[beholder3]

Today I shot a Canon 5D Mark III side by side with a Pentax K-1. Same exposure settings, same subject etc.

First I had a look at them in RawDigger to see how the metering and Raws differ. Basically pretty much the same, with the K-1 metering a tiny 1/4 stop more agressively moving the histogram to the right (kind of silly given that Canon sensor really need ETTR, not Sony sensors). Same headroom in the lights.

Then I imported them into lightroom and suddenly the K-1 pictures looked unnaturally dark compared to the Canon ones. On average the K-1 files neededa exposure adjustment of about +2/3.

Wondering how that could be I looked at Adobe's hidden exposure compensation levels in the raw metadata after import into LR and there we have the culprit: Exif:Baseline Exporsure is...
Canon 5D3: +0,25 EV
Pentax K-1: -0,5 EV

This nonsense deliberate (hidden) change by the Adobe software gives the impression K-1 images are underexposed, but they clearly are not.

If you react on this by dialing in ExpComp +2/3 in your camera you are creating more blown highlights as now you force the camera to overexpose. No wonder there still exists the myth, that Canon cameras offer more room in the highlights.

If you do not react on this, images appear too dark in lightroom which is visually not really appealing. Users now must manually correct it and wonder "why the camera doesnt meter better". Which is not the case.
Most casual users are simply unaware of what LR/PS does secretly.

I would say Adobe needs to do their homework.

[monochrome]

^^ Adobe needs to do their homework, or . . . .

Ricoh needs to spend some money and supply Adobe the data to create an accurate camera profile.

[LensBeginner]

^^ Adobe needs to do their homoework, or . . . .

Ricoh needs to spend some money and supply Adobe the data to create an accurate camera profile.

Do you mean "money" as in "money to supply Adobe with" or "money to Adobe and supply them with"?

[TonyW]

Adobe are not alone in applying baseline exposure compensation at least from my last look into this topic.

From memory others such as DxO and (I think) capture one also do this.

Apart from the baseline exposure compensation a curve is also added to brighten the image (in Process 2012). So it is possible that if you do go by the look of the image on screen of your raw converter of choice that you could think that you were overexposing by potentially up to a +2EV (baseline + brightening curve) and of course if adjusting exposure in camera to compensate in reality you would be underexposing - or perhaps more correctly giving less than optimal exposure. In cases where the baseline exposure is usually negative (Pentax?) the effect is perhaps less dramatic due to the hidden curve correction to make the image look good/acceptable on screen.

Due to the fact that Adobe and others produce raw converters for many cameras from nearly all manufacturers there is a reason for this behavior and quoting (I think Eric Chan):
In order to have common controls such as Exposure compensation behave the same way across the different cameras, Camera Raw applies a baseline exposure compensation that varies from model to model in order to get them all to behave similarly when the exposure compensation control is set to its default value of zero.

The only way to view the raw data AFAIK (after some slight application manipulation) is to view in an editor that does not apply any compensation e.g. RawTherapee or in a program such as RawDigger or FastRaw viewer.

You can of course make your own camera profile quite easily and negate both the hidden baseline exposure compensation and any hidden Adobe curve manipulations.

Still I agree it would be good for users to be aware of what is happening under the hood

EDIT: Although not a user of Digital Camera Utility (SilkyPix?) it seems to me that they do the same i.e. apply hidden correction.

---

Last edited by TonyW; 06-28-2016 at 07:06 AM.

[Dlanor Sekao]

Thank You for Posting this.....to me its very helpful knowing this and now I can deal with it.

[enoeske]

Today I shot a Canon 5D Mark III side by side with a Pentax K-1. Same exposure settings, same subject etc.

First I had a look at them in RawDigger to see how the metering and Raws differ. Basically pretty much the same, with the K-1 metering a tiny 1/4 stop more agressively moving the histogram to the right (kind of silly given that Canon sensor really need ETTR, not Sony sensors). Same headroom in the lights.

Then I imported them into lightroom and suddenly the K-1 pictures looked unnaturally dark compared to the Canon ones. On average the K-1 files neededa exposure adjustment of about +2/3.

