Expose to the right (ETTR) or underexpose?

[farnz]

Advertisement (gone after registration)

... A live histogram combined with focus peaking and the ability to hold the camera up to my eye, when combined with this Hoodman cinema kit, make my NEX extremely practical.

I do the same with my 5D2 but with a Seagull LCD viewfinder (loup) that avoids having to use those irritating straps. A low-profile self-adhesive 'seat' fits around the LCD and the viewfinder can be clipped or unclipped to it at will.

 

Pete.

[Advertisement]

The Leica Q3 is available for pre-order at B&H Photo Video and Adorama

[algrove]

Got very tired of all this back and forth.

 

Histrograms M9 +LR

 

1st Histo; Then 2nd Histo a few seconds later (the MM takes like 30 seconds) and then in the develop Module I get a 3rd and different histogram.

So after getting fed up with 1 and 2, I just bracket now when shooting landscapes.

[douglasf13]

Jaybob, I hear you about practicality. Did you guys see my add on viewfinder for the NEX-5 a couple of years ago? It pretty much works like a Hasselblad magnifier, and folds pretty flat:

 

 

 

link to article (it is missing pics now): http://forums.dpreview.com/forums/post/38124924

 

 

p.s. Jaybob, that boosted exposure of ISO 160 of course won't look like ISO 160, but I'd say it's pretty close to using a higher ISO instead.

Edited October 23, 2012 by douglasf13

[Jaybob]

boosted exposure of ISO 160 of course won't look like ISO 160, but I'd say it's pretty close to using a higher ISO instead.

 

I found the foot photos impossible to tell apart without metadata.

 

All things being considered equal, and at the risk of repeating myself, I'm not really seeing a whole lot of practical benefits to an adoption of an everyday use of underexposing or the "out of camera push" technique. Sure it can be done, there isn't a compelling enough reason to go out of your way to even think about doing it when an in camera ISO switch isn't that difficult, AND the results from base ISO to base ISO pushed 1 or 2 stops are pretty hard to tell apart from each other. You're menu-ing to to 160 anyway, menu-ing and shooting at 500 or 640 (or 800 or 1000) isn't going to tax the processor in you camera or your right thumb all that much. It's good to know that you can utilize it in a situation where you find you've underexposed by accident.

 

I find myself better served setting the lowest possible proper ISO and exposing a little more to the right.

 

Just my opinion.

 

Jay

Edited October 23, 2012 by Jaybob

[douglasf13]

I found the foot photos impossible to tell apart without metadata.

 

All things being considered equal, and at the risk of repeating myself, I'm not really seeing a whole lot of practical benefits to an adoption of an everyday use of underexposing or the "out of camera push" technique. Sure it can be done, there isn't a compelling enough reason to go out of your way to even think about doing it when an in camera ISO switch isn't that difficult, AND the results from base ISO to base ISO pushed 1 or 2 stops are pretty hard to tell apart from each other. You're menu-ing to to 160 anyway, menu-ing and shooting at 500 or 640 (or 800 or 1000) isn't going to tax the processor in you camera or your right thumb all that much. It's good to know that you can utilize it in a situation where you find you've underexposed by accident.

 

I find myself better served setting the lowest possible proper ISO and exposing a little more to the right.

 

Just my opinion.

 

Jay

 

One advantage is that you can leave the camera on one or two ISOs which saves time. The bigger advantage is that it assures that you never blow highlights by raising ISO in-camera. Also, as Bert showed earlier, boosting ISO in post has a little bit of a noise advantage over using ISO 1250.

 

Remember, exposure in digital is how much light is hitting the sensor, and ISO isn't a part of that, so simply boosting ISO to move the histogram right isn't really ETTR. You're just raising ISO to move the histogram right, not changing exposure, and, as we've been showing, that is unnecessary with the M9, outside of a few practical reasons, like needing a usable review on the back of the camera. I shoot the M9 at two ISOs, base and around 640-800. Everything else I've been pushing in the raw converter.

Edited October 24, 2012 by douglasf13

[Bill Allsopp]

A very valid question which prompted me to write a very short article on the subject. ETTR (Expose to the right)

[douglasf13]

Advertisement (gone after registration)

A very valid question which prompted me to write a very short article on the subject. ETTR (Expose to the right)

 

Again, it should be mentioned that ETTR assumes that you're at base ISO, because, if you're changing ISO to move the histogram to the right, you're not actually changing exposure (the "E" of ETTR,) but, rather, you're just boosting the camera gain after the fact. The tests in the last few pages show that is unnecessary to use the M9's gain boost (upping ISO,) at least in terms of IQ.

