his images. It will tell everyone precisely how many pixels for each of the three colors are beyond the reach of the raw file in the two images he has posted.
I agree! br br Dave and I should have this conver... (show quote)


ETTR is an important discussion for landscape. Can't the two of you settle your dispute and enter a neutral discussion? You both appear to have something worth saying.

Reply
 
 
Dec 6, 2018 17:05:45   #
selmslie Loc: Fernandina Beach, FL, USA
 
R.G. wrote:
Others have suggested that the only way to use the sensor's full capacity is to deliberately overexpose. It has also been suggested that the conditions that get the raw file to max out aren't the same conditions that get the sensor to max out (even if base ISO is used throughout).

If you use the sensors full capacity and you are not at base ISO, the gain will cause the numbers stored in the raw file to max out. For example, at ISO 200 (gain=2x) anything over 50% of the sensors capacity will max out at 16383 (14-bit). At ISO 800 (gain=8x) anything over 12.5% of the sensors capacity will max out at 16383.

The point is that you don't want to max out the raw file. It's the only record you have of the level of exposure received by the sensor.

Reply
Dec 6, 2018 17:49:04   #
John Howard Loc: SW Florida and Blue Ridge Mountains of NC.
 
This is a great thread and I appreciate having joined this group this morning.
Question for you... How do you find the Eradr and do #3?
I have been using ETTR a bit (and after reading this more) but have never tried to determine how much extra I am leaving in the tank. I know some cameras can be adjusted to determine how close the blinkies are to blown highlights (ie 95%). Don't know about my Nikons so will investigate. But how do you determine what room is left to the right?

Uuglypher wrote:
Here’s the long-story-short on EBTR landscape exposure.

1. set you camera for raw capture.

2. find the exposure you would use to get the brightest jpeg image possible without clipping highlights ( That’s the ETTR exposure)

3. ADD to the ETTR exposure your camera’s pre-determined extra raw-accessible dynamic range (ERADR) to the ETTR exposure and ... (note that the camera I used for the example below has one and 1/3 stops of ERADR at base ISO.)

4. squeeze the shutter... and Bob’s your uncle!

Then you will notice that in your camera’s display the your image is washed out with blown highlights!
“Overexposed” you s ream...stamping your feet!

Well.. Aaaaactually......it’s not! Would be, were it a jpeg file, but lucky you, you shot a raw capture and got a perfectly exposed image file of the highest possible image data quality.

Wutcha do next is open it in your raw converter, normalize its tonality by sliding the “Exposure” slider to the left...and lo and behold...there is your tonally perfect image...and nary a clipped highlight to be found!

Dave
Here’s the long-story-short on EBTR landscape exp... (show quote)

Reply
Dec 6, 2018 18:54:01   #
selmslie Loc: Fernandina Beach, FL, USA
 
John Howard wrote:
This is a great thread and I appreciate having joined this group this morning.
Question for you... How do you find the Eradr and do #3?
I have been using ETTR a bit (and after reading this more) but have never tried to determine how much extra I am leaving in the tank. I know some cameras can be adjusted to determine how close the blinkies are to blown highlights (ie 95%). Don't know about my Nikons so will investigate. But how do you determine what room is left to the right?

You might want to look at the the links in my earlier post. You may find that all Sony and Nikon cameras start to display blinkies (where the JPEG starts to blow out) one stop before the raw file gets blown out. This is by design.

Reply
Dec 7, 2018 00:54:40   #
Uuglypher Loc: South Dakota (East River)
 
John Howard wrote:
This is a great thread and I appreciate having joined this group this morning.
Question for you... How do you find the Eradr and do #3?
I have been using ETTR a bit (and after reading this more) but have never tried to determine how much extra I am leaving in the tank. I know some cameras can be adjusted to determine how close the blinkies are to blown highlights (ie 95%). Don't know about my Nikons so will investigate. But how do you determine what room is left to the right?


Hi, John,
Here is a link to an EBTR tutorial
https://www.uglyhedgehog.com/t-372364-1.html
(Tutorial-Welcome to raw exposure)...includes helpful basics of EBTR, including how to test your camera’s ERADR.

Dave

Reply
 
 
Dec 7, 2018 08:45:23   #
John Howard Loc: SW Florida and Blue Ridge Mountains of NC.
 
Uuglypher wrote:
Hi, John,
Here is a link to an EBTR tutorial
https://www.uglyhedgehog.com/t-372364-1.html
(Tutorial-Welcome to raw exposure)...includes helpful basics of EBTR, including how to test your camera’s ERADR.

Dave


Thanks Dave.

Reply
Dec 7, 2018 10:28:03   #
selmslie Loc: Fernandina Beach, FL, USA
 
R.G. wrote:
.... It has also been suggested that the conditions that get the raw file to max out aren't the same conditions that get the sensor to max out (even if base ISO is used throughout).

