Home · Register · Join Upload & Sell

Moderated by: Fred Miranda
Username  

  New fredmiranda.com Mobile Site
  New Feature: SMS Notification alert
  New Feature: Buy & Sell Watchlist
  

FM Forums | Canon Forum | Join Upload & Sell

1
       2       end
  

Archive 2013 · Exposure Latitude Inconsistency - Why?

  
 
CJBoghosian
Offline

Upload & Sell: Off
p.1 #1 · p.1 #1 · Exposure Latitude Inconsistency - Why?


Hey, Folks!

My name is Christopher J. Boghosian - this is my first post. I'm looking forward to being a part of this vibrant community.

I'm an experienced photographer, though a novice with digital. So when I recently got my 60D, I tested exposures off a gray card and am confused:

After importing my images to Lightroom, the Lightroom histogram showed that my exposures were okay 5 stops below normal exposure, BUT ONLY 2 STOPS OVER. In other words, the highlights went off the histogram only 3 stops over normal exposure of the gray card whereas the underexposure shots were still on the histogram up to 5 stops under.

I was assuming a regular range of latitude up and down the range of exposure. Why such big jumps up the histogram when over exposing?

Also, can someone please explain this statement from Lowrie's "Canon EOS 60D Digital Field Guide" (2011): "The general guideline when shooting RAW capture is to expose with a bias to the right so that the highlight pixels just touch the right side of the histogram. Thus, when tonal mapping is applied during conversion, the file has significantly more bits that can be redistributed to the midtones and darker tones, where the human eye is most sensitive to changes."

I'm thinking Lowrie's statement might be connected to my initial question regarding the inconsistency in exposure latitude?

Thank you in advance for your consideration!

- Christopher



Jul 09, 2013 at 05:53 PM
jcolwell
Offline
• • • • • • •
Upload & Sell: On
p.1 #2 · p.1 #2 · Exposure Latitude Inconsistency - Why?


Hi Christopher,

Welcome to FM.

It sounds like you're concerned about a bias in exposure latitude, not consistency. There's a whole lot of learning to be done when going from film to digital. I'm still working on it.

With respect to your results with LR, are you looking at RAW files or jpeg files?

Cheers, Jim



Jul 09, 2013 at 06:25 PM
Guari
Offline
• • •
Upload & Sell: Off
p.1 #3 · p.1 #3 · Exposure Latitude Inconsistency - Why?


Hi Chris, nice to meet you..

You should have a mindset when exposing digital files as if you were shooting E-6, not C41.

You have to protect that the highlights don't white out during exposure, while letting the shadows fall where they will. Then, with lightroom, you have to develop the shadows, bringing them back to light (or alternatively, darkening) via your sliders.

Once you loose a highlight, there's little headroom for recovery. Fr this reason, it is good practice to enable the "blinkies" in your image preview as you will have a good idea if you need to backoff the exposure a notch.

All of this assumes that you are shooting RAW files and not jpg's.

The reason why latitude is not symmetrical (as in the zone system) has to do with the fact that digital sensors gather light in discreet " tonal steps" that are controlled by the "bit depth" of your file. RAW files have more steps, but you have to develop the files in lightroom in order to make good use of all that "latent info", which lives in the shadows. Expose as if they were high latitude diapos, protectong the highlights from blowing.

If this sounds too complex, it's no worries. Shoot raw, develop in lightroom (or even photoshop) and be happy..

Hope that helps



Jul 09, 2013 at 06:51 PM
Luta13
Offline
• •
Upload & Sell: Off
p.1 #4 · p.1 #4 · Exposure Latitude Inconsistency - Why?


I have no idea what anyone is talking about here



Jul 09, 2013 at 07:28 PM
CJBoghosian
Offline

Upload & Sell: Off
p.1 #5 · p.1 #5 · Exposure Latitude Inconsistency - Why?


Thank you for the kind welcome! And very useful information - thank you...

I am referring to RAW files - I understand that JPEG is a whole other issue due to in-camera processing...

In response to your insight:

- I have enabled the "blinkies" in my camera preview; however, even when shooting RAW, am I not ultimately previewing an in-camera JPEG? I guess I'm ultimately questioning the validity of an in-camera JPEG preview when shooting RAW.

- When shooting a high contrast scene (without fill-flash; no gray card), how do digital photographers GENERALLY determine exposure? Based on your comments, it sounds like the highlight is of utmost concern, thus the beginning point. If so, does one measure the highlight and give it a couple extra stops of light, allowing the shadows to fall where they may?

