Chroma - a laser color blender

Having put a few more hours into this little project I feel it has reached a point where it might be worthy of its own thread.

What is Chroma? Well, it is a tool for calculating the chromatic coordinates of mixtures of monochromatic light and displaying them in a CIE xy chromaticity diagram. Simply put, it answers the question: If I mix these lasers, what color will I get?

It should be a very useful tool to consult when designing and building laser projectors, especially if lasers of more exotic wavelengths for which there is a lack of guidelines are involved. Systems of different wavelengths and powers can easily be evaluated for luminance and white balance.

Capabilities of the program:

• Calculate the CIE xyY coordinates for a mixture of up to 100 monochromatic sources using the Judd-Vos (1975) color matching functions.
• Display the chromatic coordinates and gamut of the system in a zoomable CIE xy chromaticity diagram in sRGB using D65 as the white point.
• Print or save diagrams in a large number of formats (including JPEG, PNG, PDF and Illustrator).

Limitations:

• Wavelengths are limited to 380 to 825 nm with 1 nm resolution (spline interpolated from 5 nm).
• The chromaticity diagram is converted to sRGB, which means that most colors are outside of its gamut (and thus not correctly displayed). Keep in mind that most computer displays can't even display the whole sRGB gamut.
• The calculated chromatic coordinates are only completely accurate for a beam reflected off a perfectly white surface. Viewing a beam in particle-free air will yield a different color due to strongly wavelength dependent Rayleigh scattering. In fog or haze the deviation will be far smaller as Mie scattering is only weakly wavelength dependent.
• Requires MATLAB Compiler Runtime (MCR) 7.9 to run.

Installation instructions:
If you have MCR 7.9 installed already, you only need the small executable. In any other case you will need the large file (250 MB), which contains an installer for MCR 7.9. Download the required executable from here to the directory you wish to run it from. Run the exe you downloaded (and follow the installation instructions if you need to install MCR).

I have rewritten a significant portion of the code since the previous release. In particular the code for the colors of the chromaticity diagram has been completely rewritten to produce a far more accurate diagram.

I have attached a screenshot so you can easily see how the program has improved.

Download mirror:
http://tocket.mine.nu/Chroma/

I can't test at the moment Tocket as I'm half way through a large video render but will do so ASAP.

looks like this will do its little revolution in our world of DIY laser combining :)

great job you did there

is this matlab, or python matplotlib? I can recognize GTK2 or wxwidgets style

Thank you. It is written in MATLAB. It is a very nice language for this kind of stuff, but unfortunately the standalone applications require the huge MCR engine to run.

It would be nice to have it ported to another language, but I fear that it would be quite a large project to retain all the functionality in doing so without spending weeks programming. Besides, the only language I can program GUIs for is MATLAB.

Hi Tocket,

New version runs OK. Like the zoom feature in the pop up window.

Not had time to play yet.

Found 2 slight issues though:

1. CMD prompt window stays open whilst Chroma is running

2. It creates a Microsoft Office Table File Shortcut on the desktop called ciexyzbg. If you delete this shortcut (which only takes you to an Office info page) then Chroma goes through 1 st run procedure again.

@tocket:

there is in fact a free open-source alternative for matlab: the pylab/scipy/numpy/matplotlib libraries for python

all are free and python is a really nice, simple and easy language, with a really fast learning

I use matlab and python/matplotlib at work to render satellite maps (gases concentrations and the like), and I have to say that I'm quite only making my figures with matplotlib now as they are nicer to the eye (and because it is open source)

python also allows to compile your code and make EXEs to redistribute them, or either use the script with python as an interpreter

GUIs are easy too, with access to many interface types

the most interesting thing is that the code would be cross-platform

I actually have some programming experience in Python as the course in applied programming I took was given in that language. I know for example string handling is excellent in python, but MATLAB has always been my choice for programming involving maths. Still, python sounds interesting if it can easily do what I want.

Do you have any recommended literature for python GUI programming?