This might help clear things up: https://www.uglyhedgehog.com/t-567359-1.html#9658823 It shows that, although the sensor is not maxed out, increasing the ISO can max out the raw file and destroy the image.

Reply
Dec 7, 2018 13:03:56   #
R.G. Loc: Scotland
 
selmslie wrote:
This might help clear things up: https://www.uglyhedgehog.com/t-567359-1.html#9658823 It shows that, although the sensor is not maxed out, increasing the ISO can max out the raw file and destroy the image.


I understand that ISO just amplifies what the sensor is outputting and I understand that to get a larger signal from the sensor it needs to capture more light - which is achieved via the shutter speed and aperture settings.

What I find difficult to understand is how quite a few competent and meticulous people are managing to find two or more extra stops of dynamic range beyond the histogram/jpeg limit (and they have the photos to prove it). That's assuming that the jpeg limit is always one stop below the raw limit. Even if the sensor itself isn't maxing out, if the raw file is maxed out the highlights will be blown. The only possibility I can think of is that the histograms aren't accurately showing the jpeg limit, but that doesn't sound very feasible, especially since the discrepancy appears to be at least one stop in most cases.

Reply
Dec 7, 2018 14:10:35   #
Uuglypher Loc: South Dakota (East River)
 
R.G. STATED:
“What I find difficult to understand is how quite a few competent and meticulous people are managing to find two or more extra stops of dynamic range beyond the histogram/jpeg limit (and they have the photos to prove it). That's assuming that the jpeg limit is always one stop below the raw limit. Even if the sensor itself isn't maxing out, if the raw file is maxed out the highlights will be blown. The only possibility I can think of is that the histograms aren't accurately showing the jpeg limit, but that doesn't sound very feasible, especially since the discrepancy appears to be at least one stop in most cases.”

Well stated! The erroneous assumption is that “That's assuming that the jpeg limit is always one stop below the raw limit.”

As you acknowledge, ‘Taint so! In fact, by testing cameras of different brands, as well as cameras of even the same model of the same brand it is obvious that the gap between the JPEG clipping limit and the clipping limit of the extant raw dynamic range in each individual camera may be as little as 1/3 stop of exposure to as much as Two and 2/3 stops (or, by reports/rumors) as much as three full stops or more.

Hence , if one’s expectation is to capture the highest quality raw image data possible, each camera must be tested for its complement of ERADR beyond the ETTR exposure (the JPEG clipping point).

Basically, the rules of JPEG exposure are those that hold for sensitive emulsion photography: expose such that the tonal value of a given detail will be rendered that same value in the rendered image. This was the system that “Sunny 16” works well for, and is the basis of the jpeg-adjusted exposure system (histogram frame and histogram) built into our digital cameras.


However, The rule for exposing for optimimal raw image data quality is to EXPOSE FOR THE BRIGHTEST POSSIBLE IMAGE WITHOUT CLIPPING HIGHLIGHT DETAIL. EXPOSE AS CLOSE TO CLIPPING (WITHIN 1/3 STOP) WITHOUT ACTUALLY DOING SO. And remember, clipping of highlight detail is revealed only in the raw converter after the series of test exposures are tonally normalized.

And the only way to do that is to use the maximum of the raw-accessible Dynamic range.

And the only way to do that is to know your individual camera’s allotment of ERADR.

And the most efficient way to determine that is by a simple series of test exposures beyond the ETTR exposure with each increasing by 1/3 stop until a raw exposure with clipped highlights is encountered.

NOT brain surgery or rocket science!

It is well to remember that 8-bit JPEG image files on the one hand and 12-bit, 14-bit, and 16-bit raw image data files on the other hand are two, significantly different imaging media.

The jpeg file, if properly exposed, is delivered close to the photographer’s previsualized image.

The raw file, properly exposed, holds the explicit promise of yielding not only the photographer’s previsualized image, but as well as the promise of providing an infinite number of different artistically creative variations that may later be “post-And the only way to do that is to use the maximum of the raw-accessible Dynamic range.
visualized”... all rendered using the highest quality image data possible of being captured, with the highest S:N ratio, and, relative to the file’s bit-depth, the greatest possible tonal and hue spectra providing the greatest possible detail delineation thanks to unexcelled limits of potential cusps of detail (every junction between two adjacent tones of gray is a potential edge of a to-be-revealed detail).

It has been suggested that a raw file has the potential of a negative awaiting further darkroom work. In truth, it is far more than that. A raw file is much akin to a pluripotent un-processed virtual image, awaiting only the photographers creative imagination to process it in an immensely wide variety of forms...from low- to high-key, color or black&white, and all possible variations in between!

8-bits versus 12-bits or more....a helluva difference in creative potential !

