forums.silverfrost.com Forum Index forums.silverfrost.com
Welcome to the Silverfrost forums
 
 FAQFAQ   SearchSearch   MemberlistMemberlist   UsergroupsUsergroups   RegisterRegister 
 ProfileProfile   Log in to check your private messagesLog in to check your private messages   Log inLog in 

10 bit per RGB in Clearwin possible ?

 
Post new topic   Reply to topic    forums.silverfrost.com Forum Index -> ClearWin+
View previous topic :: View next topic  
Author Message
DanRRight



Joined: 10 Mar 2008
Posts: 2111
Location: South Pole, Antarctica

PostPosted: Fri Nov 08, 2019 7:14 pm    Post subject: 10 bit per RGB in Clearwin possible ? Reply with quote

Currently color palette in Clearwin has 8 bits per color, or 24M colors. In 64 bit computing 24MB is considered like nothing. Even 32bit one can handle that, it is just 1% of its limit typically 2-3 and maximum 4 GB limit.

In my programs i color code the spectral lines or particles with their specific color so that with the mouse you can get the appropriate color of the screen pixel (Clearwin has special function for that for %gr or %pl. Great would be to have similar one for OpenGL %og too ).

But 24 M colors is not enough here in current complex cases involving very detailed spectra containing gazillion of lines. Scratching the head how to find the workaround with this problem for a long time and can not find one. The 10 bit color palette (1GB size) would probably ease my things. Currently all TVs can handle HDR imaging with 10 bit and cameras with 12 bit, 14 bit color depth per color. So inevitably 8 bit color depth will go away and even 1GB size become nothing if 108 Megapixel cell phone cameras are already on sale and Clearwin has to prepare to this in advance
Back to top
View user's profile Send private message
John-Silver



Joined: 30 Jul 2013
Posts: 1227
Location: Aerospace Valley

PostPosted: Fri Nov 08, 2019 9:56 pm    Post subject: Reply with quote

My guess at the reply to your question is ... No.

Obviously you need it , because you're a pioneer in the use of high-def displays.

Most people (98% ?) don't and won't ever have them until they become the 'common standard' (wh9ch depends on manufacturer's making the swithých and them becoming globally affordable.

The other question is when is that likey to be.

Here's an interesing article on 12 bit, which as I interpret it is 'not anytime soon':-

https://www.tomshardware.com/news/what-is-10-bit-color,36912.html

From my personal point of view I'm as blind as a bat anyway as far as these thngs are concerned.

What type of image are you thinking of as concerns needing 10 bit colour ? Do you have an example (which speaks a 1001 words as the sayng goes).
_________________
''Computers (HAL and MARVIN excepted) are incredibly rigid. They question nothing. Especially input data.Human beings are incredibly trusting of computers and don't check input data. Together cocking up even the simplest calculation ... Smile "
Back to top
View user's profile Send private message
PaulLaidler
Site Admin


Joined: 21 Feb 2005
Posts: 6076
Location: Salford, UK

PostPosted: Fri Nov 08, 2019 10:15 pm    Post subject: Reply with quote

In this context ClearWin+ simply reflects what is available in the Microsoft GDI and GDI+.

If and when Microsoft allow these colour depths then ClearWin+ can be adapted to pass this on to the user.
Back to top
View user's profile Send private message
DanRRight



Joined: 10 Mar 2008
Posts: 2111
Location: South Pole, Antarctica

PostPosted: Fri Nov 08, 2019 11:05 pm    Post subject: Reply with quote

Piece of spectra is here. It shows vertical bars representing position and intensity of unresolved lines which combined give you what experiment sees in yellow. Each line is unique and belong to specific atom, its specific charge, specific configuration with specific quantum numbers.


Years ago it was easy to make this plot helpful because we were restricted by computer power and memory and were not able to include a lot of such lines. Today we fill 24 million lines in seconds and it is impossible to gain any knowledge where are those lines which actually contribute the most. All now instead of being rainbow is one color soup with one element out of 100 filling the screen with its color. before I was able to put mouse cursor on this line and know everything about it. Today all lines are identified wrong way because 24 million colors is not enough to mark all lines
Back to top
View user's profile Send private message
John-Silver



Joined: 30 Jul 2013
Posts: 1227
Location: Aerospace Valley

PostPosted: Sat Nov 09, 2019 4:57 pm    Post subject: Reply with quote

there are 24 million lines in that graph ?

sorry, I'm lost as to what you're trying to achieve Dan.

