Olympus E-M5: Exposing (to the right) and Lightroom 4.1

I posted previously an article on "Exposing to the Right" (ETTR) with Olympus E-P3 RAW images. You can read it here. This blog is based on the same fundamental idea. My previous article was both praised and scorned. The latter came mostly from people who believe in a method to do ETTR known as UniWB. UniWB is a clever method of addressing limitations you encounter with practically every digital camera if you want to expose to the right. With it you have to endure either a green or a black and white live view viewfinder, and I am not too happy with either. Also it is based on using histogram which I´m not happy with either. This method I am using is based on using highlight warning blinkies as exposure meter: they are both easy to see, very graphical, and show what´s actually happening in your image, spot by spot if you want to. This method applies for Olympus PEN cameras since E-P3 (plus E-PL3 and E-PM1) and OM-D E-M5.

Since I got Olympus E-M5 I have practiced it quite happily with the same settings as with E-P3. Lots of people have written to me to write about ETTR with E-M5. I wanted to wait until Lightroom is ready for it because this method is set to be used with Lightroom. Of course you can tweak it for other RAW converters by varying camera configuration settings. At first Lightroom 4.0 was a pile of bugs. There were two Release Candidates for Lightroom 4.1, and now we have a quite satisfactory final version of Lightroom 4.1. It still has some instability issues left (and a few shortcomings with Photoshop integration if you have not updated to Photoshop CS6) but they are no more important here. What´s more important is that Lightroom 4.1 gives you more control on your image relative to its final use.

E-M5 configuration

Lets start with Olympus E-M5 configured like this:

Shooting Menu 1 > Picture Mode > 4 Muted > Contrast -2, Saturation, -2, gradation: Normal

Custom Menu D > Info Settings > LV-info > Highlight&Shadow:On, Histogram: Off

Custom Menu D > Histogram Settings > Highlight: 245

Custom Menu D > Live View Boost: Off

Custom Menu G > Color Space: Adobe RGB

Auto white balance, if used, must be set to neutral: Custom Menu G > Auto WB, Keep Warm Color: Off

Exposure

I adjust my exposure until the red warning color just appears on my chosen critical tone area. By critical tone I mean the brightest area in my image where I want to have definition of tones if the image is printed. In principle, this allows a 1/6-stop accuracy at best, but in practice, the reference tone is very close to the value of 96% almost always when the image opens in Lightroom Develop module. Same critical area will be at RGB value 245 (Adobe RGB or sRGB) in the final image with very little need to adjust as seen below. I do not need to use histogram while shooting, I do not need to check captured images, at least not because of the exposure. 

I exposed this target (right half of Spyder Checkr) with E-M5 so that red highlight warning color started to blink on the lightest area, which is the white patch, up left. This how it looks like imported in Lightroom 4.1 with Lightroom default settings. The only thing I have changed is that I chose Custom WB according to patch shown by cursor. I shot the target outside in sunlight using Daylight WB. You can use Auto-WB for general subjects but it is better to choose a preset WB or make in-camera Custom WB if your image has one or two dominant colors. 

(Note: What you see above is a sRGB image. That´s the only practical option to show in the internet. If you download it and measure patches, they have sRGB values and differ from what I show as measured from converted RAW file below.)

The image above shows the lightest and darkest patch measured (Custom WB, other LR settings at default). The lightest patch is very close to my aim, 96%. The image looks mostly fine but you may need to make some adjustments to darker tones if your subject had low or extremely high contrast.

In general photography subjects many times have brighter spots or areas than my critical value. They are mostly light fixtures and specular reflections where tonal definition is not needed. They may and must go purely white otherwise the image will look flat. I´m interested in where I want tones to end.   

Basically this was it. I am done with telling how I expose with E-M5 when images are opened in Lightroom 4.1. But I go on and write some more background information which might be useful to some people.

Quick-checking Color Channels

If you need to be more precise about colors everything is not done yet. Namely, some highly saturated colors may have a blocked channel as seen in Lightroom RGB preview. In this target all the other patches have definition but yellow patch shows red channel as over saturated and blue patch shows the same channel as zero with LR default settings.

Note: If shooting casual images to be seen just on monitor, I could expose even more to the right. There is some headroom left. By setting Custom Menu D > Histogram Settings > Highlight: 255 I would expose lighter by 1/3 of a stop. And even from there you could expose still a full stop more before the lightest patch would become hopelessly over saturated and blocked. I have tried it but the risk of saturating color channels beyond repair becomes too big AND the time needed to tweak the image colors grows too long. That´s why I have chosen my approach with Lightroom 4.1 as shown here. You may find Histogram Setting 255 worth trying. 

Here I have corrected those two patches to have definition in every channel by slightly lowering yellow and blue saturation in HSL panel.

Note: Actually I can use Spyder Checkr to calibrate my camera to show the colors in target correctly, but that would be a subject for another blog.

