Odd create anchor from image behavior

[pixel8tor]

Love all the new changes you've made since beta 10, especially to the curves display. Very nice! The control handles in the Hue/Sat dialog are still a bit too gray for my taste, and again I wish they were just solid, rather than outlined circles, but I can at least live with them this way. But they do seem a bit small.

One strangeness I've noticed, and this has been this way for a *long* time, is in the curves dialog. If you open curves, and mouse click inside the curve area to create a new anchor, before releasing the mouse button, you can drag that anchor all around the curve area without constraint and it follows the mouse movement (up, down, right, left), and as long as you don't go outside the curves area, which will delete the anchor, everything works as expected. But you can also create an new anchor by clicking on the image, and then dynamically adjust it by moving the mouse around, before releasing the mouse button. Unfortunately this is where the logic gets lost. There seems to be very little connection between how the new anchor point moves and how the mouse moves. For some reason the mouse pointer changes to a horizontal split cursor, I guess to mean move the mouse side to side, but the anchor point moves up and down. Really odd, and moving the mouse around doesn't coincide with how the anchor point moves. It seems to be only able to move up and down. It's hard to explain, and is also inconsistent. It's best if you just try it and see the difference in behavior compared to an anchor point created inside the curve area. I think the behavior should be the same whether the anchor is created inside the curve area, or by clicking on the image. Can anyone else confirm this behavior and give some feedback how they think it should work. Again, I know this is somewhat confusing to describe, but is evident if you try it for yourself.

Thoughts?

[shijan]

But you can also create an new anchor by clicking on the image, and then dynamically adjust it by moving the mouse around, before releasing the mouse button. Unfortunately this is where the logic gets lost. There seems to be very little connection between how the new anchor point moves and how the mouse moves. For some reason the mouse pointer changes to a horizontal split cursor, I guess to mean move the mouse side to side, but the anchor point moves up and down. Really odd, and moving the mouse around doesn't coincide with how the anchor point moves. It seems to be only able to move up and down.

For my opinion this is normal behave. When you select point on the image you only need to move curve point up/down to adjust curve brightness starting from that selected range of the image. Or in single channels you can manually adjust color tint or white balance in selected channel.
For example if you see blue tint in shadows, you select blue channel curve, click on the image shadows and drag mouse to move curve point up/down to neutralize that tint.
If allow to move selected point left/right, as you suggested, it is easy to lost your tonal area because human can't phisycally move mouse exactly in parallel up/down.

[pixel8tor]

For my opinion this is normal behave. When you select point on the image you only need to move curve point up/down to adjust curve brightness starting from that selected range of the image. Or in single channels you can manually adjust color tint or white balance in selected channel.
For example if you see blue tint in shadows, you select blue channel curve, click on the image shadows and drag mouse to move curve point up/down to neutralize that tint.
If allow to move selected point left/right, as you suggested, it is easy to lost your tonal area because human can't phisycally move mouse exactly in parallel up/down.

It sounds like you'd be cool with every time you pressed the {RIGHT] arrow key on your keyboard the cursor moved up! Is that true? Do you have any idea how crazy that sounds. And what key would you press to actually move your cursor right? This is not Bizarro World! At least not where I live.

[Herbert123]

I think it' s due to how the cursor movement is interpreted. The action after clicking in the image controls the strength of that curve move rather than the overall position of the anchor point within the curve.

Since screens are wider than they are high it makes sense to control that strength through moving left<-->right in the screen space rather than up down.
In apps such as Blender the same reasoning is applied: more control over the values.

What is odd though is the implementation of this behaviour in PhotoLine: it sort-of also works to move the mouse up and down, but it is entirely based on the original position of the mouse cursor whether the up and down or the left to right mouse action will work or not. And the range of movement also changes (dramatically at times) depending on where the user begins clicking in the image.

For example, begin the move at the very top of the screen: left<-->right works. Up<-->down sort-of works for up, but the range of control is limited.

If the image point that you'd like to sample is at the utter most left bottom of the screen: moving the point up in the curve is possible with moving up, but it is only possible to reach the bottom of the curve by moving the mouse cursor to the far left.

The behaviour changes depending on where the user clicks, and seems based on the relative screen positioning where the user happens to start dragging.

(I wondered if this odd behaviour also might have something to do with the Edge Width settings for image and document mode. I always set those to 1000% (max) in order to be able to pan the view freely. I tested this and it didn't make much of a difference, if any at all.)

In one case I had to first move up to the top border of the screen and then veer off towards the left to reach the max value in the curve!!!

So yes: I agree. The mouse movement after clicking to sample the curve and adjust the strength of that curve move makes little sense, because it seems entirely screen position dependent. When that context changes, the way to adjust the strength of that move via the mouse also changes.

