Understanding Selective Tone control

Hello, I was trying to figure out if this post would help me understand how Selective Tone works, or more to the point understand why the sliders don’t work as I would expect them to work.

There are a lot of posts here, a few tangents and a few discussions about ‘workarounds’ so I am struggling to understand if there is a consensus around if the behaviour of the Selective Tone sliders is good, bad or ugly.

Personally I think they are in the bad catagory as there seems to be a big overlapping of effect between the sliders and it seems like a pointless exercise using them.

Highlights seem also to effect midtones and to some extent shadows , midtones also seem to adjust highlights and shadows, even blacks seem to have an effect on the other tonal ranges.

Not sure if others feel the same or not…

1 Like

I’m neutral on the matter only because DxO’s software has worked this way for as long as I’ve used it - and I haven’t used other RAW developers for nearly as long. I don’t mind that the tonal ranges overlap: I understand that behavior and can usually make it work for me. Still, even if one’s expectations aren’t being driven by how other software’s similar sliders behave, there is the glaring problem that DxO’s own documented description of the sliders is different from their actual behavior:


If you select a file with a “normal” histogram (not clipping on either end) and manipulate the selective tone sliders, you’ll observe the following in the histogram: “Highlights” anchors the left end of the histogram while allowing the right end to clip. “Midtones” anchors both ends while shifting the center of the histogram. “Shadows” anchors the right end while allowing the left end to clip. “Blacks” does the same as “shadows” but with less shifting of the mid range tones.

1 Like

Yes, a number of people have noticed the ranges of those sliders do overlap more than in some of DXO’s competitors. The subject has been raised in at least a few different discussions here. I do not know if DXO has any plans to address this


Count me in the good category, I like the way the selective tone sliders work. I have never used Lightroom and I learned to use the sliders on DXO Optics Pro ver. 8. Therefore I didn’t have to unlearn using some other software, however I can understand the frustration experienced by those that expect it to work differently. All I had before DXOOP8 was DPP. Back then(2012) DPP was a mere shadow of it’s present self. I haven’t used DPP in a while but I don’t think that they have improved the RAW tab very much and the highlights and shadows sliders don’t have a lot of latitude to use. When I got DXOOP8 I was so happy that the I could see some difference between pre and post use of the sliders that I didn’t care that they overlapped.
I learned that if I lowered the highlights, I would have to raise the midtones, then lower the shadows slightly if they were too high. It’s all a balancing act and you have to use all 4 sliders, often more than once, but again I understand if someone finds this process to be too much work.
I like the sliders as they are because I learned it that way and I must say that I have never run into a properly exposed file that I couldn’t recover the shadows or highlights if needed.
I wouldn’t mind if some of the overlap were eliminated, but please don’t turn it into a Lightroom clone.

1 Like

Same here, the overlapping gives you a form of smoothness. That test tiff above is great to see how it effects what in order to get “friends” with the sliders.
Like the tonecurve isn’t straight with sharp corners but as a garden hose curly so some overlap is needed.
For compact control in highlight i use the controlpoints (and negative’s) or brushmasks.


Thanks for the responses all. Being reletivly new to PL I find the workings of these sliders boggling and as a result I don’t use them as I cannot get the effect I want out of them.

I will stick to using tone curve, local adjustments, smart lighting or a combination of. I find clear view excessively harsh at its default setting but sometimes use it at very low settings.

Try using Clearview selectively. I agree, on clouds etc it can be OTT, but is useful on greenery and such like.

1 Like

Do you have Filmpack also?
if so please reconsider your workflow. the combination of selective tone and advanced contrast is a great fine adjustment tool.

yes it is at 50% i have my personal preset made and placed it at 15% and default off, so i have only to turn it on to apply 15%.

best option is

1 Like

The overlap of the highlight and shadow sliders, when set to a high value, significantly affect the mid-tone values as well. That may not be what is wanted. Take for example the use of the shadows slider… Some of DXO’s competitors have shadows sliders that cover a narrower bandwidth allowing them to extract details from deep shadows without also significantly affecting the mid tones. DXO’s global shadows slider is less effective for that purpose on its own and requires an opposite adjustment of the mid-tones slider to compensate. I find it easier to use local adjustments to target specific areas from which I want to extract shadow detail.