RGB Profiles and Soft Proofing

This image shows the same situation seen as Adobe RGB Soft Proof in Lightroom 4.1. Soft Proofing is a new feature in Lightroom 4. It gives you a tool to check that your image has sound values for a needed use. If your image will be printed in a magazine (or offset printing), the general workflow is to prepare and send it in AdobeRGB color space before color separation (conversion into the right flavor of CMYK). With my exposure the lightest patch goes just where the tonal definition end generally in offset printing, at RGB values of 245. Depending on paper the darkest tonal definition would fall between RGB values 16 to 30, with 16 meaning high quality glossy magazine cover and RGB 30 meaning a cheap matte cataloque paper. Instead of Histogram it says now Soft Proofing and the histogram has been changed to show values according to Adobe RGB profile. Below histogram you can see now RGB values, these here are for the lightest patch. I have copied the corresponding RGB values on each four color pick window. Gamut warning is on, as shown by the little square in the upper right hand corner of histogram. Blue patch has still some traces of warning color showing that it is just barely inside Adobe RGB gamut. RGB numbers all show definition in every RGB channel (maybe just barely but still) and they would print out nicely enough with CMYK colors. There´s more to this (Intent etc.), but what´s important, you can see how easily this method of exposing with E-M5 gives you technically very good images. 

Here we are soft proofing the same RAW file with sRGB profile (LR at default settings, same custom WB). The lightest patch has the same values (244) as with AdobeRGB profile but the darkest patch is darker. The main difference, however, is the smaller color gamut of sRGB. Five patches are covered by red gamut warning color. 

Here I have tweaked those five patches into inside sRGB gamut by using saturation sliders in HSL panel. Again: Exposure is fine but RAW image makes a lot more colors possible than can be had inside sRGB color space. This kind of mapping is also what happens if you shoot JPEG and choose sRGB as color space in your camera. Your camera maps colors inside sRGB color space. The method may be different from what I have shown here by simple means but the idea is the same.

Note: For casual images shown on your monitor there is no need to force all colors inside sRGB (or AdobeRGB, if your monitor is capable of it). You are free to have over saturated colors if you like them, but for many professional purposes colors need to be inside gamut. Also all cameras do not behave right and map colors inside chosen color space for JPEG images. Bleak inside gamut colors are not pretty enough, it is commercially better to leave colors saturated, strong and beautiful.

The last example comes back to my reasons to do things my way. My favorite printing paper for my pigment ink printers is Canson Infinity Platine Fibre Rag. Here we have again the same target shot with Lightroom at default values (same custom WB). All the colors are beautifully inside Platine Fibre Rag gamut with this profile for Epson 9900 printer. What coud be better? Perfect exposure for the lightest tones and even strong colors have full definition out of the box. I only need to tweak dark tones as needed. Most of the time I do a lot more with my images but that´s not out of necessity. 

-p- 

Update on June 22, 2012 at 15:35 by Pekka Potka

Two examples of lots of gain with no pain

This morning, while walking the dog, I shot a couple of comparisons to show the difference in details and noise between images that are exposed "normally" with E-M5 and "my" images. "My" meaning of course using settings and blinkies as said earlier in this blog.

Example one:

Same view shot first at camera´s (E-M5) autoexposure and secondly so that red blinkies just lit the trailer´s bright side. Both shots opened here into Lightroom 4.1 with Lightroom´s default settings. Readings under histogram are from trailer´s white side.

Next I clicked Auto Tone for both images. As you can see, Lightroom gives almost exactly one stop difference between needed adjustments. Again readings from the same spot from white area. I could have gone further here with tweaking both images but doing so has no real relevance here. Like you also can see white area is not really white but has details like over painted markings. These highlight details would be quite similar after any reasonable tweaks in both images. Looking at other things in these images, you notice that I have applied an overly strong but equal combination of sharpening and no noise reduction for both images. This to illustrate how ETTR image is clearly superior in blue sky and in yellow and green surfaces. It has a lot more headroom for tweaking. These are 100% crops.

Crops from upper right hand corner show how darker but still medium blue sky behaves. Same overly strong sharpening to better show the differencies. ETTR image keeps wisps of cloud much better intact.

Example two:

Again the other picture shot with camera´s autoexposure but this time the other on was exposed to the right up to the last setting where red blinkies did not light up the lightes part of sky. Both images here opened into Lightroom 4.1 with Lightroom´s default settings. This example is to show the behaviour of darker tones.

And again, I just clicked Auto Tone to normalize images. These details from the back of chair are 2:1 screen captures from Lightroom. All noise reduction (color and luminance) is again off (at 0) and strong sharpening applied. As you can see ETTR image shows much more detail and less especially color noise. What you would get is even more detail with similar noise reduction in every image. The more you tweak the greater the gain from ETTR would be. Also note how much more details with smoother tones the background (easiest seen on the right) shows.

-p-  

Comments

[Don Pope]

This is one of the things I love the most about the EM-5. The ability to see the extreme highlight and shadow areas in the viewfinder. I can adjust my exposure right there before taking the picture.

Also, seeing the effects of white balance in the viewfinder is extremely useful. Auto WB is OK in most cases, but you can instantly see when it is not working out and you can immediately adjust. I shoot Raw + JPEG, so I can always can adjust white balance later for the Raw file. But I prefer to get the JPEGs right so I don't have to post-process so much.

