Page 2 of 4

Re: value study - blend if

Posted: Sat Jan 26, 2019 12:52 am
by minniev
Psjunkie wrote:
Sat Jan 26, 2019 12:14 am
Try holding down the alt/option key and splitting the slider..move the right half of the slider all the way right then the left half to taste...
I moved the right half of the lower left slider all the way off the right side of the screen. Then tinkered with the rest till all the funny blocked up stuff went away. The white paint area or something else seemed to lighten the shadows. But it looks kind of milky. Or is it just me?

Re: value study - blend if

Posted: Sat Jan 26, 2019 1:01 am
by Psjunkie
Really hard to say from this screen shot minniev...moving that left half of the black slider a little further to the right might resolve you milky issue I dunno..compare the black square right to the one I showed..you see the smaller lighter square, that kinda gives you an idea of where you are....the smaller square does not always have to disappear, it all depends on what you are wanting the blending to look like....the whole point is to use blend-if to blend and have a natural look...whether it be blending in a colors or filters or an adjustment layer or whatever...the power to blend whatever you want into whatever you want, highlights, mid tones, or shadows...you'll get the hand of the more you play with it..

Re: value study - blend if

Posted: Sat Jan 26, 2019 1:08 am
by minniev
PietFrancke wrote:
Fri Jan 25, 2019 10:54 pm


We will get it... When I did mine, the darks were on the bottom, then the Mids, then the Lights on top. The slider you have to adjust is the underlying layer (the bottom one) - but you know that NO I DON'T, I KNOW WHICH LAYER IS WHICH BUT NOT WHAT SHOULD BE DONE TO AFFECT ANYTHING IN A PARTICULAR WAY! You have to set the sliders to set matching ranges. Luckily there are numbers that document the Left Side (dark) and the Right Side (light). My values were:

Light Layer - 133-255 -- on top -- the light layer is applied to anything beneath that is between 133 and 255.
Mid Layer - 79-133 -- in middle -- the mid layer is applied to anything beneath that is between 79 and 133.
Dark Layer - 0-79 -- on bottom -- the dark layer is applied to anything beneath that is 0-79.

I think truly it would be better if it was 0-78 and then 79-whatever, etc.. You see what is happening. Yeah, I saw the go gray sometimes too, I haven't figured out why it does that.

You can see how those ranges would "carve up" the image into three distinct groupings. Of course you could always "soften" the edge by using Alt on the slider button, but then you would lose the crisp edge.

We aren't going to lose you on this one!! Once you see how those ranges work, you can add more another one and divy it into four groups for example. The secret is to have the ranges be distinct and mutually exclusive.

Edit - and when you replace a value-range, it is best to replace it with a color/tone that is in the SAME range, otherwise, things get messed up above.

LOL - Frank answered too!
I am not even sure what the problem is that I am trying to solve here. I googled a few tutorials and they mostly demonstrated how to blend text or shapes of one tonality into a background of another tonality. I can see how that might help with designing ads or something but I'm trying to figure how to use it for art or photos to achieve something. Frank's experiment did show me a little something with that splat of white paint blended in to brighten (I think that's what it did).

My other exercise was to try to create a single image out of two images of an HDR sequence using blend/if. Which is a kind of blending of lights and darks from two images into one. I did finally achieve that, and got rid of all the gray blocked up stuff to make a reasonably decent blended image. I think that's the same principle, just applied differently. Which makes me wonder if HDR software is like a blend/if command on steroids.

Re: value study - blend if

Posted: Sat Jan 26, 2019 1:57 am
by PietFrancke
I like the HDR - and discussing how to use it, let's defer that for a moment, but I want to ensure that we are first on the same page for how it works. Study this table - If you understand it, we can move on. (my purpose for using it was as a "simplification" - an effort to reduce detail yet to still see the forms).. I am thinking that Blend-If with the underlaying layer is effectively a type of luminosity mask..

Now to address your note in RED. (but I might be misunderstanding, I think you've got the mechanics, but in case not)...

First, for the underlying layer, we will set up the ranges to fall into exact thirds. 256/3 = 85... so:
Dark layer - set underlying values on slider to 0-85
--- Pick a color for your fill layer that has HSB (B is for Bright) Bright percent less than 34 - (make it 20 percent Bright)

Mid layer - set underlying values on slider to 86-170 (even though my example was 79-133) 86-170 puts it directly in middle third.
--- Pick a color and ensure the B of HSB falls between 34 and 66 (make it 50).