I do have film pack. I will give it a try, thanks. I will also try and create a preset with a reduced level of clear view.

I do use clear view in a local adjustment and I find this quite effective.

1 Like

selective tone is luminance driven and contrast is detail driven.
So both at -30 is the same lowering as -60 only selective tone, using both preventing you to get grayish artefacts in the “blown sections”
Pushing selective down and contrast up gives you more detail in the highlights wile luminance is tempered.
pushing shadow you want details so use the contrast also to push up.
getting blotches in shadows pull contrast down. smoothing.

(test it with a shot/image of a plastered white wall, you will se what i meant.)

Don’t be afraid to use Exposure compensation in center weighted average and Smartlighting spot weighted at the same time by using both you can leave SL in a slight modes most of the time. and push exposure wile SL react on your change.

1 Like

Yes, I agree and this what I was trying to say but apparently didn’t succeed…lol.

I understand the frustration experienced by someone who has learned shadows and highlights adjustments on another software and expects DXOPL3 to behave similarly. I also understand that the procedure that I use will be considered to be too much work or too complicated by some. I forgot to mention that I sometimes have to adjust the sliders multiple times before I am happy. I also forgot to mention that I sometimes have to use local adjustments to push shadows or highlights that are too far out of bounds for the global sliders. As @OXiDant mentions I also use the advanced contrast sliders in FP5 to enhance the SA adjustments. I understand someone feeling that this is waaaaaay too much trouble.
I am no longer a professional photographer(but I’ve been there) and do not have to process large batches of photos under a deadline any more. I am now just an hobbyist so I don’t mind the trouble, in fact I enjoy it. So, I’m happy with the sliders the way they are and as Peter pointed out, sometimes the overlapping of the sliders provides subtle gradations that wouldn’t be addressed if the sliders had a very narrow band of effectiveness.
I am not opposed to eliminating some of the overlap of the sliders but I hope that some is left in place. Perhaps the solution could involve a switch between a “narrow band” selective adjustment tab and a “legacy” version where the user could decide which to use based on the photo that is being edited at the moment.

Mark and Mark,(edit @mwsilvers and @rrblint ; we have too much Marks here…:woozy_face:)
Despite i use the sliders as they work there is an improvment possible.
I don’t know the reason for the spread overlap and we talked about our expectations above as i remember correctly, also i was for less overlap.
The workaround is local adjustmenttools which works fine.

A switch would be a problem i think then you need TWO algoritm’s in the air.

Step 1 is understanding WHY dxo made the behaviour as it is.
Step 2 is determine IF there assumption of the way it works is good or bad judgement.
Step 3 is is it easy to solve?: Like the opacityslider in local adjustment palette, can we have a fine tuning slider in contrast and selective tone so we set our own prefered overlapping area.

Personaly i think a “opacityslider” aka “featheringslider” to set the hole palette’s behaviour in overlapping from all sliders would be the holy grale for most of us. But as i allready said, first i need to understand why both work as they work. Maybe it’s part of a much greater algorithm which is the core of the raw conversion software and changing this is breaking up the hole program.

We need the staff for that.

I started this thread a long time ago. The purpose was to understand how the tone controls work and to discuss work-arounds to any problems.

This thread’s purpose was not

Anyone can start a new discussion. Could I suggest that, if you are not refining our knowledge of how the controls work or suggesting a work-around, you create a new topic?

For those unhappy with the tonal sliders, as they operate today, consider that a tonal slider is a combination of a fixed luminosity mask and and adjustment applied to the pixels selected by the mask. The adjustment is either a brightness or exposure adjustment. On some systems, a contrast or micro-contrast adjustment is also made.

What this means is that, if PL were to add luminosity masks, you could define your own ranges and have these do whatever you wanted. You might want to read about and vote for the following feature:


Thanks. It seems like the same topic to me. Is creating multiple discussions on the same topic the done thing in a forum?

I have the information I need now, so probably not worth creating a new discussion.

I wouldn’t say it was the same topic. This discussion is on how the features work. Your discussion is on whether how they currently work is a good thing. These are two completely different topics.

I think they are somewhat related. So we will have to agree to disagree.

Thanks Peter, your idea could very well be the solution that we are looking for. I also like the idea put forth by @freixas using luminosity masks. Surely one of these ideas is do-able.