-
Welcome to the new forums! Please read this first. For known issues we are working to resolve, click here.
ETTR in 2023
- Thread starter anfat
- Start date
No, the utility of ETTR ceased to exist on January 1 of this year. All MF cameras have firmware that enforces this.
Seriously, this keeps coming up, but the physics of photon counting hasn't changed.
Loading…
blog.kasson.com
MoreCowbell
Leading Member
I see no reason to stop
Erik Kaffehr
Veteran Member
#1 More photons yield a cleaner image
#2 Maximum exposure yields mor photons
For optimal IQ we need as much exposure as possible.
Recent improvements don't really change that.
Best regards
Erik
#2 Maximum exposure yields mor photons
For optimal IQ we need as much exposure as possible.
Recent improvements don't really change that.
Best regards
Erik
But not more. A little underexposure is better than clipping areas that you care about.
And the most important -- maybe the only important -- change over the last decade that affects DR is dual conversion gain, which offers no advantage at base ISO, where ETTR is the biggest win.
SrMi
Veteran Member
Yes, but I am confused why you would think it is not.
The only reason I can think someone would believe that times have changed is thinking that modern cameras have DR to burn.
SrMi
Veteran Member
Or maybe because ChatGPT has doubts about it
.In the hands of less competent photographers, ETTR may lead more often to clipped highlights. Therefore, if one does not know how to expose without clipping relevant highlights, it would be better to avoid ETTR.
Hard to argue with that.
yogi4fitness
Veteran Member
That would be me. I underexpose a bit because I’m better able to pull up shadows than recover highlights.Or maybe because ChatGPT has doubts about it
In the hands of less competent photographers, ETTR may lead more often to clipped highlights.
Does this mean we are back to “expose for detail in the highlights” from the old slide film days?
Erik Kaffehr
Veteran Member
The idea of using ETTR is to fully utilize the full well capacity of the sensor. SNR (Signal Noise Ratio) is proportional to the square root of the utilized well capacity.
Development in modern sensors mostly improves readout noise, that mostly affects very deep shadows.
So, for best image quality we would use techniques that yield the best signal noise ratio and that would be to use base ISO and expose ETTR, avoiding clipping significant highlights.
Increasing ISO mostly means using less of the full well capacity.
Now, it could be argued that signal noise ratio is very good on any sensor and also that sensors work well at reasonably high ISOs. That may relate a bit to improvements in processing.
Going to high ISOs essentially means giving up on SNR, at high ISOs readout noise will move upwards on the visibility ladder. Modern sensors have very low readout noise, especially when combined with dual gain conversion.
Some sensors apply noise reduction to the raw file at high ISOs.
Also, with older sensors, increasing ISO may reduce readout noise, thus improving dynamic range, but if exposure is reduced, shot noise will always increase relative to signal.
Best regards
Erik
Development in modern sensors mostly improves readout noise, that mostly affects very deep shadows.
So, for best image quality we would use techniques that yield the best signal noise ratio and that would be to use base ISO and expose ETTR, avoiding clipping significant highlights.
Increasing ISO mostly means using less of the full well capacity.
Now, it could be argued that signal noise ratio is very good on any sensor and also that sensors work well at reasonably high ISOs. That may relate a bit to improvements in processing.
Going to high ISOs essentially means giving up on SNR, at high ISOs readout noise will move upwards on the visibility ladder. Modern sensors have very low readout noise, especially when combined with dual gain conversion.
Some sensors apply noise reduction to the raw file at high ISOs.
Also, with older sensors, increasing ISO may reduce readout noise, thus improving dynamic range, but if exposure is reduced, shot noise will always increase relative to signal.
Best regards
Erik
Erik Kaffehr
Veteran Member
That has always been the case...
But, with slide film we needed to care about mid tones, as slide film was intended for direct viewing, on a lightbox or screen.
Best regards
Erik
--
Erik Kaffehr
Website: http://echophoto.dnsalias.net
Magic tends to disappear in controlled experiments…
Gallery: http://echophoto.smugmug.com
Articles: http://echophoto.dnsalias.net/ekr/index.php/photoarticles
Last edited:
Not a huge fan of ETTR in practice. I’ve often found that bringing the exposure, or bits of it, back down results in colour shifts and odd transitions. The result can look off or like a poorly done HDR. Usually, I’ll choose whether to exposure for highlights or shadows and let the rest fall where it will. I rarely feel the need to go for verisimilitude, since imho all images are interpretations anyway.
still a case of blow the highlights and there was nothing you do about it and with negatives, underexpose the shadows and there was no silver to work on.
Last edited:
anyway with the X2D the histogram or the blinking lights refer to the jpeg image and not to the raw...
I work this way: I use a bracket set of 3 exposures for each shot; the base exposure is 1+1/3 stops more than the camera meter, the next two bracketed exposures are +2/3 and -2/3 stops from the base exposure; the result are 3 exposures ranging from +2/3 to 2 stops more than camera meter. Then in the RAW developer I choose the exposure which has no clipping highlights.
I work this way: I use a bracket set of 3 exposures for each shot; the base exposure is 1+1/3 stops more than the camera meter, the next two bracketed exposures are +2/3 and -2/3 stops from the base exposure; the result are 3 exposures ranging from +2/3 to 2 stops more than camera meter. Then in the RAW developer I choose the exposure which has no clipping highlights.
Last edited:
Every shot, seems like a good way to wear out the shutter. After a while, don’t you just get a feeling of what is needed from looking at which of the three exposures you tend to use?
There's a way to improve that:
Using in-camera histograms for ETTR - the last word
If you're interested in the whole topic, it's probably a good idea to read each section in order. If you just want to do the calibration and have access to Rawdigger, Photoshop, and Excel, read this: Preparing for monitor-based UniWB and do this: A one-step UniWB procedure. Acknowledgments Why...
blog.kasson.com
SrMi
Veteran Member
Blown highlights cause color shifts and odd transitions. That is a common complaint about ETTR but is caused by incorrect exposure (loss of relevant data). Since I have never seen it in my files, I would like to see an example where ETTR hurts colors. FWIW, if you clip relevant highlights, you are not doing ETTR.
An often-heard argument against ETTR is that the sensors are non-linear, and moving the exposure brings the data into non-linear areas, which then look weird when bringing the brightness down. AFAIK, sensors work only (mainly?) in their linear range; the post-processor adds non-linearity.
SrMi
Veteran Member
It is OK to "believe" in ETTR usefulness and not use it because:
- UniWB too complicated
- No time to optimize exposure
- Plenty of performance with MF sensor even if reducing the IQ to that of FF or m43
- Too tired (synonym for lazy?) to use bracketing
We should not avoid using ETTR because of urban legends.
- UniWB too complicated
- No time to optimize exposure
- Plenty of performance with MF sensor even if reducing the IQ to that of FF or m43
- Too tired (synonym for lazy?) to use bracketing
We should not avoid using ETTR because of urban legends.
Keyboard shortcuts
- f
- Forum