 

Also, it should also be mentioned that exposing mid tones too far to the right can have consequences on color, which may or may not be a trade off worth making.

 

A third issue is that, if you focus too much on protecting highlights, you may severely underexpose your subject, so sometimes it is worth the trade off to blow out highlights for the sake of the subject.

Edited October 24, 2012 by douglasf13

[Jaybob]

YAWN....

[douglasf13]

YAWN....

 

I'm not sure why this is a worthwhile post in a thread about ETTR?? If you're not interested in discussing the nuance anymore, then no need to join in.

 

I do think this thread has been helpful for some, because there are a lot of misconceptions about how ETTR and camera settings work.

 

ETTR is not about increasing ISO, but, rather, increasing exposure, and it has always been about base ISO only.

[dwbell]

YAWN....

 

Come on Jay. Your contributions have been helpful and much appreciated. Don't ruin it by being that guy now.

 

Douglas's points, together with your and others info have actually increased my understanding. For what it's worth I do agree with Douglas. ETTR kinda only makes sense in terms of IQ if you're changing the amount of light hitting the sensor, through shutter speed or aperture.

[Jaybob]

ETTR is not about increasing ISO, but, rather, increasing exposure, and it has always been about base ISO only.

 

 

A discussion impies that there will be back and forth. I came out swinging, relented, tested, and am not convinced. and yet...

 

If I polled a group of photographers, who were working the World Series, or a presidential event, or an NFL game, or a wedding this weekend, and I and I asked them how they determine exposure, what percentage of them would say "Depending on the ambient light, I set my ISO, white balance, meter something close to a middle gray, and check this histogram to make sure I'm not blown or blocked, and I'm good..." 15%? 30%? 70%? I'm going to say without blinking an eye, CONSERVATIVELY it would be over 95%. That's the way it's done, and the way most photogs brains are wired to process the act of photography, using the camera to make "exposures".

 

How many of them would say "I meter at 640, then set my Base ISO (and white balance if I have time...), and shoot using the 640 setting, effectively underexposing 2 stops, dont bother with the histogram and I'm good...oh and then I'll drag it up in Lightroom later, with absolutely no benefit to the image quality." I dont think I'm going out on a limb by saying the percentage would be really low. It's just not practical. WHY is the question I keep asking myself. Not HOW does this work, but (assuming that "image quality" is not "improved", whatever that means) WHY would I use this method?

 

That being said, I, (and I'm not the only one) consider my ISO dial to be a functioning, helpful part of my camera. Like choosing a film stock, TMax 100 looks different than Tri-X exposed at 1250 or 1600. It's always been a conscious personal choice. I understand that "ISO" is just gain amplification, amplification that can result in noisy-ness and artifacts and whatnot. I don't think it's a bad thing follow the accepted conventions of the practice of photography by setting an effective film speed and then adjust my aperture and shutter speed for convienice and effect. Dont get me wrong, too much "camera induced noise" is too much, and, for me, that occurs around 2500 on my D3. For me, it certainly doesn't happen at 640 or 800 on a properly exposed M9 file. The practical use of ETTR doesn't assume that you're at base ISO, the concept works perfectly fine when set at ANY ISO. ETTR as a concept actually works better when the camera is set on a higher ISO because, you're protecting shadow and black information, information that is already the first to feel and show the stretch.

 

ETTR, or exposing for the highlights in digital regardless of model, is hardly news from bizarro land, or some radical concept, for alot of photographers it's the way pictures get made in the real world. I'm not convinced that shooting at base but underexposing and fixing it in post is a better way to go about it. If I was seeing noticeable better image quality (I'm not), I'd be on board. Shot at any ISO, ETTR (with no blown highlights) or as close to proper exposure, will result in more controllable and fixable images. When faced with any histogram there's exponentially more information to correct in the higher registers (exposure, highlights, and whites) of the histogram than there is in the lower areas of exposure, shadows and blacks. If that is accepted, why would I want to start the correction with all of my information bunched up to the left of center?

 

I am beyond frustrated with this "discussion", hence the yawn. Again, a discussion impies that there will be a back and forth, not just "assumptions" and the same patronizing comments made to anyone who strides into the room with a valid point to make. It's not personal. Just an observation of what's being typed.

 

I'm a big boy and I realize no one's "making" me use this underexpose method. I'm just not a fan of it, unless, of course, it's to fix a mistake.

 

Jay

Edited October 25, 2012 by Jaybob

[dwbell]

That's a lot of text for someone who's bored with the non-discussion Jay!