Quote:

---

Originally Posted by White-Light

1. CMD prompt window stays open whilst Chroma is running

---

Well, some useful information (as well as any errors) is printed in it. I have no idea how to get rid of it, but I think it's nice. :)

Quote:

---

Originally Posted by White-Light

2. It creates a Microsoft Office Table File Shortcut on the desktop called ciexyzbg. If you delete this shortcut (which only takes you to an Office info page) then Chroma goes through 1 st run procedure again.

---

It is actually a binary MATLAB formatted file containing the background image which is rendered at first run. Since it takes a while to generate it I thought this solution would be the best. It is also possible to embed it in the exe, but that would make it significantly larger. Anyway, this file is created in whatever directory Chroma is in, so just put the exe somewhere else and create a link to it on the desktop instead.

That software is really quite good! :D
I've been playing around with a few options to see what sort of powers I'd need as my build progresses and attached a few screenshots.

Chroma1.jpg shows what I'll have initally. Look how much of the green/blue spectrum I'm missing! :eek:

Chroma2.jpg shows when I add 150mW of 473nm (but I have to bump the 660nm red upto 0.9W! :eek:)
So I then swapped that out for 250mW of 635nm and it's perfect :D (Chroma3.jpg)
Sad thing is, my green is 300mW, and will have to be throttled right down (but I knew that when I bought it - good for non-balanced things I guess then!)

Tocket: Fantastic piece of software. Dead useful for planning ;)
Just one thing: Would it be possible to get it to solve the ideal colour powers if say you fix one colour variable. e.g. have a fixed amount of 473nm and out pop the appropriate mW power values for the other wavelengths for that perfect white? *retardo

Ta,
Dan

@tocket:

the matplotlib library has in fact all which is needed to plot everything like matlab

it is to be installed with the scipy/numpy libraries which have everything needed to do scientific computation

the two are really nice, and really close to matlab language

for python GUI programming, I cannot point some lectures to you, but rather recommand what you could find on google about python with GTK+ or python with WxWidgets as both the subjects are well discussed in the open-source community

also, compiling python to exe, as you surely know, would intensely reduce the size of your resulting package (especially if the user already have python installed :p )

Something that deserves to be mentioned more than once is that the power you enter into the program needs to be the actual power coming out of your projector, after all optics!

You can enter mathematical expressions into the fields in the program however, so if you're planning your build you could for example write 0.95*0.95*0.98*0.98*0.98*0.5 for a 500 mW green passing through 2 dichros (with 5% loss for each), bouncing off 2 98% reflective galvo mirrors and finally passing through an AR coated piece of glass. Perhaps I make the power input field wider to facilitate this.

Quote:

---

Originally Posted by danielbriggs

Fantastic piece of software. Dead useful for planning ;)
Just one thing: Would it be possible to get it to solve the ideal colour powers if say you fix one colour variable. e.g. have a fixed amount of 473nm and out pop the appropriate mW power values for the other wavelengths for that perfect white? *retardo

---

Glad to hear that you like it. I had already written the code to optimize mixtures towards a given white point, but I have not yet figured out how to implement it into the GUI. It can easily be made to work with mixtures of only 3 lasers, but I want to integrate it nicely into the GUI, which supports up to 100 lasers. The solution to these optimization problems can either be a single number, multidimensional function or non-existant, depending on what parameters you supply. I have summarized the solutions for commonly used wavelengths in this table though:
http://tocket.mine.nu/laser/redtable.gif

It was first posted in another thread, where it's also explained how to use it.

Where can I get this? I don't see a link

The word "here" is linked in the original post, but since the theme of the boards changed links are the same color as normal text, which makes them very hard to see. I'll update the post.

Here's the link again anyway:
http://blue.orgchem.kth.se/Chroma/distrib/

Chroma.exe doesn't run and the other file is 256mb big...

Yes, look at the installation instructions in the first post.

The server is very fast though; I download the file in just 40 seconds.

Quote:

---

Originally Posted by lamborgini8

Chroma.exe doesn't run and the other file is 256mb big...

---

If it doesn't run you need the big file as it contains the Matlab software re distributable needed to run it.

doc here at pl suggested the croma also zoof suggested it at the harlem meet and i just want to say that is really awesome. thanks tocket you with out a doubt Mr Color at PL:D

I'm glad to see my program come to use.

