Intelligent Discussion of Desired Wavelength Software

Let's Have a Meaningful Discussion, Shall We?

As of late, there have been a plethora of poll threads regarding "true" wavelengths of particular colors. The fad has produced a myriad of fluff threads that seem to annoy many members. Many of us thought it would end. When it did not, we begged why.

The answer, it seems, comes down to a dissatisfaction with the current "Wavelength to RGB / HEX Converter Tool" made by our very own RHD. The issue stems from perceived inaccuracies in the program's output along a few key wavelengths.

Questions have been raised regarding facilitating a new conversion tool that would take into account things like:

• Beam Brightness
• Room Lighting
• Monitor/Display Type
• Monitor Settings/Calibration
• Beam vs Dot Appearance

Personally, I consider the current tool to be an amazing tool, one which I use to code my signature. If the forum members want changes, however, we should create an intelligent discussion on the topic and not a whole lot of wasted breath with polls of little to no consequence. Change is instigated when the desire for change has been made known and steps have been taken to troubleshoot the issue.

Troubleshoot the issue.

Please do not use this thread as a flame wall either for or against. This needs to remain level headed. Think objectively. If a member, in possession of specialized programming skills, tells you something is not possible, do not just mouth off and be a jackass. Pursue results in a competent manner and who knows? Maybe one day we will have a highly detailed and accurate conversion tool. At the very least we could create an interesting website, which once setup by the end user(individually), would give them the ability to explore wavelengths in ways they might never otherwise be able to due to economic constraints.

Might I suggest creating a list of everyday items with known spectroscopy? If we know the wavelength of a green stoplight is always 5xxnm for example, we can use that as a landmark for additional wavelengths.

Anyway, enough of my babbling. Converse!

Thanks,
Isaac

wannaburnstuff said:

Personally, I consider the current tool to be an amazing tool, one which I use to code my signature.

Click to expand...

Awesome program:wave:! Didn't know about it until now:yabbem: Now I no longer need pure wavelenght pools.
Edit: Now I understand why you said 650nm is pretty orange in comparasion with 690-700nm.

The calculator doesn't do it justice lol. It actually gives the exact same values from 650 all the way up to like 730 or something. This is one area that could be explored.

Yea... only after ~710 becomes to be darker.

Do you have access to the numerical output of the calculator, so as to see if the program is making delineations that are not being represented on the screen?

The limiting factor will probably be the screen resolutions of color.

For example, there may not BE a color CODE that represents a particular nm value....the moral equivalent of there being no NAME for that wavelength. If there's no name for the wavelength, the program has no way to tell the monitor what to display, so it displays whatever is closest that DOES have a name.

If the codes exist, but are not in the look up table of the program for example, then, they could be added to fix the problem.

If the codes don't exist, then, someone would have to create them, and, add them to the database, and so forth.


The other option is that our eyes simply don't have the color resolution in some wavelengths needed to tell the difference between some colors. This is typically manifested by seeing a color as getting lighter or darker, as opposed to being perceived as a different color.

Useful reference on "Color", code examples and information links:
Color Science

Teej said:

Do you have access to the numerical output of the calculator, so as to see if the program is making delineations that are not being represented on the screen?

The limiting factor will probably be the screen resolutions of color.

For example, there may not BE a color CODE that represents a particular nm value....the moral equivalent of there being no NAME for that wavelength. If there's no name for the wavelength, the program has no way to tell the monitor what to display, so it displays whatever is closest that DOES have a name.

If the codes exist, but are not in the look up table of the program for example, then, they could be added to fix the problem.

If the codes don't exist, then, someone would have to create them, and, add them to the database, and so forth.


The other option is that our eyes simply don't have the color resolution in some wavelengths needed to tell the difference between some colors. This is typically manifested by seeing a color as getting lighter or darker, as opposed to being perceived as a different color.

Click to expand...

At the top of the page it says: "221647 wavelenghts served to date". I don't know what exactly this means, but I saw it isn't any difference between ?nm and (?nm +10nm) at the red range. You can take it like I did: 740nm-750nm; 750nm-760nm; 760nm-770nm; and see ~no difference between them but after comparing the first with the last (740nm-770nm) you can see the difference. I think the range that must be explored is 770nm-800nm. 800nm is a wavelenght that is pretty visible but on that program doesn't apear.

BurningLaserGuy said:

At the top of the page it says: "221647 wavelenghts served to date". I don't know what exactly this means, but I saw it isn't any difference between ?nm and (?nm +10nm) at the red range. You can take it like I did: 740nm-750nm; 750nm-760nm; 760nm-770nm; and see ~no difference between them but after comparing the first with the last (740nm-770nm) you can see the difference. I think the range that must be explored is 770nm-800nm. 800nm is a wavelenght that is pretty visible but on that program doesn't apear.