Thanks!



Jul 09, 2013 at 07:33 PM
snapsy
Offline
• • • • • •
Upload & Sell: On
p.1 #6 · p.1 #6 · Exposure Latitude Inconsistency - Why?


Camera meters are designed meter middle gray around 2 1/2 stops below highlight clipping. It does this by assuming the content in the scene has average reflectivity; if the content is more reflective than normal (like snow) then the camera may underexpose; if the content is less reflective than normal (like black) then the camera may overexpose. Evaluative/matrix metering also applies scene recognition to help predict what reflectivity the scene has and also what aspects of the scene contains content important to meter from. It also can average meter regions across the frame.

Even when the camera meters correctly the in-camera histogram/blinky display is based on a raw-converted (post-gamma) conversion applied to the raw using the user-selected JPEG scene settings (contrast, saturation, white balance, etc..), even when shooting raw. This may lead to an indication of highlight clipping where there is none or more rarely, vice versa...such that when you load the raw into your PP app like Lightroom there will be no clipping of raw data for an image the camera indicated there was clipping. Do a google search for "UniWB" for more information.



Jul 09, 2013 at 08:01 PM
StillFingerz
Offline
• • • •
Upload & Sell: Off
p.1 #7 · p.1 #7 · Exposure Latitude Inconsistency - Why?


Hi Christopher and welcome to FM!

Here is a good article discussing the ETTR 'Expose to the Right' technique.

Maximizing S/N Ratio in Digital Photography
http://www.luminous-landscape.com/tutorials/expose-right.shtml

There is a ton of information on the web, just put the above 'text' into Google search and you will get quite a few hits. Luminous Landscape is another wonderful website and if you want to fully understand 'RAW' data I can highly recommend the folllowing book that details RAW files.

The Digital Negative: Raw Image Processing in Lightroom, Camera Raw, and Photoshop
By Jeff Schewe

Here's a link to the book, it's on Amazon...in paperback or eBook formats.
http://www.amazon.com/Digital-Negative-Processing-Lightroom-Photoshop/dp/0321839579/ref=sr_1_1_title_0_main?s=books&ie=UTF8&qid=1373418937&sr=1-1

Hope this helps, good luck!
Jerry



Jul 09, 2013 at 08:18 PM
CJBoghosian
Offline

Upload & Sell: Off
p.1 #8 · p.1 #8 · Exposure Latitude Inconsistency - Why?


Thank you all again! "Exposing to the right" is exactly what I'm talking, so now I know what to search for.

I quickly glances at the luminous landscape link (will read tomorrow) and it said that a digital sensor has about 5-6 stops latitude. That's exactly what I discovered with my gray card test!

LOOKS LIKE I GOT A LOT OF READING TO DO!!!

This thread has really helped me put terms to what I was dealing with, so now I'm able to do my own research. Thanks again..



Jul 09, 2013 at 08:25 PM
CJBoghosian
Offline

Upload & Sell: Off
p.1 #9 · p.1 #9 · Exposure Latitude Inconsistency - Why?


WHOA!!!! I absolutely cannot believe how TRUE ETTR is and it's a bit unnerving.... Coming from a film background, I'm shocked at how "abnormal" the analog sensor metering method is.

I did a quick test of a dark scene:

1) Exposed it using a gray card.

2) Exposed it +5 stops (to the extreme right of the histogram).

Opened both in Lightroom. I didn't touch the first exposure (normally exposed). I brought down the exposure of the second.

Alas, the first, "normally exposed" was incredibly more noisy! The second, "overexposed" by 5 stops looked fantastic.

****Note: I cannot post photos

Thank you so much for all the help. I'm both elated to have learned so much, but really annoyed that I need to overexpose (not underexpose) dark scenes.



Jul 09, 2013 at 09:53 PM
jimmy462
Online
• • •
Upload & Sell: On
p.1 #10 · p.1 #10 · Exposure Latitude Inconsistency - Why?


CJBoghosian wrote:
...I recently got my 60D, I tested exposures off a gray card and am confused:

After importing my images to Lightroom, the Lightroom histogram showed that my exposures were okay 5 stops below normal exposure, BUT ONLY 2 STOPS OVER. In other words, the highlights went off the histogram only 3 stops over normal exposure of the gray card whereas the underexposure shots were still on the histogram up to 5 stops under.

I was assuming a regular range of latitude up and down the range of exposure. Why such big jumps up the histogram when over exposing?