I have done some thinking about the algorithms used and so far I can only come to the conclusion that they are good the way they are. It uses more recent color matching functions than many professional applications, so it should be quite accurate unless I made an error in the implementation.

As for being Mr Color... well, I do work with dyes in real life, so that may be an accurate description. ;)

The download link is broken

Fixed. Stupid kernel updates sometimes disable httpd on my workstation.

Quote:

---

Originally Posted by tocket

Yes, look at the installation instructions in the first post.

The server is very fast though; I download the file in just 40 seconds.

---

smokin' along at 51.2 KB/sec...

Very nice software! thanks for sharing.

It motivated me to add 200mW 405 to my projector to get a nicer white.

Thanks. Keep in mind 405 nm makes the beam look a bit weird as it's hard to focus on. At least for me. It does have a significant impact on the color though. Mixing deep red and violet gives some amazing colors.

Quote:

---

Originally Posted by laserist

smokin' along at 51.2 KB/sec...

---

Hehe, guessing you don't live in Europe. I get 5-6 MB/s, but I live only 2 km from the server.

The speed is very ok! I really do love my 100/100Mbit fiber connection.
Thanks Tocket for this excellent and interesting piece of software.
http://www.xs4all.nl/%7Ehobbybob/Fas...20download.jpg

Quote:

---

Originally Posted by tocket

Thanks. Keep in mind 405 nm makes the beam look a bit weird as it's hard to focus on. At least for me.

---

How thick is your 405 beam?, if its a normal 5mm beam it makes sense becouse it fatter than your other beams.

I have replaced my original aixis lenses with 5mm beam for special lenses with 1mm beam diameter and 0,8mrad and this made a huge improvement for my red and color mixing

Is the beam diameter not an factor that's relevant to the color table?

Quote:

---

Originally Posted by hobbybob

The speed is very ok! I really do love my 100/100Mbit fiber connection.
Thanks Tocket for this excellent and interesting piece of software.

---

Nice to see that people down on the continent can enjoy the high speed as well. Wish I had 100/100 myself, 50/10 (cable) is the best I can get.

Quote:

---

Originally Posted by mccarrot

How thick is your 405 beam?, if its a normal 5mm beam it makes sense becouse it fatter than your other beams.

I have replaced my original aixis lenses with 5mm beam for special lenses with 1mm beam diameter and 0,8mrad and this made a huge improvement for my red and color mixing

---

I only made a temporary setup to test 405 and 532 nm mixing. You're right though, the beams didn't quite match. Especially since it was a low power green laser with a tight beam. Those are some very impressive specs for a direct injection laser though. Care to share what lens it is you're using?

Regardless, I have trouble focusing on a 405 nm beam alone as well, so I suspect the problem is in fact in the wavelength. Or rather, a combination of the nearsightedness I seem to have developed lately and the wavelength.

Quote:

---

Originally Posted by cyberb0b

Is the beam diameter not an factor that's relevant to the color table?

---

It is relevant, yes. I chose not to include it and many other factors that also may or may not have an effect for the sake of simplicity. A perfect beam is assumed. In most cases that's close enough.

It is of course possible to calculate the impact of different beam geometries, distributions and alignments separately. Then enter the spectral power distribution for different positions in the beam in my application to get a more accurate color.

The lenses are aspherical glass lenses available with broadband and red AR coating.

I have purchased them from Pritzler.

They are about 50 euro/pc :eek:, but compared to the Rötner Lens27 it got better specs.

Lens27 got 1,5mRad :(
this lens 0,8mrad :cool:
and only 1% loss on the optics :) and it comes with the complete diode mount.
the beam will be 1mm x 2mm so very good for stacking 2 beams on top of each other.

Could it be possible to add some more functionality to the calculated cumulative luminous flux ?
I can imagine that the term luminous flux is not that clear for many users.
You could add a functionality that that says:

This cumulative luminous flux of 7698 lm equals the brightness of a single [enter wavelength here]nm beam.

For many (starting) laserists who want to know how their rgb projector compares to a single 532 projector, this might be a valuable addition.

That's pretty easy to add, but I can only compare luminance, not brightness. Two beams of equal luminance don't necessarily have the same brightness. It's a bit complicated. I can certainly imagine that people have trouble understanding these weird units and concepts. Even I do.

