COlor Picker seems odd at highest and lowest values in gradient section

[Bill Christens-Barry-2]

Hi,

I've noticed what I think is a changed behavior of the Color Picker concerning its leftmost section that displays as a grayscale gradient. In that section, low values are displayed as dark at the top and high values displayed as light at the bottom.

In macOS 10.14.6 (Mojave), to pick the lowest value (0) you must pick just beneath the highest value in the gradient. The lowest value that can be picked at the top of the gradient is not 0 but is 1. This behavior also holds for different image types and how their values are displayed numerically. Moreover, if you click in the regions above or beneath the gradient near its lowest or highest values, the lowest non-zero value (1) or the highest value (255), respectively, is selected.

This seems counterintuitive; would it be better to have the lowest value be at the top of the gradient region? I wonder if this might be a regression, perhaps due to a non-zero based 8-bit index rollover (255 + 1 = 0).

On Windows 10 this behavior is slightly different.

Bill

--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html

[Michael Schmid]

Hi everyone,

my 3 cents: I think it would be even nicer to have an area a few pixels
wide for picking black or white.


Michael
________________________________________________________________
On 31.08.20 15:45, Bill Christens-Barry wrote:

> Hi,
>
> I've noticed what I think is a changed behavior of the Color Picker concerning its leftmost section that displays as a grayscale gradient. In that section, low values are displayed as dark at the top and high values displayed as light at the bottom.
>
> In macOS 10.14.6 (Mojave), to pick the lowest value (0) you must pick just beneath the highest value in the gradient. The lowest value that can be picked at the top of the gradient is not 0 but is 1. This behavior also holds for different image types and how their values are displayed numerically. Moreover, if you click in the regions above or beneath the gradient near its lowest or highest values, the lowest non-zero value (1) or the highest value (255), respectively, is selected.
>
> This seems counterintuitive; would it be better to have the lowest value be at the top of the gradient region? I wonder if this might be a regression, perhaps due to a non-zero based 8-bit index rollover (255 + 1 = 0).
>
> On Windows 10 this behavior is slightly different.
>
> Bill
>
> --
> ImageJ mailing list: http://imagej.nih.gov/ij/list.html
>

... [show rest of quote]

--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html

[Gabriel Landini]

On Monday, 31 August 2020 16:51:04 BST you wrote:
> my 3 cents: I think it would be even nicer to have an area a few pixels
> wide for picking black or white.

Isn't that what the Reset "region" (a small white and black rectangle) does?

Cheers

Gabriel

--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html

[Bill Christens-Barry-2]

Thanks, Gabriel,

I hadn't realized that the tool you mention handles this and provides a nice way to pick white or black. For some reason I associated it only with the foreground/background tool next to it. My takeaway: I need to read/play/explore more...

Bill

--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html

[Michael Schmid]

 >> my 3 cents: I think it would be even nicer to have an area a
 >> few pixels wide for picking black or white.

On 31.08.20 19:22, Gabriel Landini wrote:
 > Isn't that what the Reset "region" (a small white and black
rectangle) does?

Yes and no - the "Reset" sets white foreground AND black background.

Michael

--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html

[Herbie]

""Reset" sets white foreground AND black background."

...and that's what should be accepted as default because having other
conventions in parallel is a reason for never ending problems. I judge
it being one of the essential short-comings of ImageJ that is difficult
to remedy without breaking existing code.

Just my two Euro-cents

Herbie

::::::::::::::::::::::::::::::::::::::::::::
Am 01.09.20 um 11:28 schrieb Michael Schmid:

>  >> my 3 cents: I think it would be even nicer to have an area a
>  >> few pixels wide for picking black or white.
>
> On 31.08.20 19:22, Gabriel Landini wrote:
>  > Isn't that what the Reset "region" (a small white and black
> rectangle) does?
>
> Yes and no - the "Reset" sets white foreground AND black background.
>
> Michael
>
> --
> ImageJ mailing list: http://imagej.nih.gov/ij/list.html

--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html

[Wayne Rasband-2]

> On Aug 31, 2020, at 9:45 AM, Bill Christens-Barry <[hidden email]> wrote:
>
> Hi,
>
> I've noticed what I think is a changed behavior of the Color Picker concerning its leftmost section that displays as a grayscale gradient. In that section, low values are displayed as dark at the top and high values displayed as light at the bottom.
>
> In macOS 10.14.6 (Mojave), to pick the lowest value (0) you must pick just beneath the highest value in the gradient. The lowest value that can be picked at the top of the gradient is not 0 but is 1.

I am not able to reproduce this problem but the Color Picker in the ImageJ 1.53e12 daily build adds a 2 pixel high row of black pixels at the top of the canvas that makes it easier to pick black. You can also click on the “reset” icon in the lower right corner to set foreground to white and background to black.

-wayne

> This behavior also holds for different image types and how their values are displayed numerically. Moreover, if you click in the regions above or beneath the gradient near its lowest or highest values, the lowest non-zero value (1) or the highest value (255), respectively, is selected.
>
> This seems counterintuitive; would it be better to have the lowest value be at the top of the gradient region? I wonder if this might be a regression, perhaps due to a non-zero based 8-bit index rollover (255 + 1 = 0).
>
> On Windows 10 this behavior is slightly different.
>
> Bill

--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html