Click to expand...

Right, but do we know if it doesn't appear because there are no names for those wavelengths, no way to tell a monitor what to make...? Or, is it because the monitor makes it, but our color vision can't tell the difference?

Typically, if a color is perceived as getting darker or lighter when the wavelength is supposed to be changing, it means the monitor IS displaying something different, but, we can't tell the difference.

wannaburnstuff said:

Let's Have a Meaningful Discussion, Shall We?

As of late, there have been a plethora of poll threads regarding "true" wavelengths of particular colors. The fad has produced a myriad of fluff threads that seem to annoy many members. Many of us thought it would end. When it did not, we begged why.

The answer, it seems, comes down to a dissatisfaction with the current "Wavelength to RGB / HEX Converter Tool" made by our very own RHD. The issue stems from perceived inaccuracies in the program's output along a few key wavelengths.

Questions have been raised regarding facilitating a new conversion tool that would take into account things like:

• Beam Brightness
• Room Lighting
• Monitor/Display Type
• Monitor Settings/Calibration
• Beam vs Dot Appearance

Personally, I consider the current tool to be an amazing tool, one which I use to code my signature. If the forum members want changes, however, we should create an intelligent discussion on the topic and not a whole lot of wasted breath with polls of little to no consequence. Change is instigated when the desire for change has been made known and steps have been taken to troubleshoot the issue.

Troubleshoot the issue.

Please do not use this thread as a flame wall either for or against. This needs to remain level headed. Think objectively. If a member, in possession of specialized programming skills, tells you something is not possible, do not just mouth off and be a jackass. Pursue results in a competent manner and who knows? Maybe one day we will have a highly detailed and accurate conversion tool. At the very least we could create an interesting website, which once setup by the end user(individually), would give them the ability to explore wavelengths in ways they might never otherwise be able to due to economic constraints.

Might I suggest creating a list of everyday items with known spectroscopy? If we know the wavelength of a green stoplight is always 5xxnm for example, we can use that as a landmark for additional wavelengths.

Anyway, enough of my babbling. Converse!

Thanks,
Isaac

Click to expand...

:yabbmad: The original polls actually used the word "pure," a concept fundamental to color vision research as well as to the world of RGB. The word "true" was only used as a veiled insult to suggest that the whole concept is so subjective that it smacks or religion or ideology. :cryyy: Rather than a fad, I like to think of it as a slow, thoughtful approach to a very complex problem. Instead of trying to tackle it in one thread, better to focus on various aspects of the problem under the rubric of how various laser pointer colors look to various people. As for people hoping it would end, I thought this forum was supposed to be about laser pointers. While some people want more power, others want to build their own lasers, many enjoy them for the colors and for the fact that they come with wavelength numbers. We even have people here with digital spectrometers who can give us some insights about how reliable the numbers are.

:can: "Troubleshooting the issue" is great if you're an engineer tackling a problem within your field of expertise. So far they've spent billions of dollars on color vision research, and have made about as much progress as finding the cure for cancer. Which is to say, they've made amazing progress, but huge problems still remain.

As for everyday items of known spectroscopy, I actually spent a lot of time on that. Unfortunately, I ran into complications. For example, amber traffic signals used to be 594, but the new ones are 590. The old 3M green ones were 507 (a wavelength known as "traffic light green") but then 3M left the business to Dialight, which set the standard at 505. Now they're going 500 because of new international standards. Worse yet, all these colors look different depending on how far away you are, the time of day, cloud cover, and what other lights are in your field of view (like red tail lights). I still think it's good to collect wavelengths like that, just for educating yourself, but not as the basis for this project.

Instead, I think they already have vision studies that answer the most basic questions about colors and wavelengths. In particular, they've got red, green, blue, yellow, and cyan figured out, plus an educated guess that violet is in the low 400s. Usually these studies are done by shining monochromatic light into 10-minute dark-adapted eyes from straight ahead, covering either 1 or 2 degrees of the fovea.

The next step to to get a list of things that can affect how we see color, and figure out which ones are important enough to worry about. For example, men and women see color differently by an average of 2.2 nm. Not a big difference. Brightness, on the other hand, makes a huge difference. Then we need some data on the most common standards like sRGB or Adobe RGB, starting with finding out how many people have which ones.

That's it for now. I'll try to post some links that might be helpful. :wave:

http://www.efg2.com/Lab/ScienceAndEngineering/Spectra.htm

I used this program back when I had a projector, and it worked pretty well in my experience.

