Rongnongno Loc: FL

Difference?

Magenta filter reduces the green and equalize RGB toward a unitary value app.
uniWB equalize RGBG to 1.

Ysarex Loc: St. Louis

The filter would not be as precise as setting a unity WB.

selmslie Loc: Fernandina Beach, FL, USA

Guillermo Luijk says you need to base it on a target set to R=162, G=64 y B=104. There is no magenta filter than will do that. The resulting WB setting may be different for different cameras.

Luijk did not invent UniWB in 2008, he gives credit to Iliah Borg for getting there first. Then Borg went on to develop RawDigger, a much better way to see what's happening.

In theory it's supposed to align the camera's JPEG histograms but you have to look at the camera's JPEG histograms to see that. With spot metering the WB setting has no effect on the meter's response.

You seem confused about how to use it in conjunction with ETTR. Luijk does not address that.

I am still waiting for an answer from you or Bob. Why are you using UniWB?

dpfoto Loc: Cape Coral, FL

How do you pronounce your name, "Rongnongno"? Is it Rong-nong-no?

Rongnongno Loc: FL

gn is equivalent to the ñ in Spanish so ron-ñon-ño

Best bet thought is to use RGG...

Ysarex Loc: St. Louis

No he doesn't. Those values are the result he arrived at from the example test he presented with one specific camera. The point of the example was that each camera has to be tested to determine what will be unique values for that camera make/model.

He does address UniWB and ETTR -- in great detail: "It is well known how useful is looking at the histogram and the clipping information warning that our cameras provide. Specially if we are exposing to the right, taking a look on them is very interesting and almost mandatory." In that quote from his article "exposing to the right" is a link: http://www.guillermoluijk.com/article/ettr3/index.htm

selmslie Loc: Fernandina Beach, FL, USA

I never got past Lujik's original article from 2008 because the whole process seemed to be totally academic in nature. In that article he says nothing about how or why to use it, just how to set it up. Luijk's 2009 article finally addressed this but I had already moved on. UniWB may appeal to some people but I am too pragmatic to bother with it.

The problem with UniWB is that it requires the camera to be set to a WB that makes it difficult to review the image. Then you have to attempt to get all three histograms to the right of the display if you if you want to expose to the right.

My cameras have four histograms including a luminance histogram which closely tracks the green histogram. That's not surprising there since with daytime landscapes the green histogram represents 50% of the pixels (56% for X-Trans). Red or blue only go farther when the brightest part of the image is red or blue and that is rare. But even if all you see is the luminance histogram you are getting all you need to get to ETTR. Many cameras, especially mirrorless, will show you that live.

But the key to getting useful luminance information is to not mess with the camera's JPEG by deviating from the camera's default JPEG rendition (see https://www.uglyhedgehog.com/t-691294-4.html#12124757).

It doesn't take much research to discover that the camera's meter is not affected by the white balance setting. This is because it is linked to the potential raw values, not the JPEG. That's why it seems silly to set the camera up to use UniWB and then use the camera's spot meter to determine the exposure while ignoring the histograms. You use the spot meter with any WB setting including Daylight and end up with the exact same exposure.

So we have one user here who religiously promotes ETTR but sees through the folly of UniWB and doesn't use it. We have another who has his camera set to UniWB but does not use it to determine his exposure. And finally we have the OP trying to understand how they work together. But none of the three is willing to look at the raw histograms to discover the relationship between the camera's JPEG histograms and the raw histograms. A little research will reveal that, at the extreme right end where ETTR is based, they both come together so that a blown JPEG highlight is extremely likely to be accompanied by a blown raw highlight (see https://www.uglyhedgehog.com/t-691294-1.html#12120601).

The rest of us just look at UniWB as a big waste of time.

Ysarex Loc: St. Louis

Yes he does mention why to use it. I quoted that article in my response to your original error: https://www.uglyhedgehog.com/t-691512-1.html#12125285 Here's the quote again: "It is well known how useful is looking at the histogram and the clipping information warning that our cameras provide. Specially if we are exposing to the right, taking a look on them is very interesting and almost mandatory."

selmslie Loc: Fernandina Beach, FL, USA

I can get all of that without using UniWB.

But like UniWB, ETTR is a waste of time in broad daylight anyhow. See https://www.uglyhedgehog.com/t-691294-4.html#12125526

Rongnongno Loc: FL

You are engaging with someone who has no clue and just likes to read himself. Most folks have given up him. I have long done so long ago.

Ysarex Loc: St. Louis

I couldn't care less.

I'm just helping you to be accurate about what Guillermo Luijk wrote. Twice now you said Luijk did not mention ETTR and why and how to use UniWB to achieve ETTR.

First you said: "You seem confused about how to use it in conjunction with ETTR. Luijk does not address that."
Second you said: "I never got past Lujik's original article from 2008 because the whole process seemed to be totally academic in nature. In that article he says nothing about how or why to use it, just how to set it up."

You're wrong. He does mention why and how UniWB is useful for people interested in ETTR. I've provided a quote for you twice now. And again: "It is well known how useful is looking at the histogram and the clipping information warning that our cameras provide. Specially if we are exposing to the right, taking a look on them is very interesting and almost mandatory."

selmslie Loc: Fernandina Beach, FL, USA

Yes,I saw that in the second article.

But UniWB is still misunderstood by the OP and not really being used as intended by anyone else here. It only appeals to the incurably curious and the lunatic fringe.

Ysarex Loc: St. Louis

Yep, surest proof that you're dealing with 100% pure bovine excrement is when the purveyor of said BS cites themselves as a reference.

Rongnongno Loc: FL

Beside the 'intermission'.

I have been thinking about your answer. You are correct as the effect will be different depending on the opposite color and a few other things. Setting the magenta bias in the sensor on the other hand might be a different kettle.

To be tested.

Basically getting R=1, G=.5, B=1, G=.5

Advantages, I think, JPG histogram is reported correctly and so is the raw despite the DR compression.

Am I crazy to think this way?

Ysarex Loc: St. Louis

Which is linked from the first article using the term "expose to the right."