p.6 #1 · Easy ETTR technique using the Zebra pattern
joakim wrote:
This question doesn't make any sense to me, maybe I am missing something.
If you shot a scene with a 1s "normal" exposure and then the same scene properly adjusted for ETTR at 2s, how could the 1s exposure in any way be better than the 2s exposure? In what sense could shadow recovery be worse if you have doubled the amount of light, you have more data to work with and you have to do less recovery?
The goal of this ETTR technique is to preserve all highlight data while capturing as much shadow detail as possible.
Using the zebra technique, the first exposure will have all highlight data needed and as long as it's not too underexposed, it's all we need.
However, if the first capture (exposed to the highlights using the zebra technique) is underexposed by more than a stop, bracketing is the solution as described on the first post.
p.6 #2 · Easy ETTR technique using the Zebra pattern
Fred Miranda wrote:
The goal of this ETTR technique is to preserve all highlight data while capturing as much shadow detail as possible.
Using the zebra technique, the first exposure will have all highlight data needed and as long as it's not too underexposed, it's all we need.
However, if the first capture (exposed to the highlights using the zebra technique) is underexposed by more than a stop, bracketing is the solution as described on the first post.
I'm with you on that and I think I wasn't weary clear in my example. I meant a scene where there was no problem with dynamic range but enough headroom to increase the camera suggested exposure with one stop without burning any highlights. Compared to doing shadow recovery in post that must always be the better option?
p.6 #3 · Easy ETTR technique using the Zebra pattern
MJKoski wrote:
It makes a difference with compressed RAW. There are gaps in highlights and pulling those back may cause unwanted banding in tonal transitions.
Ok, I read the post to quickly and was thinking about ETTR in general even though the A7RII was mentioned.
p.6 #4 · Easy ETTR technique using the Zebra pattern
joakim wrote:
I'm with you on that and I think I wasn't weary clear in my example. I meant a scene where there was no problem with dynamic range but enough headroom to increase the camera suggested exposure with one stop without burning any highlights. Compared to doing shadow recovery in post that must always be the better option?
Yes, the ETTR Zebra technique makes a difference even for scenes where dynamic range does not exceed the sensor's capacity. You will still get much cleaner shadows after post-processing. For very high contrast scenes (example: sun in the frame), ETTR Zebra will give us an underexposed image, so using the bracketing technique is always needed for optimal shadow detail.
I find myself usually overexposing my images 1 or 2 stops (As long as ETTR Zebra technique allows me) and getting very clean shadows without any loss of highlight detail.
p.6 #5 · Easy ETTR technique using the Zebra pattern
True in theory, I've often wondered if true in practice. Do you have any examples to hand of compressed vs uncompressed files if the same scene with pulled back shadows? Yep
MJKoski wrote:
It makes a difference with compressed RAW. There are gaps in highlights and pulling those back may cause unwanted banding in tonal transitions.
p.6 #8 · Easy ETTR technique using the Zebra pattern
White highlights on predawn sky which was red like the reflection below horizon. I should have been able to pull soft reds out of the highlights but there is nothing there. Just an crude bands of washed out tones. According to RawDigger nothing was clipped.
p.6 #9 · Easy ETTR technique using the Zebra pattern
joakim wrote:
I'm with you on that and I think I wasn't weary clear in my example. I meant a scene where there was no problem with dynamic range but enough headroom to increase the camera suggested exposure with one stop without burning any highlights. Compared to doing shadow recovery in post that must always be the better option?
Increasing exposure will always give you better shadows. That's the whole point of ETTR.
p.6 #10 · Easy ETTR technique using the Zebra pattern
retrofocus wrote:
Have to test this with my A7R, I suspect it works there, too. So far I decided to turn the whole zebra thing off after I experimented with it for a while. I found the blinking highlights too distracting especially for landscape photography where the sky is always brighter than the foreground. Also, I find the percentage number very confusing since I expected to see more zebra lines at higher percentage which is not the case how it works (instead it is more a sensitivity number for highlights).