Dave

Reply
 
 
Dec 7, 2018 14:55:26   #
R.G. Loc: Scotland
 
Uuglypher wrote:
.........The raw file, properly exposed, holds the explicit promise of yielding not only the photographer’s previsualized image, but as well as the promise of providing an infinite number of different artistically creative variations that may later be “post" And the only way to do that is to use the maximum of the raw-accessible Dynamic range visualized”... all rendered using the highest quality image data possible of being captured, with the highest S:N ratio, and, relative to the file’s bit-depth, the greatest possible tonal and hue spectra providing the greatest possible detail delineation thanks to unexcelled limits of potential cusps of detail (every junction between two adjacent tones of gray is a potential edge of a to-be-revealed detail).
.........The raw file, properly exposed, holds the... (show quote)


My rather simplistic take on what you just said is that a file that's been over-exposed (as described above) will take much more pushing and pulling, the noise will be less or non-existent, the colours will be truer, especially where subtle gradations are concerned. That all adds up to a file that's going to produce better results and it will be easier to achieve those superior results.

That will be especially true if the image has shadow areas that need brightening. Even a shot taken at base ISO can get noisy and flaky if it's brightened significantly. But darkening isn't going to produce or aggravate any noise, so coming down from the bright end of the luminosity spectrum is the safer option. Getting good results easily sounds like my kind of editing .

Reply
Dec 7, 2018 15:22:27   #
Uuglypher Loc: South Dakota (East River)
 
R.G. wrote:
My rather simplistic take on what you just said is that a file that's been over-exposed (as described above) will take much more pushing and pulling, the noise will be less or non-existent, the colours will be truer, especially where subtle gradations are concerned. That all adds up to a file that's going to produce better results and it will be easier to achieve those superior results.

That will be especially true if the image has shadow areas that need brightening. Even a shot taken at base ISO can get noisy and flaky if it's brightened significantly. But darkening isn't going to produce or aggravate any noise, so coming down from the bright end of the luminosity spectrum is the safer option. Getting good results easily sounds like my kind of editing .
My rather simplistic take on what you just said is... (show quote)


As usual, you excell at extracting from my logorrheic dissertations the basic points that I had hoped to impart !

However...you used the word “overexposed” which requires that I quickly take another dose of my high blood pressure medication. Proper exposure of a raw capture, when viewed in-camera as if it were a jpeg image (the only way it can be viewed) will be thought to be “overexposed”...when, in fact,as a raw image file, that “washed-out, blown-highlight-laden image” is just as a perfectly exposed raw capture ought appear in the camera display that knows only how to display jpegs !!

Harrumph!

I might also have added that proper raw exposure assures that every level of brightness in the captured scene (every vertical line in the histogram) will have been captured at the greatest level of brightness possible, relative to all the other brightness levels in thehistogram...from the darkestdiscernible shadow detail to the highlight detail barely discernible adjacent to a specular highlight. This means that the raw file so exposed does, in practical fact, contain the absolute, greatest creative potential of any image able to be made of the captured scene.

I know you know that...I just hate to encourage the term “overexposed” in the context of raw exposure unless it is observed as an undeniable blown highlight with tonal normalization in the raw converter.

Best regards,
Dave

Reply
Dec 7, 2018 15:35:51   #
R.G. Loc: Scotland
 
Uuglypher wrote:
....I just hate to encourage the term “overexposed” in the context of raw exposure.....


Again I'll reduce things to a level of simplicity that I feel comfortable with . "Overexposed" means "brighter than the camera's suggested exposure level (with exposure compensation set to zero)".

Reply
Dec 7, 2018 15:56:57   #
selmslie Loc: Fernandina Beach, FL, USA
 
R.G. wrote:
... What I find difficult to understand is how quite a few competent and meticulous people are managing to find two or more extra stops of dynamic range beyond the histogram/jpeg limit (and they have the photos to prove it). ... The only possibility I can think of is that the histograms aren't accurately showing the jpeg limit, but that doesn't sound very feasible, especially since the discrepancy appears to be at least one stop in most cases.

I don't doubt the earnestness of many proponents of ETTR/EBTR.

The bottom line is that the only way to be sure about what gets into the raw file is through the use of a tool like RawDigger. It reports the raw numeric values before the camera or computer translates that information into a visible image. Showing an image after it has been converted from raw is simply not the a way to prove anything.

I have been presenting the RawDigger evidence of images captured as non-ETTR, ETTR, EBTR and with blown out raw highlights. You can see this in ETTR vs. ISO Invariance where I explore both extremes from severe underexposure to overexposure with blown highlights. The images posted include the RawDigger histograms and the full-size images that were not processed in any way other than to move the Exposure slider in Capture One Pro where indicated.

Reply
Page <<first <prev 5 of 7 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.
Landscape Photography
UglyHedgehog.com - Forum
Copyright 2011-2024 Ugly Hedgehog, Inc.