Wondering how that could be I looked at Adobe's hidden exposure compensation levels in the raw metadata after import into LR and there we have the culprit: Exif:Baseline Exporsure is...
Canon 5D3: +0,25 EV
Pentax K-1: -0,5 EV

This nonsense deliberate (hidden) change by the Adobe software gives the impression K-1 images are underexposed, but they clearly are not.

If you react on this by dialing in ExpComp +2/3 in your camera you are creating more blown highlights as now you force the camera to overexpose. No wonder there still exists the myth, that Canon cameras offer more room in the highlights.

If you do not react on this, images appear too dark in lightroom which is visually not really appealing. Users now must manually correct it and wonder "why the camera doesnt meter better". Which is not the case.
Most casual users are simply unaware of what LR/PS does secretly.

I would say Adobe needs to do their homework.

Do you have highlight protection on?

[beholder3]

Do you have highlight protection on?

I always have it on Auto in the camera, but in case of the image file I am discussing here, it did not kick in (still ISO 100).

I just adjusted my K-1 .dcp camera profiles to counter the effect and now am quite happy with the results I am seeing upon loading in LR. Much more so than before.

My most liked profile now contains the following ingredients:
a) colorchecker based profile using two extreme color temperatures
b) Anti-Baseline Exposure compensation curve adjustment
c) add a tiny contrast S-curve enhancement to my liking
d) correct the red color magenta tint and oversaturation that can appear on really red subjects

[Bruce Clark]

Can this controlled by the default develop settings in Preferences ie. "Make defaults specific to camera serial no" ?

[Kunzite]

^^ Adobe needs to do their homework, or . . . .

Ricoh needs to spend some money and supply Adobe the data to create an accurate camera profile.

Which should be done in a luxury cruise in Caribbean?

[beholder3]

In order to have common controls such as Exposure compensation behave the same way across the different cameras, Camera Raw applies a baseline exposure compensation that varies from model to model in order to get them all to behave similarly when the exposure compensation control is set to its default value of zero.

While I could accept that they want to level the starting situation, the scenario we now have here is exactly the opposite:
The starting look and histogram situation in LR is not made more similar across two cameras using the hidden adjustments, but it is made more different.

[TonyW]

While I could accept that they want to level the starting situation, the scenario we now have here is exactly the opposite:
The starting look and histogram situation in LR is not made more similar across two cameras using the hidden adjustments, but it is made more different.

If you only subtract baseline exposure you will only be part way there toward seeing what your camera produces.

Ideally you should be also change to Process version 2010 and taking the sliders to zero. Default are Blacks = 5, Brightness +50, Contrast +25 - reduce all these to 0.00 and switch back to Process Version 2012 and look at the sliders.

You should see the values applied between Proc 2010 to Proc. 2012 i.e Exposure = -1.00, Contrast = -33, Blacks = +25. If you now subtract your baseline exposure calculation you should see a very similar presentation of the image data between RawDigger, FastRaw Viewer and Raw Therapee at least as far as density and contrast go.

Worth bearing in mind that initial conversion of raw data is somewhat subjective and what one person views as an acceptable starting point may well differ from another.

In general it appears Adobe have chosen to present an image from each camera raw data that represents a reasonable starting point (at least to them) relating to colour, contrast and brightness with Adobe Standard profile. With some cameras Nikon, Canon they have chosen to offer settings that mimic the look of the presets e.g. Vivid, Camera Neutral etc. and from what I have seen with Nikon have done a very good job. Bearing in mind that neither may be accurate or even desirable renditions of the scene and of course these settings are really only relevant to JPEG (mimicking what is seen on the limited camera LCD)

---------- Post added 06-28-16 at 09:07 AM ----------

Can this controlled by the default develop settings in Preferences ie. "Make defaults specific to camera serial no" ?

Yes they can but unless you create your own profile the settings will not show as zero defaults but reflect the removal of the Adobe baseline and brightness settings. Baseline exposure compensation may vary with ISO used so should be checked.