Re: 1 billion colors
Re: 1 billion colors
- Subject: Re: 1 billion colors
- From: Wire ~ via colorsync-users <email@hidden>
- Date: Mon, 6 Jan 2020 15:16:03 -0800
John, thank you for the insight about "plaid". This is exactly the point.
Experiments the CIE devised are comparative. They dispensed with the
qualia—thew it out—and built a model based on population sampling. Thanks
to this model, we have a useful concept of dE. You don't want to upend this
and claim that because you have a dE you have a definitive claim on a
countable qualia. Absurd. Trivial cases violate the model. Try closing your
eyes: thought experiment done!
On Mon, Jan 6, 2020 at 3:07 PM John Gnaegy via colorsync-users <
email@hidden> wrote:
> > #SomeHugeNumberOfColors
>
> "#SomeHugeNumberOfRGBTriplets" is more accurately what I meant, if you're
> defining "color" as a user's perception of light stimuli. But then what's
> the name for what a spectrophotometer reads? Is that a "color"? Maybe it's
> a set of measured values.
>
> At any rate, you've got:
>
> 1) the number of RGB triplets (let's just stick with RGB for now) so
> 8 bit: ((2 to the power of 8) to the power of 3) = 16,777,216
> 10 bit: ((2 to the 10th) to the 3rd) = 1,073,741,824
> 12 bit: ((2 to the 12th) to the 3rd) = 68,719,476,736
> 2) the number of different measured values your spectrophotometer will
> will read when pointed at your display as it's trying to display those
> triplets. How many? It depends on both your display and your spectro.
> 3) the number of individually perceptible colors you could perceive given
> any of those large sets of inputs, for a given environment, meaning for a
> given light level and color surround. Your eye (well, your brain) adapts
> constantly for your surround.
>
> And it all seems pretty unnecessary and over the top until you're trying
> to make a nice smooth gradient between two similar colors. 10 bit vs 8 bit
> gives you four times as many bands...and although 4 is better than 1, it
> can still be obviously noticeable depending on the two colors you're trying
> to go between. Same with 12 vs 10 I'd think, 16 is better than 4 but it
> still might not be enough to get an imperceptibly smooth gradient between
> two similar colors.
>
> So is there anywhere you can see a nice, smooth, imperceptibly changing
> gradient? Turns out yeah. The sky. Go out and look at how smoothly the sky
> transitions from one color overhead to a wildly different color at the
> horizon (I mean without clouds). It's pretty amazing. And it kind of makes
> me want to never look at a display again.
>
>
> _______________________________________________
> Do not post admin requests to the list. They will be ignored.
> colorsync-users mailing list (email@hidden)
> Help/Unsubscribe/Update your Subscription:
>
>
> This email sent to email@hidden
>
_______________________________________________
Do not post admin requests to the list. They will be ignored.
colorsync-users mailing list (email@hidden)
Help/Unsubscribe/Update your Subscription:
This email sent to email@hidden