Light layer - set underlying values on slider to 171-255 ---
--- Pick a color and ensure that B of HSB falls between 67 and 100 (make it 80 percent Bright)

First, I just want to ensure we are on the same page - that is - You understand the mechanics!
img5.jpg

Re: value study - blend if

Posted: Sat Jan 26, 2019 4:43 am
by minniev
PietFrancke wrote:
Sat Jan 26, 2019 1:57 am
I like the HDR - and discussing how to use it, let's defer that for a moment, but I want to ensure that we are first on the same page for how it works. Study this table - If you understand it, we can move on. (my purpose for using it was as a "simplification" - an effort to reduce detail yet to still see the forms).. I am thinking that Blend-If with the underlaying layer is effectively a type of luminosity mask..

Now to address your note in RED. (but I might be misunderstanding, I think you've got the mechanics, but in case not)...

First, for the underlying layer, we will set up the ranges to fall into exact thirds. 256/3 = 85... so:
Dark layer - set underlying values on slider to 0-85
--- Pick a color for your fill layer that has HSB (B is for Black) Black percent greater than 67 - (make it 80 percent black)

Mid layer - set underlying values on slider to 86-170 (even though my example was 79-133) 86-170 puts it directly in middle third.
--- Pick a color and ensure the B of HSB falls between 34 and 66 (make it 50). The B of HSB is expressed as a Percent of Black.

Light layer - set underlying values on slider to 171-255 ---
--- Pick a color and ensure that B of HSB falls between 0 and 33 (make it 20 percent black)

First, I just want to ensure we are on the same page - that is - You understand the mechanics!

img5.jpg
Probably I don't. Here is what happened. The B values seem to inversely correspond to brightness. The higher the B value, the brighter the color, and the lower the B value the duller and darker the color? Doesn't look like anything to do with black at all.

And you mean the sliders values are set to these numbers on the blending options attached to the layer in question.

I tried a simpler picture converted to monochrome to see if I could understand better.

Re: value study - blend if

Posted: Sat Jan 26, 2019 4:45 am
by minniev
Psjunkie wrote:
Sat Jan 26, 2019 1:01 am
Really hard to say from this screen shot minniev...moving that left half of the black slider a little further to the right might resolve you milky issue I dunno..compare the black square right to the one I showed..you see the smaller lighter square, that kinda gives you an idea of where you are....the smaller square does not always have to disappear, it all depends on what you are wanting the blending to look like....the whole point is to use blend-if to blend and have a natural look...whether it be blending in a colors or filters or an adjustment layer or whatever...the power to blend whatever you want into whatever you want, highlights, mid tones, or shadows...you'll get the hand of the more you play with it..
I watched another couple of videos and found one that brightened up a gal's belly with some white paint kinda like you did above. They didn't seem to have the milky problem, maybe some extra contrast would help. I saw one where they used it to change out a sky, which seemed useful, especially when trees are involved.

Re: value study - blend if

Posted: Sat Jan 26, 2019 6:12 am
by Duck
 ! Message from: Duck
Sorry Piet, I moved this to the Post Processing forum as your original post, nor the subsequent posts, fit under the "tutorial" qualifications.

If you feel I am wrong, please PM me and we can discuss it further. I hope you understand.

Re: value study - blend if

Posted: Sat Jan 26, 2019 12:41 pm
by PietFrancke
Duck wrote:
Sat Jan 26, 2019 6:12 am
 ! Message from: Duck
Sorry Piet, I moved this to the Post Processing forum as your original post, nor the subsequent posts, fit under the "tutorial" qualifications.

If you feel I am wrong, please PM me and we can discuss it further. I hope you understand.
Makes sense to me, Thank you Duck!

Re: value study - blend if

Posted: Sat Jan 26, 2019 12:53 pm
by PietFrancke
minniev wrote:
Sat Jan 26, 2019 4:43 am
Probably I don't. Here is what happened. The B values seem to inversely correspond to brightness. The higher the B value, the brighter the color, and the lower the B value the duller and darker the color? Doesn't look like anything to do with black at all.

And you mean the sliders values are set to these numbers on the blending options attached to the layer in question.

I tried a simpler picture converted to monochrome to see if I could understand better.
You are SO correct. I updated/edited my post above, hopefully it will make more sense now. The mechanics. Yes, the B of HSB IS Brightness. So, please for your example, use the new numbers which are the inverse of what I told you earlier. Please give it a shot, it should work properly now. I appologise so very much, I don't know what I was thinking.

Re: value study - blend if

Posted: Sat Jan 26, 2019 3:44 pm
by minniev
PietFrancke wrote:
Sat Jan 26, 2019 12:53 pm

You are SO correct. I updated/edited my post above, hopefully it will make more sense now. The mechanics. Yes, the B of HSB IS Brightness. So, please for your example, use the new numbers which are the inverse of what I told you earlier. Please give it a shot, it should work properly now. I appologise so very much, I don't know what I was thinking.
This seems to be more what you are getting at. Now what?