Do you think the user should be able to choose what wavelength to compare it against or should I simply set it to 532 nm?

532 would make sense, but then again it would be useful to be able compare say 671nm with 445 or 405 for instance.

Comparing against 532nm would suffice for 90% of the time.

In a wider perspective, right now chroma is a tool for calculating theoretical ideal white balance. Not every bodies cup of tea.
If you add selectable wavelength comparison, it would widen its use to all kinds of chromatic laser issues.
If you want more people to discover the possibilities of chroma, adding this functionality might help it reach more laserists.

Real nice software.. Good job :D ..

This software is really helpful! Now all I need is a power meter :)

BTW: I noticed that using the 660 + 532 + 405 (is really popular here) yields a not so great blue range (as others have very well pointed out). However if you add 488 and even just a little such a 1/10 the power of the 660 and 405 that you get a really nice blue coverage. I did see some 488nm solid state lasers for sale too. I used ~100 mW for the 660 ~150 mW and 532 ~25mW. Now if we can just find a 510nm laser that would be totally awesome.

just get a small argon head, you'll have 515 and 488 ;)

Yes that would appear to the best answer... however the size/weight/power and modulation requirements are going against the grain of the rest of the projector. A nice little omnichrome and a PCAOM would fit the bill... Direct modulated solid state would be better... although the price on those is high too.

Anyone out there already setup similar to this?

http://blue.orgchem.kth.se/Chroma/distrib/
don't open it. have anybody can test it.

It's up again. This shit happens way too often with linux...

Quote:

---

Originally Posted by laserpic

don't open it. have anybody can test it.

---

It's been sat on many of our computers for a long time and there are no reported viruses or spyware.

It's installed on 2 of my home machines and on my work PC (with mega security) with no warnings.

Ian

No problems here either... Program works well and does exactly what it is supposed to do...

Quote:

---

Originally Posted by laserpic

http://blue.orgchem.kth.se/Chroma/distrib/
don't open it. have anybody can test it.

---

I'm not that sure he means we should not open it.

Laserpic could just as well mean:

Not able to open the link, could anybody check it.
It sounds very chinenglish to me.

Quote:

---

Originally Posted by -bart-

I'm not that sure he means we should not open it.

Laserpic could just as well mean:

Not able to open the link, could anybody check it.
It sounds very chinenglish to me.

---

yes, my English is poor. but go with trend of chinenglish.

Thanks tocket for up it again.

Every good program needs an icon :p so.. :)

Sorry its in a rar.. but the board doesnt alow me to upload a icon file. :)

Quote:

---

Originally Posted by -bart-

Not able to open the link, could anybody check it.
It sounds very chinenglish to me.

---

Same problem here, DNS is good but the server seems to be down.

Would love to give it a test run, but of course then I need the prog and the MCR from somewhere! Mirrors! Mirrors! ;)

it is probably just me but I cant get it to work.

Scenario one - 100mw 532 250mw 405 250 mw 642

Scenario two- 100mw532 250mw 405 125mw 642 125mw 660

Scenario three- 100mw 532 250mw 405 250mw642 250mw 660

which scenario is best?

Quote:

---

Originally Posted by Stoney3K

Same problem here, DNS is good but the server seems to be down.

Would love to give it a test run, but of course then I need the prog and the MCR from somewhere! Mirrors! Mirrors! ;)

---

Sorry about that, it was time for updates again and I forgot to type in sudo httpd as usual. Mirrors seems a bit overkill, but I guess I could put it up at home as well.

Figure 1 = scenario 1 etc

added scenario 4 - 100mW 532, 125mW 642, 500mW 405




http://img687.imageshack.us/img687/2724/chromalarge.jpg

Thanks Doc. Appears I need to bump my green up to get use of my dual 642's and dual 405's if I am reading charts correctly.

DOH!!

I had a problem with the program at first, but got it running smooth now! It's so fun to play with and imagine what kind of laser I could I have. My girlfriend could care less though.

Is it just me, or does the MCR 7.9 not work on Vista 32/64 or W7 32/64?

--DDL