I can understand you wanting to be able to distinguish between individual bars on the plot (to hover the mouse over and pop up info if I understand correctly) but even if you had 24 million bars (each with a different colour) the majoriy would be hidden 'behind' the upper ones.

Do you mean you want to allocate a single colour to each atom regardless of whether it's plotted/visible or not ??

Just one observation also, the 'yellow' 'cumulative' (if I understand correctly ?)plot in the background appears to be a peak where the bars are less dense, which is a bit counter-intuitive
_________________
''Computers (HAL and MARVIN excepted) are incredibly rigid. They question nothing. Especially input data.Human beings are incredibly trusting of computers and don't check input data. Together cocking up even the simplest calculation ... Smile "
Back to top
View user's profile Send private message
DanRRight



Joined: 10 Mar 2008
Posts: 2111
Location: South Pole, Antarctica

PostPosted: Sat Nov 09, 2019 6:38 pm    Post subject: Reply with quote

All these millions of lines are not visible if use linear scale like here, they are there in LOG scale which allows to see orders of magnitude weaker lines.

Problem is with the slowness of method of assigning color to each line when there are more lines then colors. As a result you can not afford to do that before seeing plot each time as it may take few seconds (1000 times per day) and do that only ones at the beginning for all strongest lines

And why method slow is because color-associated integer number (one number which is obtained from three RGB numbers from 0 to 255) per each color if move from red to violet do not spread consecutively. As a result I assign color randomly (within 3 RGB limits for specific ion) and then check if this color was already taken.

For linear scale it is possible to speed up the method cutting all bellow 1/1000. Still it is annoying slow. And after 100 attempts to do good algorithm I suspect it is possibly malfunctioning

Good discussing this stuff again, the new ideas always come.

This plot has three different plots in it - line (red), bars, and line with the fill (yellow). All you see was done using draw_line@ and draw_characters@. There also exist Clearwin function to place color background underneath the %gr plot or on Clearwin window
Back to top
View user's profile Send private message
wahorger



Joined: 13 Oct 2014
Posts: 654
Location: Morrison, CO, USA

PostPosted: Mon Nov 11, 2019 11:55 pm    Post subject: Reply with quote

Dan, I'm not sure I understand the issue fully. The image is hard for me to see if there are gradations either along the spectral lines, or if the changes in color are on adjacent lines. The yellow peak in the background is, of course, very obvious.

A while back, when I was working full time, I ran across the issues of sufficient colors AND the ability of the average human (and CRT and color laser printer) to distinguish adjacent colors that were only a few "counts" apart in either the R, G, or B.

While it is true that 10 or 12 (or even 16-bits) are available, the human eye has a difficult time when viewing similar colors when the colors are adjacent. Think on when you and your SO are out picking paint colors. Some paint samples are really close in color, even when viewed in the same lighting. I, for example, have a very difficult time with dark blues and blacks (specifically socks!)

With having small features of a color and a similar color adjacent, it might be impossible to distinguish them. The human eye (and brain) tend to wash them together.

In my work case, I was trying to distinguish between detector responses of 0.05 dB as I mapped the face of a specialized detector that had unique physical, optical, and mechanical features. In order to get the "zones" defined, I found that a big change of color in the step, and changing even more as the differences were greater, was visually useful. So, a totally green transition when from a light green to a brilliant green in steps of 5-10 counts from a starting value of 100 going to 255. That might step into RED for the next band. Etc.

Never used greys, whites or blacks. Hard to distinguish. But yellow, reds, greens, blues, orange, etc. could give the gradation I wanted to represent 20 dB of signal change across the face of the detector.
Back to top
View user's profile Send private message Visit poster's website
DanRRight



Joined: 10 Mar 2008
Posts: 2111
Location: South Pole, Antarctica

PostPosted: Tue Nov 12, 2019 1:25 am    Post subject: Reply with quote

Bill, this need for larger color depth is not about human, though 10 bit is definitely needed in some cases, just look at color banding examples in Google search for images. Just 1 in 256 change is very noticeable in dark homogeneous change or blue sky. It is also needed in photography to keep larger dynamic range for HDR. In my case it is computer which distinguishes color when i point my mouse on the bar line above and tells me which line out of 24 million it is.