...

- Christopher


Hi Christopher, and welcome,

I am curious...in your grey card experiment, how did you meter your exposures for those shots with your camera? What I'm not understanding is the meaning behind your usage of "normal exposure" and "underexposure" here. And, what do you mean, or how would you define (describe) "regular range of latitude" here? I'm just trying to establish a common reference point from where we can explore what it is you're encountering.

The transition from film to digital for me came with some frustrations and confusions...as an example: I did not (and still do not) like the unforgiveness of digital when it came to highlight retention. Once one reaches the full-well capacity of their sensor information gathering gets abruptly cut-off and all of one's highlights beyond that point are gone...film emulsions work on a much smoother exposure latitude curve with no abrupt (I prefer: rude) cutoffs.

And I bring that particular example up as it may have some bearing on what you are experiencing. Learning the exposure behaviors—its benefits and idiosyncrasies—of the sensor in your 60D will be akin to learning how a new emulsion behaves for you. But digital is not film, it is merely an electronic simulation (recreation), thereof. A lot of minds and effort went into the sensors, electronics and software allowing you to take that 50mm f/1.4 lens off of your film camera and slap it on your digital body so that when you set your exposures identically on both cameras the digital one would yield results similar to what the film one had been giving you all along!

Anyhoo, here are a couple of interesting shootouts between digital sensor and film behaviors in both over- and under-exposed situations...

Twin Lens Life ~ Fine Art Film Photography ~ Los Angeles Southern California ~ Bwright Photography: Digital vs. Film (The Real Deal) - Nikon D300 vs. Fuji GS645s:
http://www.twinlenslife.com/2009/05/digital-vs-film-real-deal-nikon-d300-vs.html

Twin Lens Life ~ Fine Art Film Photography ~ Los Angeles Southern California ~ Bwright Photography: Digital vs. Film - Canon 5D Mark II vs. Kodak Ektar 35mm "Pound for Pound, Pixel for Pixel":
http://www.twinlenslife.com/2011/01/digital-vs-film-canon-5d-mark-ii-vs.html

...and welcome to the world of digital image making. Things are not always as they appear.

Best to you,
Jimmy G



Jul 09, 2013 at 11:10 PM
scalesusa
Offline
• • • •
Upload & Sell: Off
p.1 #11 · p.1 #11 · Exposure Latitude Inconsistency - Why?


Your premise is that Camera makers calibrate exposure in their cameras to a 18% gray card.

They also do not calibrate to reflectance, so there are two incorrect assumptions.

http://www.bythom.com/graycards.htm




Jul 09, 2013 at 11:29 PM
gdanmitchell
Offline
• • • • • • •
Upload & Sell: Off
p.1 #12 · p.1 #12 · Exposure Latitude Inconsistency - Why?


Think of "expose to the right" (a guide and not a rule) as meaning "expose close to the right but (almost) never too far to the right."

You often want to expose to the right - e.g. keep things as bright as you appropriately can - in order to avoid two problems in shadows of scenes with very large dynamic ranges, both of which can be challenging if you need to bring up the dark areas in post: noise in shadows, and banding (not noise banding) if you expand the shadows. (The former is because the signal level ends up not far enough above the noise floor and the latter because fewer data levels are available for recording the darkest portion of the image.)

However, you (generally) do not want to go too far in pushing the highlights since you risk "blowing out" the brightest areas of the scene. Here digital behaves quite differently than film. With film, (esp. negative film) the response to light "rolls off" at the brightest levels, allowing you to push those highlights a bit and still get some detail. But with digital, the failure is abrupt - once the level in a color channel saturates, no detail will be recorded. If you saturate all three channels, basically you just get plain white. (We've all see photographs by people who missed this - for example, beautiful textured clouds with gross pure white bright areas that have no detail left.)

So, yes, you have fewer stops to work with above the supposed optimal exposure in digital. While you may be able to get away with blowing a few small specular highlights, you almost never want to allow an area of the scene with important texture, etc. to blow. Further, you need to watch carefully that you don't blow an individual channel. (The classic error is with a subject that is very hot in one channel, usually the red channel. The red and green channels can be fine, as can be the average of the three channels, but the red channel blows badly. See many overexposed flower photos for examples. )

Sometimes you don't want to expose all the way to the right. A very low contrast scene with a lot of interesting detail in the regions of near-white tonality can sometimes give you a bit more to work with in post if you hold back from the right just a bit.