[Alan Fairley]

Nice method, but I am too lazy for it! I shoot raw and use the default matrix metering with autobracketing set up to give me +0 and +1 exposures. With the E-M5, usually there is enough headroom in the +1 exposure to be able to print from that in LR 4.1 or ACR 7, otherwise I drop down to the +0. The nice thing about this method is that I can stack the images by time in LR and use the normal exposures for review, and then unstack to print the ETTR file for the image I choose. It is very rarely that this won't work for me.

[David Tenser]

Thanks for a great article, even though some of the color spaces and settings are over my head. Two questions:

1. Why do you need to change the in-camera contrast to -2 and other settings like the color tones and gradation? I thought those settings just applied to in-camera JPEGs, not the 12-bit raw files. You said elsewhere that raw files are linear.

2. I don't really understand much about color spaces yet, but just to be sure: is there any advantage for me with switching in-camera color space to Adobe RGB? (E-PM1, shooting raw, using LR 4.1) or should I leave it at sRGB?

Thanks again!

[Pekka Potka]

David: Everything you see in camera viewfinder or monitor has gone through in-camera JPEG process. You can actually never see an RAW image, just an RGB approximation of it. (JPEG means actually a method used to compress data into a smaller space while saving and opening an RGB image. So, a JPEG image is a flavor of RGB images.) With camera´s standard settings you get an RGB (JPEG) image that shows a fairly limited range of colors and tones compared to what actually could be extracted from a RAW file. Now to your questions:
1. The meaning of these settings is to get an RAW approximation from cameras processing, shown in viewfinder or monitor, which is as close to RAW´s potential as possible from an OM-D or PEN camera. It is just to better see what can be possible. Still limited but anyway.
2. If you are shooting RAW then choose the widest color space available in you camera. AdobeRGB is wider than sRGB. Again, at this point it is just to better see what´s possible with RAW. Later (in LR) you can convert (Export) your image into sRGB color space for web.
-p-

[Olympus Maitani Fan]

Hi Pekka,

thanks for your contribution. Since I'm just amateur/weekend photographer I can get more out of my camera. I read your ETTR- E-P3 post some time ago and used about similar way with my E-P2 and it help me to get more information to my RAW files. With your help I have understand that it's more easy to expose to left later on example in lightroom than opposite with "under exposed" images.

Since we all know that four thirds sensor is a bit compromise and we are forced to think more how we are going to expose our images. You really help to get best out of our four thirds.

I'm now playing with OM-D and I still have E-P2. I'm going to try your method now with OM-D but without datacolor color checkr since I do not have it. I do calibrate my cheap panel with spyder 3 pro tool and I use different paper/printer ICC profiles with my prints, so far this have been enough for me. Do you think I should get color checkr? As you can see I'm still learning.

[Pekka Potka]

Maitani Fan: Spyder Checkr is not important per se. I used it here primarily as a subject with various colors and to illustrate the behaviour of saturated colors. You are doing things the right way as you take care of using profiles. Just enjoy shooting!
-p-

[Olympus Maitani Fan]

Thanks Pekka, I'll pass that color checkr and it is overkill for my use. I don't need that much color accuracy since I might change colors as I want and like them to be. I did try your settings yesterday in heavy rain with kit lens 12-50mm on streets. Your settings works great and I'm happy with outcome after processing them in lightroom. Olympus could have made larger lens hood for 12-50mm, water drops were all over filter :)

[David Tenser]

I forgot to get back to this thread: Thanks Pekka for your response. I think I understand what you mean, but wanted to ask a couple of followups to be sure.

So, changing gradation to standard and reducing contrast and saturation to -2 is just done so that the histogram and high- and lowlight clipping blinkies are as close approximation to the actual raw data as possible? In other words, if the saturation was set to +2, the highlight clipping blinkies might start to blink earlier than when the raw sensor data actually reaches its saturation limit, since the highlight clipping blinkies report based on your in-camera jpeg settings? If so, that makes sense.

If the above is true, I guess those in-camera saturation and contrast and gradation settings really don't change anything about the raw data saved -- it's just processing instructions that can be overridden in LR4.1 anyway. Correct?

Then my last question is: which settings actually doimpact the raw data -- is it correct that changing the sharpness and noise reduction settings actually will affect the pixel information stored in the raw file? And in other words, if you want to let LR4.1 handle all of the post processing, you're better off turning sharpness and NR down to preserve as much detail as possible and worry about sharpness and noise in PP instead?

Thanks again for your answers, they really help me understand this better!

Cheers from Sweden,
David

[Jussi]

Huh, this is great and goes deep. I will start using it right now.

Thanks Pekka.

[Kielitaidoton kuvaaja]

Hieno blogi, mutta toivottavasti ehdit kääntää näitä myö kotimaiselle kielelle, tuo googlen kääntäjä kun ei oikein ymmärrettävää tekstiä tuota.

[Pekka Potka]

Kyllä tarkoitus on pitää blogi kaksikielisenä. Juuri nyt, maalla terassilla istuessani, en uskalla luvata aikatauluja ennen elokuuta. -p-