I will think about good fast method of assigning color for lines again. Clearly i do not need to see each of 24 million lines. The method i used was too universal, hence was too slow. Just to color 1000-10000-100000 lines would be OK probably

As to what humans distinguish i like this example which gives people good psychological and philosophical lesson, this is my own invention Smile Take two piles of 30 and 31 candies or 30 and 31 coins. Only one out of 1000 people will for sure find the correct pile 10 out of 10 times. Hence ~3% of your income can be wasted, or saved, or donated on anything and will not be noticed, ok, the 1% for sure. Think about how much 1-3% is per year - pretty noticeable amount!


Last edited by DanRRight on Tue Nov 12, 2019 10:42 pm; edited 1 time in total
Back to top
View user's profile Send private message
wahorger



Joined: 13 Oct 2014
Posts: 654
Location: Morrison, CO, USA

PostPosted: Tue Nov 12, 2019 2:18 am    Post subject: Reply with quote

Dan, thanks for the explanation!
Back to top
View user's profile Send private message Visit poster's website
John-Silver



Joined: 30 Jul 2013
Posts: 1227
Location: Aerospace Valley

PostPosted: Tue Nov 12, 2019 7:24 pm    Post subject: Reply with quote

the need for many colours as I see it is for obtaining much better visualisations of say contour plots for very detailed models.

In todays 'globally socially aware environment' (in the context of computing) a very, very very pretty picture is worth a lot compared with a very pretty picture ... well as far as convincing managers are concerned who can't understand numbers :O) LOL

An anecdotal example is a coleague of mine who back in the 80's gave a presentation of some FE results to a packed room of managers.
After a whole hour of presentation he said at the end ... "and for those of you who hadn't noticed' (no one had made any comment) ' the plot of the model is of course updýside down'.
Done deliberately of course, he had a healthy disdain for managers ability to understand the technical :O)

Managers in my domain (aeroSPACE) have an unnerving propensity to believe they can actually do the work for themselves !
I wouldn't be surprised if the Boeing 737/747max fiasco doesn't emanate from decision-making 'on a whim' and cost/schedule basis, and disregard for technical expertise/advice. It wouldn't be the first time.

So, give 'em a very, very , very pretty picture to keep 'em away from the details is the rule of thumb as far as I'm concrned !
_________________
''Computers (HAL and MARVIN excepted) are incredibly rigid. They question nothing. Especially input data.Human beings are incredibly trusting of computers and don't check input data. Together cocking up even the simplest calculation ... Smile "


Last edited by John-Silver on Tue Nov 12, 2019 10:23 pm; edited 1 time in total
Back to top
View user's profile Send private message
LitusSaxonicum



Joined: 23 Aug 2005
Posts: 2049
Location: Yateley, Hants, UK

PostPosted: Tue Nov 12, 2019 9:04 pm    Post subject: Reply with quote

Blimey, the 747 has a Max that's in trouble?
Back to top
View user's profile Send private message
John-Silver



Joined: 30 Jul 2013
Posts: 1227
Location: Aerospace Valley

PostPosted: Tue Nov 12, 2019 10:22 pm    Post subject: Reply with quote

... there's no proof to the contrary now is there Wink :-O
_________________
''Computers (HAL and MARVIN excepted) are incredibly rigid. They question nothing. Especially input data.Human beings are incredibly trusting of computers and don't check input data. Together cocking up even the simplest calculation ... Smile "
Back to top
View user's profile Send private message
mecej4



Joined: 31 Oct 2006
Posts: 1205

PostPosted: Tue Nov 12, 2019 11:44 pm    Post subject: Reply with quote

Eddie's comment is about your writing 747 in place of 737, I think.
Back to top
View user's profile Send private message
LitusSaxonicum



Joined: 23 Aug 2005
Posts: 2049
Location: Yateley, Hants, UK

PostPosted: Wed Nov 13, 2019 1:38 pm    Post subject: Reply with quote

Mecej4, Of course.

John must use a touch screen and have fat fingers, methinks, as his pieces are usually full of typos. On the other hand, as he's in aerospace, he should know that 747 for 737 is inexcusable, hence why I pulled his leg, and he took it accordingly.

Perhaps the whole of the airline industry believes that 737s should have the passenger capacity of 747s, which is why there is so little legroom in the former - even in Business Class!

Eddie
Back to top
View user's profile Send private message
Display posts from previous:   
Post new topic   Reply to topic    forums.silverfrost.com Forum Index -> ClearWin+ All times are GMT + 1 Hour
Page 1 of 1

 
Jump to:  
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum


Powered by phpBB © 2001, 2005 phpBB Group