i do gradient beetween 0,0,0 and 0,0,1 (rgb values) suprising things?

[MxHeppa]

i do gradient beetween 0,0,0 and 0,0,1 (rgb values) suprising things? i flatten such box (which is whole document)
and i dont use nosie setting. and after this enchament and  auto contrast makes stuff other side is yellow (255,255,0) and another is white. and there in beetween parts noise. imagined i get only two diffeent coloured bars not like this. i dont know if this what i should expect.
 

[Dan C]

Hi MxHeppa

My apologies, I'm struggling to follow the workflow you've described. Could you please provide a screen recording showing the steps you've taken? As a gradient using your RGB values would be black fading to blue (assuming 1=255).

[MxHeppa]

On 10/29/2019 at 4:38 PM, Dan C said:

Hi MxHeppa

My apologies, I'm struggling to follow the workflow you've described. Could you please provide a screen recording showing the steps you've taken? As a gradient using your RGB values would be black fading to blue (assuming 1=255).

1 is 1. i mean two colours what nearest possible in 8bit rgb values. there is no inbeetween colours for these two. i dont know how do screen recording (windows 10 pro).

[MxHeppa]

On 10/29/2019 at 4:38 PM, Dan C said:

Hi MxHeppa

My apologies, I'm struggling to follow the workflow you've described. Could you please provide a screen recording showing the steps you've taken? As a gradient using your RGB values would be black fading to blue (assuming 1=255).

my gradient is beetween black and almost black blue.(one set blue direction).

 

 

[Dan C]

I'm sorry, I'm still not sure what colour values you're using. A gradient with RGB 0,0,0 to RGB 0,0,1 would be black fading to a slightly lighter black (which is almost inconceivably different).

If your values are RBG 0,0,0 to RGB 0,0,100 then you would see a gradient as follows - 

Is this the gradient (or similar) that you're starting with?

I've made a quick screen recording following the workflow you've described - are these the steps you're taking?

[Old Bruce]

Don't do the Auto Levels first, just the Auto Contrast after flattening, and yes use the 000, 000, 001 and 000, 000, 000 for the colour stops.

8 hours ago, Dan C said:

Is this the gradient (or similar) that you're starting with?

I've made a quick screen recording following the workflow you've described - are these the steps you're taking?

Wow, I have been messing about just guessing at what could do this. And then I re-read the original post more closely and yes, the recipe does do this. A gradient from almost pure black, just a 001 value for the blue, to pure black will result in the yellow to white when the Auto Contrast is used. I would like to see the algorithm used for the Auto Contrast which results in this.

On 10/29/2019 at 5:24 AM, MxHeppa said:

i do gradient beetween 0,0,0 and 0,0,1 (rgb values) suprising things? i flatten such box (which is whole document)
and i dont use nosie setting. and after this enchament and  auto contrast makes stuff other side is yellow (255,255,0) and another is white. and there in beetween parts noise. imagined i get only two diffeent coloured bars not like this. i dont know if this what i should expect.
 

[Dan C]

Thanks Bruce, I too have found this causes strange results - selecting Auto Contrast again from the white/yellow image changes this to black/blue!

I've logged this with our developers and will provide any updates here as available.

On 11/5/2019 at 12:43 AM, Old Bruce said:

I would like to see the algorithm used for the Auto Contrast which results in this.

Unfortunately I don't believe this is something I can share publicly - however if given permission I will post as much information here as possible.

[Old Bruce]

I am only curious as to why there is not a clean break from the yellow to the white when starting with  a gradient from 0, 0, 1 to 0, 0, 0. My intuition says there should be a clean break, I am quite often very wrong.

[Dazzler]

40 minutes ago, Old Bruce said:

I am only curious as to why there is not a clean break from the yellow to the white when starting with  a gradient from 0, 0, 1 to 0, 0, 0. My intuition says there should be a clean break, I am quite often very wrong.

I imagine this is a dithering algorithm making a blend between two values that won't interpolate further?
 

[Dazzler]

Talking of dithering ... I've just found a 'dither gradients' option available in the preferences under the performance tab, but turning this off made no difference - still got the same dithered result. Maybe a bug?

[Dan C]

The 'dither gradients' option relates to how live gradients are drawn, once the gradient has been rasterised this setting is no longer in effect, so I don't believe this is related to the issue here.