I again tested the ETTR method recently with my A7R - it works as shown, but I decided that I rather do this manually with the exposure wheel than using the zebra pattern function. I still don't like the blinking distractive pattern which the zebra provides. With digital I simply tend more to underexpose to avoid clipping highlights (opposite to film where you need to expose for the shadows and overexpose between 1/2 to 1 stop or more). This rule of thumb works pretty well.
p.6 #11 · Easy ETTR technique using the Zebra pattern
retrofocus wrote:
I again tested the ETTR method recently with my A7R - it works as shown, but I decided that I rather do this manually with the exposure wheel than using the zebra pattern function. I still don't like the blinking distractive pattern which the zebra provides. With digital I simply tend more to underexpose to avoid clipping highlights (opposite to film where you need to expose for the shadows and overexpose between 1/2 to 1 stop or more). This rule of thumb works pretty well.
My general workflow:
Into the light (and the light is an important part: -1EV with +-2EV bracket (or even 5 shots when I'm lazy)
Normal Situation: +0.3EV to +1EV
Low DR Situation: +1EV up to +1.7EV
If I'm very lazy, I'll shoot just with +0EV
If I plan my shot of the year, I use the described ETTR method and add a bracket afterwards.
p.6 #12 · Easy ETTR technique using the Zebra pattern
sebbe wrote:
My general workflow:
Into the light (and the light is an important part: -1EV with +-2EV bracket (or even 5 shots when I'm lazy)
Normal Situation: +0.3EV to +1EV
Low DR Situation: +1EV up to +1.7EV
If I'm very lazy, I'll shoot just with +0EV
If I plan my shot of the year, I use the described ETTR method and add a bracket afterwards.
Very nice workflow Seb! In optimal low contrast light, I find myself having the dial at +1.3EV most of the time.
p.6 #13 · Easy ETTR technique using the Zebra pattern
Thanks for the information on this technique, seems to be great...
In real world usage, is it only really for the most high contrast scenes, like sunrise/sunsets that often also require the use of a tripod?
When I have tried it handheld, at base 100 iso around dusk time, I generally find that opening up two stops often leads to an overall much brighter image overall and will require a shutter speed that is borderline too slow for handholding?
p.6 #14 · Easy ETTR technique using the Zebra pattern
Apologies for bringing up an old thread, but I did some testing over the weekend and would like to share the results.
I followed methodology outlined in the video tutorials on digital camera exposure by Greg Benz in order to find out how much highlight information is preserved in the RAW compared to in-camera clipping warnings, which for my tests were based on the Zebra 100+ exposure warnings described for ETTR in this thread:
I had the camera on a sturdy tripod when capturing the images, with a Canon FD 2.8/200mm at f5.6. I framed for a blue part of the sky and took all exposures within a couple of minutes. The idea is that any shifts in color in the brightest parts of the image will indicate that one of the channels–in this case the blue channel–has been clipped beyond recovery. The images were reviewed in Capture One Pro 10, with the Base Curve set to "Linear Response". This camera profile documents the image information available without requiring highlight recovery. For the first set of images (leftmost column) I exposed the image to the brightest setting before encountering the Zebra 100+ warning. I then continued to increase the exposure 1/3 EV for every consecutive shot. In post the exposure was normalized to reveal color shifts due to clipping of the blue channel.
For these tests I was particularly interested in seeing if different camera settings (silent shutter, compressed/uncompressed raw, base vs. elevated ISO) would influence how much highlight information was preserved beyond the "Zebra 100+" exposure warning.
To my surprise I found significant color shifts happening when increasing the exposure beyond 2/3 (0.67) EV relative to the Zebra 100+ clipping warning. The ISO 640 images seemed to have 1/3 EV more headroom in the highlights. The dropdown from 14 to 12-bit image data associated with the Silent Shutter had little deterioration at ISO 100, but significant negative effect at ISO 640. Using the uncompressed RAW option gave at ISO 100 approximately 1/3 EV more highlight headroom, and had negligible effect at ISO 640.