Dan



Jul 10, 2013 at 02:59 AM
Ferrophot
Offline
• • •
Upload & Sell: Off
p.1 #13 · p.1 #13 · Exposure Latitude Inconsistency - Why?


Although I've been photographing since 1960 I still think nailing exposure is harder than nailing focus. Not sure all the fancy algorithms and weightings make the task easier either. Normal scenes no woories, get a vast tonal range and it gets much harder. Agree about the blinkies, but then I only see them after the shot is taken, that's too late sometimes. I generally take a test shot, check historgram, switch to M, adjust exposure, take another, check etc until I'm sure that the white objects are not blown. Yes, I can sit on the shutter and take 6 identical shots, there's always one that's underexposed.


Jul 10, 2013 at 08:14 AM
gdanmitchell
Offline
• • • • • • •
Upload & Sell: Off
p.1 #14 · p.1 #14 · Exposure Latitude Inconsistency - Why?


About not knowing until after the first shot - just think of your camera and its histogram display as a very fancy and expensive meter. Go ahead and use that first shot to zero in on the best exposure. No embarrassment in doing that!

Dan



Jul 10, 2013 at 09:55 AM
CJBoghosian
Offline

Upload & Sell: Off
p.1 #15 · p.1 #15 · Exposure Latitude Inconsistency - Why?


JIMMY: Thank you for the links - I will check them out! In regards to your questions:

- I used my camera spot meter pointed at an evenly lit gray card, which completely filled the frame.

- I used the meter to determine "normal exposure," i.e., set at 0 on the Exposure Level Indicator.

- I then took a series of photos of the same gray card in the same lighting in 1-stop increments, 5 down and 5 up.

This is a link to my series of photos: A screengrab of my series of exposures.

What confused me was that ALL 5 down (underexposed) stayed within the histogram; however, only 2 up (overexposed) stayed within the histogram. The apparent imbalance in the range of latitude really confuses me. Why 5 down the range but only 2 up?

I suspect it has something to do with the way the sensor works, i.e., it assigns half the data to the brightest highlight in the scene.

SCALESUSA: thank you for the link. Very helpful. I too noticed that my 18% gray card exposure was 1/2 stop to the left! I was so confused; now I know why - thanks....

DAN: thank you so much for the helpful info, especially the point about the red channel - very interesting. But I still have this one question: if the in-camera histogram is based on a camera generated JPEG, then how much discrepancy is there between that JPEG reading and the actual RAW image I ultimate care about? In other words, I'm not getting a histogram of my RAW image, but a camera generated JPEG, so how accurate is that?

Thank you all!!!



Jul 10, 2013 at 10:41 AM
Guari
Offline
• • •
Upload & Sell: Off
p.1 #16 · p.1 #16 · Exposure Latitude Inconsistency - Why?


The in-camera preview (Jpg) is good enough for judging when clipping occurs (blinkies)... As long as you don't see blinks, you are good to go


Jul 10, 2013 at 11:03 AM
goosemang
Offline
• • •
Upload & Sell: Off
p.1 #17 · p.1 #17 · Exposure Latitude Inconsistency - Why?


i usually let things clip a little on the in-camera histogram, because i know i can recover a certain amount in camera raw. as for how much you can recover... depends on the scene, the camera, etc. you just have to get to know how the files from your particular body react.


Jul 10, 2013 at 11:58 AM
jimmy462
Online
• • •
Upload & Sell: On
p.1 #18 · p.1 #18 · Exposure Latitude Inconsistency - Why?


CJBoghosian wrote:
JIMMY: Thank you for the links - I will check them out! In regards to your questions:

- I used my camera spot meter pointed at an evenly lit gray card, which completely filled the frame.

- I used the meter to determine "normal exposure," i.e., set at 0 on the Exposure Level Indicator.

- I then took a series of photos of the same gray card in the same lighting in 1-stop increments, 5 down and 5 up.

This is a link to my series of photos: A screengrab of my series of exposures.

What confused me was that ALL 5 down (underexposed) stayed
...Show more

Hi again, Christopher,

Okay, thanks for the clarifications! Here goes with some more homework for you...

The problem lies not with your camera but with our understanding with what our camera's histogram is actually telling us. Most folks (and a good portion of the internet) have this belief that 18% grey is middle grey to our cameras, or value 128 (on a scale from 0-255). Such is not the case. It turns out that our cameras use an sRGB-adjusted gamma for their histograms, not a linear gamma of 1. As a result the bit value for middle grey is more like 117 in sRGB. There's an excellent (read: understandable) write-up on this under the Simplest test to show obvious truth - Histograms show gamma data header a bit more than half-way down this page...

