Initial Color Eq docs based on documentation by Boris. #668
Conversation
|
Hi Mica, here I am. How do we proceed now? I'm almost finished with Manual as a document. Tomorrow I can offer it. |
|
@s7habo if you click on the "Files Change" tab on this page, you'll see the page I added. You can either add comments, click and edit it to propose changes, or make comments that I can accept as changes. You can see the flow for that here: https://docs.github.com/en/pull-requests/collaborating-with-pull-requests/reviewing-changes-in-pull-requests/reviewing-proposed-changes-in-a-pull-request Feel free to add whatever you'd like! |
| ## options | ||
|
|
||
| white level | ||
| : Set the upper bound for the brightness correction via the slider or picker. |
|
FYI we want to try and avoid too many images if possible, since it's a pain to re-create them every time the UI or functionality changes. |
|
@paperdigits is it ok for you if I comment like this? I would hate to make sugestion about changes in the code itself because I have no idea how to use this language.
@elstoc I understand that this means more work, but it also means a lot of work to explain things to users afterwards, when there are hardly any visual examples for functions that are sometimes explained in a very technical way. I will be happy to re-create them afterwards if the changes require it. When the time has come, you are welcome to contact me. |
|
@s7habo yeah this is fine. If you can "suggest" the changes, then I can just click "accept" instead of having to copy and paste. Thanks for the work, the range selection stuff is completely unrecoverable from just the UI itself :D |
|
@s7habo I don't mean no images but let's try to limit the number if we can and make them simple enough that they won't need too much maintenance later on. If we are using images to illustrate concepts, it's better to include only as much of the UI as you need to explain things. For example if you're trying to illustrate the shape of the curve when a parameter is applied, then if you only show the curve itself and not the rest of the module UI, it will probably never need to be updated again. |
Ok. I will try to pay attention to it. I will continue now and then we can talk about the details. |
|
The point “effect radius” is still open. I'll deal with that tomorrow. |
|
I am now finished. I also have the whole thing as an odt document which I can offer together with screenshots. |
Co-authored-by: s7habo <[email protected]>
Co-authored-by: s7habo <[email protected]>
Co-authored-by: s7habo <[email protected]>
Co-authored-by: s7habo <[email protected]>
Co-authored-by: s7habo <[email protected]>
Co-authored-by: s7habo <[email protected]>
|
|
||
| _Visualization indicator_ | ||
|  | ||
| shows which pixels have been selected. Red color indicates the selection, and pixels in the blue area are not taken into account. |
There was a problem hiding this comment.
Note that if a blend mask is active, the visualization can only be seen in areas where the blend mask selects the module's output.
|
A BIG thank you to all of you. BTW I will check how/if the mask visualizing restriction mentioned by @ralfbrown can be avoided. |
|
I haven't looked at the code, but it's pretty clear the mask visualization uses the RGB channels rather than the alpha channel. It would take a reworking of the yellow masks generation in gamma.c to generate the red/blue masks with a new mask display mode (or two) and have the equalizer put the "strength" value on the alpha channel. |
|
Yes but we have a way to pass thru the RGB data. |
|
@elstoc I hope this is good enough for a first draft. |
|
|
||
| This module works in darktable's UCS 22 color space. | ||
|
|
||
| The entire hue range is represented as a curve with eight overlapping sections divided by eight fixed nodes. Choose whether you wish to adjust (select) pixels based on their hue, saturation, and brightness. You can then use the equalizer interface nodes, on their respective tabs, to adjust the hue, saturation, and brightness of ranges of pixels selected via this method. |
There was a problem hiding this comment.
You can't choose to adjust/select by saturation or brightness. All of the selections are by hue. It should perhaps say something like "Pixels are selected based on their hue, and the module can be used to adjust the hue, saturation, or brightness of those selected pixels"
|
|
||
| The entire hue range is represented as a curve with eight overlapping sections divided by eight fixed nodes. Choose whether you wish to adjust (select) pixels based on their hue, saturation, and brightness. You can then use the equalizer interface nodes, on their respective tabs, to adjust the hue, saturation, and brightness of ranges of pixels selected via this method. | ||
|
|
||
| The distances between nodes cannot be changed but all nodes together can be moved by +/- 23 degrees along the color wheel using the node placement slider, allowing you to fine-tune the selection of a specific color. |
There was a problem hiding this comment.
No need to specify +/- 23 degrees here
|
|
||
| ## hue tab | ||
|
|
||
| Click+drag the control points on the curve to change the color of all pixels where the mask includes the given color. |
There was a problem hiding this comment.
This is the first time we've mentioned the word "mask" (without defining what is meant)
|
|
||
| Color Equalizer is essentially an implementation of the Color Zones module, but works in the scene referred part of darktable's pixelpipe. | ||
|
|
||
| This module works in darktable's UCS 22 color space. |
There was a problem hiding this comment.
I don't think most users are going to care about this
|
BTW I've created a separate issue around central documentation of color pickers (#673) |
|
As the discussion is getting a bit hard to follow, and I wanted to do a bit of reorganisation, I have created another draft in a new PR #674 |
Color Eq docs based on a document sent to me by Boris.
darktable issues that might need labels adjusted if this is merged:
darktable-org/darktable#16633
darktable-org/darktable#16543
darktable-org/darktable#16446
darktable-org/darktable#16400
darktable-org/darktable#16301
darktable-org/darktable#16279
darktable-org/darktable#16236
darktable-org/darktable#16089