I'm very interested in hearing your thoughts on this test, and would on the base of these results warn against overexposing the highlights by more than 2/3 (0.67) EV relative to the Zebra 100+ clipping indicator.
p.6 #15 · Easy ETTR technique using the Zebra pattern
andersd wrote:
Apologies for bringing up an old thread, but I did some testing over the weekend and would like to share the results.
I followed methodology outlined in the video tutorials on digital camera exposure by Greg Benz in order to find out how much highlight information is preserved in the RAW compared to in-camera clipping warnings, which for my tests were based on the Zebra 100+ exposure warnings described for ETTR in this thread:
I had the camera on a sturdy tripod when capturing the images, with a Canon FD 2.8/200mm at f5.6. I framed for a blue part of the sky and took all exposures within a couple of minutes. The idea is that any shifts in color in the brightest parts of the image will indicate that one of the channels–in this case the blue channel–has been clipped beyond recovery. The images were reviewed in Capture One Pro 10, with the Base Curve set to "Linear Response". This camera profile documents the image information available without requiring highlight recovery. For the first set of images (leftmost column) I exposed the image to the brightest setting before encountering the Zebra 100+ warning. I then continued to increase the exposure 1/3 EV for every consecutive shot. In post the exposure was normalized to reveal color shifts due to clipping of the blue channel.
For these tests I was particularly interested in seeing if different camera settings (silent shutter, compressed/uncompressed raw, base vs. elevated ISO) would influence how much highlight information was preserved beyond the "Zebra 100+" exposure warning.
To my surprise I found significant color shifts happening when increasing the exposure beyond 2/3 (0.67) EV relative to the Zebra 100+ clipping warning. The ISO 640 images seemed to have 1/3 EV more headroom in the highlights. The dropdown from 14 to 12-bit image data associated with the Silent Shutter had little deterioration at ISO 100, but significant negative effect at ISO 640. Using the uncompressed RAW option gave at ISO 100 approximately 1/3 EV more highlight headroom, and had negligible effect at ISO 640.
I'm very interested in hearing your thoughts on this test, and would on the base of these results warn against overexposing the highlights by more than 2/3 (0.67) EV relative to the Zebra 100+ clipping indicator.
Thanks for posting your findings!
After a couple years using the zebra technique posted here for landscapes, I am certain it's able to retain all highlight data when normalized in post. I've tested side-by-side images exposed for the highlights against the same scene using the zebra technique and after recovering the highlights for the latter, there is absolutely no visible difference. Perhaps there will be a slight color shift but the highlight data is all there. The benefit of much cleaner shadows is very noticeable.
p.6 #16 · Easy ETTR technique using the Zebra pattern
Fred Miranda wrote:
Thanks for posting your findings!
After a couple years using the zebra technique posted here for landscapes, I am certain it's able to retain all highlight data when normalized in post. I've tested side-by-side images exposed for the highlights against the same scene using the zebra technique and after recovering the highlights for the latter, there is absolutely no visible difference. Perhaps there will be a slight color shift but the highlight data is all there. The benefit of much cleaner shadows is very noticeable.
Thanks for chiming in Fred! Could it be the case that the Zebra 100+ warning is based on luminosity (mean R, G and B value) rather than any individual color channel being clipped?