Histograms show gamma-encoded data:
http://www.scantips.com/lights/gamma.html

...explaining why you're seeing 2 1/2 stops up and 5-stops down. Your experiment, as it turns out, recreates the author's experiment and, not surprisingly, corroborates his findings!

Another page that utilizes this experiment, but where the author both struggles for but never figures out the math behind the effect is here...

Digital Tones and Exposure Zones:
http://www.rags-int-inc.com/PhotoTechStuff/TonesnZones/

And, another, still, under the Calibrating Your Brain to Understand the Histogram header, with no attempt at understanding or explaining the "why it is"...

Interpreting the Camera Histogram:
http://super.nova.org/DPR/Histogram/

FWIW, you can actually view the RAW images, and their RAW histograms, in Linear Gamma from your 60D using Canon's DPP software.

Hopefully, again, this is of some help!
Jimmy G



Jul 10, 2013 at 11:49 PM
gdanmitchell
Offline
• • • • • • •
Upload & Sell: Off
p.1 #19 · p.1 #19 · Exposure Latitude Inconsistency - Why?


CJBoghosian wrote:
DAN: thank you so much for the helpful info, especially the point about the red channel - very interesting. But I still have this one question: if the in-camera histogram is based on a camera generated JPEG, then how much discrepancy is there between that JPEG reading and the actual RAW image I ultimate care about? In other words, I'm not getting a histogram of my RAW image, but a camera generated JPEG, so how accurate is that?!


You bring up an interesting question. I could answer several ways. One would be to go look up which picture setting is the one that supposedly provides the most accurate histogram display since, as you mention, it is based on (or at least used to be based on) a jpg version of the image.

But... in reality the histogram image, no matter how "accurate" it is, must still be interpreted by the photographer. The red channel issue is a good example. Here is a little list of issues in this particular situation:

1. Some don't realize that they can display a multiple channel histogram that separately displays the red, green, and blue channel histograms. The combined ("white") histogram display averages the RGB values. This works fine for many sorts of scenes, but in those that are strong in one channel but weak in the others, that strong channel can be over-exposed with no indication from the white histogram, since it averages that over-exposed channel's value with the less intense values of the other two channels.

2. With this in mind, at least with subjects that are very strong in one of the color channels (but why not all the time?), it is better to use the three channel display and watch for the strongest channel to determine your ETTL ("expose to the right") based exposure settings.

3. However, this isn't enough in all cases. There are certain subjects in which you might still need to drop your exposure by a stop or so. Here I commit the sacrilege of actually looking at the image in the LCD, where I have learned to recognize how it responds to near-blown or blown channels by changing the image color or saturation in those areas. This is not an objective measure nor an accurate view, but it tells me that I'm pushing it in situations where the camera says I'm not. Another situation - which I think I mentioned in my earlier post - is when the image is very high key and I want to retain sufficient detail in the near white tones. Here I will not push all the way to the right, especially if there are few or no dark tones in the scene. I find that I'll have better luck working those high key tones in post if I "underexpose" a bit from the supposed ideal ETTR setting.

Among the many themes that I would like to see more forum photographers understand is that there is no such thing as a camera-determined ideal or perfect exposure. There are many ways to decide to expose a thing, and we still must use our experience and exercise judgment (and occasionally make an educated guess) about things. Some, especially new photographers and double-especially new "digital" photographers hope that the camera will make the right decisions in all cases. This simply is not the case.

Take care,

Dan



Jul 11, 2013 at 04:40 AM
CJBoghosian
Offline

Upload & Sell: Off
p.1 #20 · p.1 #20 · Exposure Latitude Inconsistency - Why?


Thank you again, Gentlemen.

I checked out the articles you linked to, Jimmy, and I have to say they got a bit too technical for me! However, I did grasp the general ideas and am finding more and more that Dan's overall position is the way to go: personal experience and discernment.

However, I am very happy that I am gaining a better sense on how digital exposure works. I cannot approach it in the same what as I did with film - that's certain.

I'm definitely going to give more attention to the three channel histogram in-camera. Dan's warning regarding scenes that are monochromatically heavy is great to know and something to watch out for!




Jul 11, 2013 at 12:31 PM
1
       2       end




FM Forums | Canon Forum | Join Upload & Sell

1
       2       end
    
 

You are not logged in. Login or Register

Username       Or Reset password



This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.