Chroma - a laser color blender

This is the code that it runs on (converted to something other than FORTRAN of course). You should be able to look at this code and understand that RGB blending - it's fairly straightforward.

Spectra Code

221,647 means exactly that - 221,647 wavelengths (not necessarily different ones) have been checked with the tool (and counting). It's a counter.

Grats on your 7000th post rhd! By the way you will never receive enough rep for the work you have done for the forum. If only it would let me.

Next three people to rep my man rhd will get repped be me. Hit him up

wannaburnstuff said:

Grats on your 7000th post rhd! By the way you will never receive enough rep for the work you have done for the forum. If only it would let me.

Next three people to rep my man rhd will get repped be me. Hit him up

Click to expand...

Yes, a celebration is in order.

Wow 7,000 posts ----that represents an amazing amount of time and effort..

+ Rep to you!

rhd said:

This is the code that it runs on (converted to something other than FORTRAN of course). You should be able to look at this code and understand that RGB blending - it's fairly straightforward.

Spectra Code

221,647 means exactly that - 221,647 wavelengths (not necessarily different ones) have been checked with the tool (and counting). It's a counter.

Click to expand...

Thanks for the link. I already had a look the site a year or two ago. The problem is, this can't be done with just RGB blending, in part because the human visual system is not exactly an RGB system. It has an RGB system inside of it, but it's way more complicated than that. Going by simple formulas is exactly what I'm trying to avoid.

zyxwv99 said:

Thanks for the link. I already had a look the site a year or two ago. The problem is, this can't be done with just RGB blending, in part because the human visual system is not exactly an RGB system. It has an RGB system inside of it, but it's way more complicated than that. Going by simple formulas is exactly what I'm trying to avoid.

Click to expand...

You refer to wavelength SOFTWARE. I hate to be a nay sayer here, but software running on your machine is going to display its output in RGB values. You can adjust the location in the chain where the relevant conversion occurs, but you can't get away from that conversation occurring.

Further, you're citing a whole bunch of things that DO NOT change the color of a wavelength:

Beam Brightness
Room Lighting
Beam vs Dot Appearance

These factors do impact color:

Monitor/Display Type
Monitor Settings/Calibration

There are significant acknowledged variabilities in terms of color representation from monitor to monitor.However, RGB codes are INHERENTLY STATIC in that they're fixed at time of content authorship. It's axiomatically impossible to generate static codes that are better or worse at being variable.

The current hex converter is based on information from Dan Bruton's website, which in turn is based on CIE 1931. The following diagram illustrates how it works. (Warning: this my own crude diagram.)



The inner triangle represents the sRGB color space, which is what most people have on their computers and handheld devices. To find out the wavelengths of the color dots on such a computer, draw a straight line from the white point marked D65 through any corner of the triangle, then keep going until you hit the edge of the horseshoe-shaped thing with wavelength numbers on it. For sRGB you hit 611nm. Adobe RGB has a bigger triangle, so the wavelengths of the color dots are different.

Dan Bruton had to make an educated guess about what kind of computer you have. I would like to offer two options: sRGB and Adobe RGB. Maybe in the future also some kind of Apple RGB. The idea is that people with different monitors would get different hex codes tweaked for their own monitor. Incidentally, Microsoft Windows already does something related with "rendering intent." If I'm looking at a photograph on my computer, Windows automatically put in the "Perceptual" gamma correction. For business graphics like bar charts, it gives you the "Saturation" gamma. Windows also has two more: "Relative colorimetric" and "Absolute colorimetric." All four can appear on your screen at the same time. Also, who said hex code has to be static? I think it should be generated on the fly, since there may be too many variables.

Meanwhile, the CIE chart has problems. It's supposed to represent a "grand unified theory" of how color vision works, but since 1931 they've found out it's a little more complicated than that. I don't propose to fix all the problems because this is a supposed to be a simple project. Besides, we're just dealing with lasers, which are light-emitting sources with pure spectral colors. That greatly simplifies the problem.

One issue that I would like to address is brightness. You mentioned "Beam Brightness, Room Lighting, Beam vs Dot Appearance." The following diagram illustrates the problem.



Colors shift dramatically with brightness. CIE 1931, like most vision studies, was based on people with dark-adapted eyes, in a dark room, looking into a monochromator. The typical brightness levels are pretty dim. Unfortunately, it's hard to get reliable data on "no contrast shift" which is the variant of Bezold–Brücke shift we need to use.

One way to deal with that problem is to get data on additional colors at different brightness levels, just for error-checking purposes. In color-vision research they use the four "unique" colors (red, green, blue, and yellow) plus some "binary" colors (even mixtures of two colors) such as cyan (b/g), orange (r/y), yellow-green, etc.