Color Vision Project

[lunchtable]

Hey Folks,

From viewing some threads in this forum, I realize that many of the users on this forum are very knowledgeable when it comes to the field of lasers; I am thus coming to you guys for some help, as I am very new to this topic.

Anyways, I am attempting to design a device for a color vision study using laser diodes. I wish to achieve pseudo-monochromatic light of red, green and blue colors. A human subject should be able to adjust the intensities of these laser diodes to produce a given color.

I'm not sure which laser diodes would best suit this application. Additionally, the human subject should be able to view the color of laser light. Is there some sort of diffuser screen or device that allows humans to accurately view the color of light emitted from a visible laser without damaging the retina?

Any help is greatly appreciated; thanks!

 

[Morgan]

An interesting project you have there.

There are available diodes to do this sort of thing, it just depends on budget really.

As to a device or diffuser screen to safely view then, as long as power outputs are kept sensible, (others may be able to help you calculate this more scientifically), you could just use an unfocussed diode. Lasers don't have to be pencil thin beams you know .

Low powered greens are pretty cheap now; reds are the same and there is a new kid on the block in blue but this maybe, (almost certainly!), overkill due to the power.

Have you fully explored the route of simple filtering to attain the required colours? You mention pseudo-monochromatic. Nothing pseudo-monochromatic about laser light! This experiment could probably be achieved more cheaply and hassle free with conventional lighting. IMO.

Others, however, may have their own opinions and will no doubt be as happy to share them as I will be to read them.

Good luck with your project!

M


(P.S. Welcome to LPF! :wave: )

 

[lunchtable]

Hi Morgan,

Thank you for both the welcome and the quick reply!

Concerning the diffuser screen versus using an un-collimated laser or light filter: the idea had come up. However, the initial thoughts were that using the three lasers (red, green and blue), a person could adjust intensities to achieve any color within the visible spectrum. I'm not sure how I would accomplish this using filters or un-focused laser light. More specifically, I'm not sure how I would "combine" these primary colors. Any suggestions?

The benefit of using a focused laser beam is that I could have the three beams cross each other onto some diffuser screen no larger than a penny at the same point, and by varying intensities of the three laser, hopefully could also produce any color light.

Thanks in advance!

 

[Morgan]

Well, if you have an unfocussed laser you can have a large area of light, say the size of a dustbin lid from a few feet away, and where the three colours overlap you will get colour mixing. The same process would also work with coloured gels over three seperate white light sources. Varying their individual intensities will also produce a colour mix on whatever surface, (probably best to use some sort of projection screen), where the light patches overlap. After all this is the principle behind presentation projectors; three different colour light sources mixing on a screen... !

I don't think you would need to combine the three light sources from a directional point of view as the combining would happen on the surface and where reflected into the eye, (bearing in mind researched safety concerns). Whichever light source you choose, I would consider using a larger target then a penny as it may be easier to view and could be considered safer as any light, (laser or not), would be spread over a larger area.

I appreciate it's always cool to use lasers, and it may turn out to be the best light source for your project, but you can go a long way to set this up with conventional sources to prove your ideas will work and then install lasers at the end if truly necessary. I'm not trying to put you off at all, just may be the way I would go, (and then claim I couldn't achieve anything significant until somebody, somewhere gets me some g*ddamn lasers!)

Helpful?

M

 

[pullbangdead]

Just throwing this out there, but I assume you've read everything from the CIE experiments? The 1931 stuff from Wright and Guild? The ever-more-complicated follow-up stuff in the 60s and 70s?

You can read their papers to see how they did their color mixing experiments, because they developed ALL the chromaticity stuff based on the same kind of experiment you describe: letting test subjects adjust intensities in order to achieve certain colors. And they did it all in the 1920s, and others have been doing it ever since.

Hopefully you've already looked at all this...if not, I recommend looking into their stuff very heavily, so as to avoid as much wheel reinvention as possible.

And just one adjustment: you can't get achieve every possible color using 3 monochromatic light sources. The human eye can still see more colors beyond everything that is possible from mixing 3 monochromatic sources. You can get close, but not quite there.

 

[lunchtable]

Thanks guys!

I will definitely look into the CIE experiments and develop a design without the use of lasers (dang, the laser idea was so cool!). But yes, I will make sure to keep you updated with my project :yh:

If I run into more design problems (which I undoubtedly will), I'm confident that someone here can help me. Thanks again!

 

[HIMNL9]

If i can suggest you a cheaper and less dangerous way (and more easy to find), use backlighting leds, or also better, those 3x4mm oval leds that are used in megascreens ..... they have a decent (also if not perfect) monochromaticity, are not dangerous as lasers, speaking of vision experiments, more cheap in proportion of power, can be drived more easily and are more resistant .....

And always as suggestion (experiment already done ) ..... two separate "chambers" with 3 of these leds that reflect on a white surface, and separate controls, can give you more details in case of sight disease (i made one for a friend that is optic doctor ..... the patient is asked to regulate the colors for get the same level of white for both the eyes, so can be seen the difference also from one eye to the other, not just the chromaticity differences ..... uhm, maybe i have to patent this )

 

[Benm]

I'd suggest using LED's too in this application. Lasers are always cool to use, but i doubt they serve much of a purpose here.

Red, green and blue power led's are widely available, and the manufacturers often show a spectral diagram in the datasheet. If they are not monochromatic enough, you could even use gels to correct that to some degree.

Also, if you use leds without any focussing optics, it wouldn't be that hard to evenly light a piece of white paper or screen from behind, with the test subject looking at its front. With some distance betweed led's and screen color blending isn't that hard to achieve with leds.

 

[lunchtable]

Wow, thanks for the help guys!

I'm thinking LEDs may be the way to go. Do you have any suggestions for a monochromatic LED?

 

[HIMNL9]

The ones for "sport megascreens" are the better ones for spectral results, as far as i know ..... i use some of the Nichia ones, in some projects, and are very bright and "color cleans" .....

As example, the NSPR336S, NSPG336S and NSPB336S serie, usually for the white balancing, i use one blue, one green and two reds for each "cell" .....

Also the SMD ones are good, if you can work with SMD components .....

Their catalog is here (huge PDF), if you want to see the specifications.

 

[Benm]

There are various suppliers of usable leds. Lumileds and Cree spring to mind besides nchia, but seoul semiconductor has some good color ranges on some of its models too.

The red arent likely to pose any problem, but for the green and blue an additional gel could be desireable since they overlap a little bit in the ~500 nm range.

For white balancing you'll need to adjust the currents to obtain a decent white, and using 2 red leds on each green and blue is a common approach since red often is the least bright in a series. You can also consider one green, two red and two blues - even with that green is probably still more bright than you need to create white.

 

[pullbangdead]

It depends on how monochromatic and how well controlled you want your sources to be. LEDs are much wider than laser diodes, which are wider than other lasers, which are wider than using monochromators.

The people doing this in the 1920s certainly didn't use any LEDs or laser diodes.

 

[Benm]

True, but there are many sources of monochromatic light. Using, for example, atomic emission lines, can result in much more accurate wavelengths produced than even solid state or dpss lasers, let alone diode lasers.

You can afford to have some spectral width with visual tests usually - unless you want to create response curves or such that are accurate to the nanometer.

 

[LSRFAQ]

I suggest you look at books by Munsell on color.

I'd remind you that if you draw wavlength points along the edge of the CIE curve, that you can see the box or triangle that defines the available colors.

Steve