That is inconsistent UX behaviour in my book.

[shijan]

It sounds like you'd be cool with every time you pressed the {RIGHT] arrow key on your keyboard the cursor moved up! Is that true? Do you have any idea how crazy that sounds. And what key would you press to actually move your cursor right? This is not Bizarro World! At least not where I live.

Read my post again please. Seems you don't understand what i wrote at all.
It is classic curves behave when you click on desired image area by mouse and quickly (and not too precise) adjust brightness up/down only.
By the way, you can click directly in image with Curves picker and instantly use arrow keys to move point up/down AND left/right! It is way more precise way to work than all those mouse dragging.

[pixel8tor]

You state quite clearly in your first sentence "For my opinion this is normal behave." Everything after that is only trying to justify your pinion. It completely misses *my* point and you go off on your unsupported tangent - well unless you live on Bizarro World. So maybe *you* should reread my OP and try to really understand it. And then your second post, well it's just more of the same.

As Herbert123 says "What is odd though is the implementation of this behaviour in PhotoLine", and "That is inconsistent UX behaviour in my book.", and "So yes: I agree. The mouse movement after clicking to sample the curve and adjust the strength of that curve move makes little sense . . .".

[Juan]

If I understood right, when you click on the image while having the curves, the anchor that is created on the image is pixel value (color) based, it is not meant to move the anchor by any means,, it is just registering the pixel value of the image under the mouse while dragging.

Another example is when using Hue/Saturation or Hue editor, while dragging the mouse over an image it samples the pixel color.

[pixel8tor]

If I understood right, when you click on the image while having the curves, the anchor that is created on the image is pixel value (color) based, it is not meant to move the anchor by any means,, it is just registering the pixel value of the image under the mouse while dragging.

Another example is when using Hue/Saturation or Hue editor, while dragging the mouse over an image it samples the pixel color.

No Juan, you don't understand. Reread my OP and *follow it* - don't just think about it! You have to see the effect, or should I say crazy behavior. Also read herbert123's earlier post. He obviously understands the issue, because he tried it, and didn't just think about it.

[Juan]

If I understood right, when you click on the image while having the curves, the anchor that is created on the image is pixel value (color) based, it is not meant to move the anchor by any means,, it is just registering the pixel value of the image under the mouse while dragging.

Another example is when using Hue/Saturation or Hue editor, while dragging the mouse over an image it samples the pixel color.

No Juan, you don't understand. Reread my OP and *follow it* - don't just think about it! You have to see the effect, or should I say crazy behavior. Also read herbert123's earlier post. He obviously understands the issue, because he tried it, and didn't just think about it.

What made you think I didn't open PL and tried?

Following your logic, when you click and drag in the picture while having the curves adjustment layer open you want to see created anchor point moving all around?
Not sure if you have noticed but while hovering (not clicking) over a picture, the curves dialog shows a red anchor of that pixel value.

[pixel8tor]

Following your logic, when you click and drag in the picture while having the curves adjustment layer open you want to see created anchor point moving all around?
Not sure if you have noticed but while hovering (not clicking) over a picture, the curves dialog shows a red anchor of that pixel value.

No, once again, it's that the logic of creating/moving a new anchor are completely different depending on whether you click inside the curves dialog box, or whether you click on the image. They should produce identical results. Clicking inside the dialog box it's very rational. The new anchor movement mirrors your mouse movement, until you release the mouse button, or hit the edge of the curves window, which causes that anchor to be deleted. Now, with a cleared curves dialog, click somewhere on an image, keep the mouse button pressed, and move your mouse around. Welcome to Bizarro World! There is *no* mirroring of your mouse to what's going on on the screen. Move you mouse up/down, then try right/left. Notice anything odd? First off, it makes *no* sense that moving your mouse right/left actual affects things on the screen in an up/down direction. Since the correct logic is already in Photoline, as is evident by creating anchors inside the curves dialog works properly, I don't understand why this Bizarro World reasoning was specially coded this way. It makes *no* sense! But since there is good logic already available, it shouldn't be too difficult to call the rational code function/subroutine instead of running the Bizarro World code. HTH

[Martin Huber]

(...) it's that the logic of creating/moving a new anchor are completely different depending on whether you click inside the curves dialog box, or whether you click on the image. They should produce identical results.

That's your opinion.
But these are two different thing:

• Clicking in the curves control creates a new point that can be modified.
• Clicking in the document, you choose a brightness level and you can directly modify that brightness level. So you don't modify the anchor point, but the clicked brightness level. And you can do that by either moving the mouse up/down or left/right.
  The mouse cursor isn't ideal as it doesn't reflect the possibility to drag up/down, but it is difficult to find a meaningful cursor that isn't confusing on the other side.