With that being said, I do agree that an increase in exposure and associated benefits to shadow detail in many cases will be preferable to (slight) highlight clipping of one of the color channels in the highlights.
p.6 #17 · Easy ETTR technique using the Zebra pattern
Hi - I’ve been reading up on ETTR via this thread and tried it out for the first time this past weekend, and think I’m not doing something quite right as the images are so overexposed that to get the information in some of the highlights (the sky) that in post processing I would have to decrease not just highlights all the way, but decrease exposure by 1.5 to 2 stops and then the rest of the image is way too dark, even after pulling up shadows. Here’s how I’ve set things up - would greatly appreciate any input on what I may be doing wrong:
Camera: A7RII
File format: RAW, compressed
ISO: 100
Zebra: 100+
Creative style : Standard with all settings at 0 (although I read a thread on DPreview that suggested using Neutral with all settings at -3... I’m not clear how that would help with ETTR)
WB: Auto (Although I read on DPreview to use UniWB... I don’t even see that as a setting option on the camera... sorry for the newbie question)
Mode: Aperture priority
Technique: I’m adjusting the exposure of a frame until the zebras disappear completely by using the exposure compensation dial (shutter speed adjusts), and then increasing the exposure by 2 stops (again using exposure compensation dial).
Finally, more broadly speaking, for those who are using ETTR, are you using it all of the time or just in certain situations where the lighting range of a shot is particularly high? Appreciate any input here, because I’m currently getting worse results with ETTR than simply disabling zebra and loosely following the on-screen histogram like I’ve done in the past.
p.6 #18 · Easy ETTR technique using the Zebra pattern
I'm normally either a fully AF guy, taking the shot whenever the focus confirmation comes on, or a manual focus guy, shooting only magnified view with no zebras or other obscuring focusing aids.
I find I am at the ragged edge of shutter to freeze motion shooting "normally" with ISO 100-400, which is to say the metered exposure minus around -0.7EV to give me something like 1/60-1/125 sec at the slowest.
If I am following the ETTR technique discussed here, and adding +2.0EV in EC, doesn't that put me at 1/15 sec for hand held?
Maybe I'm just getting confused here, but this "ETTR +2" seems best suited for either tripod mounted or bright sun.
p.6 #19 · Easy ETTR technique using the Zebra pattern
I hear ya. I tried the technique but Lightroom was showing that the highlights were clipped with the blinkies and histogram, and I didn't know if that was accurate or if it's like in camera where it is just showing data from the JPEG, so I stopped doing the ETTR till I could confirm either way.
dukester82 wrote:
Hi - I’ve been reading up on ETTR via this thread and tried it out for the first time this past weekend, and think I’m not doing something quite right as the images are so overexposed that to get the information in some of the highlights (the sky) that in post processing I would have to decrease not just highlights all the way, but decrease exposure by 1.5 to 2 stops and then the rest of the image is way too dark, even after pulling up shadows. Here’s how I’ve set things up - would greatly appreciate any input on what I may be doing wrong:
Camera: A7RII
File format: RAW, compressed
ISO: 100
Zebra: 100+
Creative style : Standard with all settings at 0 (although I read a thread on DPreview that suggested using Neutral with all settings at -3... I’m not clear how that would help with ETTR)
WB: Auto (Although I read on DPreview to use UniWB... I don’t even see that as a setting option on the camera... sorry for the newbie question)
Mode: Aperture priority
Technique: I’m adjusting the exposure of a frame until the zebras disappear completely by using the exposure compensation dial (shutter speed adjusts), and then increasing the exposure by 2 stops (again using exposure compensation dial).
Finally, more broadly speaking, for those who are using ETTR, are you using it all of the time or just in certain situations where the lighting range of a shot is particularly high? Appreciate any input here, because I’m currently getting worse results with ETTR than simply disabling zebra and loosely following the on-screen histogram like I’ve done in the past.
p.6 #20 · Easy ETTR technique using the Zebra pattern
seurot wrote:
I hear ya. I tried the technique but Lightroom was showing that the highlights were clipped with the blinkies and histogram, and I didn't know if that was accurate or if it's like in camera where it is just showing data from the JPEG, so I stopped doing the ETTR till I could confirm either way.
It's called ETTR. Some of your images will be 'intentionally' overexposed but the highlights should still be preserved. There will be a need to edit your images by reducing "highlights" and/or "exposure" in post depending on the scene. The difference is that your shadows will be much cleaner.