 

Seriously, I tend to agree with all your points. However, the discussion was helpful to me. The points Douglas raised, even if pedantic, or semantic to some, I found useful to crystallise my knowledge on ETTR. Even if the weakest, most lame-arsed take away of this to me is "hmmm, I could bump the ISO here to get the histogram sensibly 'right' but that won't affect IQ in any way. What I need to do is let more light in, which means a tripod" then, for me, it's worth a re-discussion of the issue.

 

Peace guys, learnt from both of you.

[Jaybob]

it's not boredom, trust me.

 

it's exhaustion from the same point being repeated ad nauseam, a point that has been delivered in thinly veiled yet still somehow completely condescending manner.

 

What are the real benefits of underexposing, other than just being able to do it and having it all turning out fine in post production (or giving you exactly the same results as adjusting an apparently ficticious thing called "ISO" on camera)?

Edited October 25, 2012 by Jaybob

[k-hawinkler]

You got it!

 

Here a three way comparison with the M9. The standard deviation of the same dark area:

 

ISO 1250 uncompressed stdev 7.51

ISO 160 + 3stops uncompressed stdev 6.02

ISO 160 + 3stops compressed stdev 6.74

 

Exposure time of all three: 0.5 sec. at f/4 with the 75/1.4

 

Finally we find a difference in compressed and uncompressed! But only if you push 160 ISO by 3 stops. Still the 160 ISO is less noisy than the equally exposed 1250 ISO and uncompressed by quite a margin.

 

Tiff file at 100% of original, standard setting. LightRoom 4.1, process 2012

 

 

 

I think this is worth keeping in mind.

[Jaybob]

I think this is worth keeping in mind.

 

No, I'm still not convinced.

 

It's not a real world situation. Playing the Devil's advocate, If I was faced with that tripodded photograph to take, and "image quality" was of the utmost consideration, I don't know one photographer who would ever expose those frames in that way. Why not just go to 4 or 5 seconds at f4 at 160, uncompressed of course? You're on a tripod already. You have the instant feedback of a histogram. Why not just shoot it properly since you're at the base ISO as opposed to fixing it later? Shooting it deliberately wrong and then being able to (and having to) fix it in the machine isn't interesting to me.

 

The scientific methodology seems as though it may be suspect as well, (perhaps it isn't) there are external influences such as the default LR import settings, in camera white balance, sharpening, lens corrections, color noise reduction, not to mention starting with a 3 stop underexposed image?

 

Discussion, not dismissing.

Edited October 25, 2012 by Jaybob

[k-hawinkler]

The numbers speak for themselves.

[Lindolfi]

It may be usefull to repeat the experiment yourself Jaybob. It is easy. Just keep all conditions the same, including settings of the RAW converter, with the exception of the Exposure slide. Open in photoshop en look at the histogram of a dark part of the image. Observe how the noise is higher in the high ISO photograph and see whether it works. Draw your own conclusions.

 

The fact that a tripod is used is just to keep the conditions equal. It is necessary for the experiment. A real world photograph would also benefit fromt the same principle without a tripod.

 

How important is it for a particular photograph you want to make? All depends on what you require. But it may make just the difference that you need.

 

I should indicate to Jaybob and Douglasf13 that both have good arguments, but they are used in a different context and can therefore never lead to agreement: Jaybob claims that most photographers don't work with underexposure and therefore it must be wrong (context of convention) while Douglasf13 points at a principle that may be worth looking into (context of discovery).

Edited October 25, 2012 by Lindolfi

[dwbell]

I should indicate to Jaybob and Douglasf13 that both have good arguments, but they are used in a different context and can therefore never lead to agreement: Jaybob claims that most photographers don't work with underexposure and therefore it must be wrong (context of convention) while Douglasf13 points at a principle that may be worth looking into (context of discovery).

 

Bingo. You're talking past each other. Also I don't find Douglas' tone condescending. Perhaps it's an interpretation the written word problem so common on the Internet?

[Advertisement]

Hallo,

Look here: Leica M

[Jaybob]

It may be usefull to repeat the experiment yourself Jaybob. It is easy. Just keep all conditions the same, including settings of the RAW converter, with the exception of the Exposure slide. Open in photoshop en look at the histogram of a dark part of the image. Observe how the noise is higher in the high ISO photograph and see whether it works. Draw your own conclusions.

 

The fact that a tripod is used is just to keep the conditions equal. It is necessary for the experiment. A real world photograph would also benefit fromt the same principle without a tripod.

 

How important is it for a particular photograph you want to make? All depends on what you require. But it may make just the difference that you need.

 

 

I tested already. I understand the concept. I just fail to see real world application.

http://www.l-camera-forum.com/leica-forum/2214324-post168.html

[k-hawinkler]

Lack of imagination, maybe?