200mW of red is equivalent to how much green?

[qumefox]

qumefox, given that the commonly-used LPC LOC is typically spec'd at 660nm, I am curious what equivalent 532 green mw your program gives you for that wavelength?

You'll have to wait until I get back to office the for 660 numbers unless you just want to install chroma and matlab yourself. Chroma was written by a PL member. I'm not going to install it on customer's machines and i'm onsite at a school atm.

link: tocket.mine.nu - /Chroma/distrib/

and here's the PL thread on it: Chroma - a laser color blender

I always had trouble with the cChroma package not working right, so if you use a 32bit OS, you'll have better luck downloading the matlab runtime straight from MCR and just downloading the chroma executable from the link. If you use 64bit the just grab the two files out of the x64 directory, they both work fine.

Awesome data btw.

 

[JaiNobeZ]

That's correct, but remember: Lower beam diameter means higher divergence (usually), so it's opposite after about 5 meters, then red/violet beams are thinner than green beams

Really? Interesting. I thought that difference was only fairly marginal.

Whichever way it works out, it's worth bearing in mind when comparing visibility. It's a property of wavelength and power PER UNIT AREA.

 

[ElektroFreak]

This thread got me to thinking, and then to doing some digging!

I wanted to find-out how much Rayleigh scattering varies by wavelength, so that we could adjust qumefox's #'s to account for the change in actual beam intensity based on a laser's wavelength.

Turns out a fella by the name of Lord Rayleigh (hence the name, and also why the first word should be capitalized ) figured this out way back in the 1800's!

He determined that this amount varies inversely proportional to the wavelength of the light raised to the 4th power.

So we can now calculate the difference in Rayleigh scattering between a 532nm green laser and other wavelengths, in order to determine how much we need to adjust the apparent beam intensity to correct this difference!

So based on that, here is the "approximate" difference in Rayleigh scattering between a 532nm green laser and other common laser wavelengths -

635nm - 0.49266490683388707592096813556886

650nm - 0.44873826309162844438219950281853

660nm - 0.42215388175097666707984764583813

445nm - 2.0427089195236108955687445081855

405nm - 2.9773263176444960813623876252967

But that's not the end of the story! One of the articles I came across indicated that modern references to Rayleigh scattering often forget to include an additional factor - one that also varies by wavelength!



Atmospheric Extinction

Well, it turns-out that the refractivity of air also varies with wavelength! -

Dispersion

Now normally, when a web page claims other references are wrong, I take it with a grain of salt. But this is coming from an extremely authoritative source. These folks' careers depend on making EXTREMELY accurate calculations regarding such things.

So, unlike crazy Google ATM, I'm going to assume they are a more accurate reference than Wikipedia!

According to this reference, accounting for this extra factor, the change in Rayleigh scattering actually varies by the inverse of the wavelength raised to approximately the 4.08th power across the visible spectrum.

Using these revised calculations, the approximate difference in Rayleigh scattering between a 532nm green laser and other common laser wavelengths is now -

635nm - 0.48573864900825677601320307593276

650nm - 0.44160396556074654015153702131358

660nm - 0.41493512744940499532617990819131

445nm - 2.0720995430911201003047269051986

405nm - 3.043007095689047987747722875912

So, assuming the quoted Chroma values were correct, the revised beam intensity values (including Rayleigh scattering wavelength adjustments), would be approximately -

200mw 635nm red = 23.827mw green

200mw 650nm red = 10.680mw green

And for the other examples given -

1W 445nm = 99.270mw green

600mw 405nm = 9.6077mw green



qumefox, given that the commonly-used LPC LOC is typically spec'd at 660nm, I am curious what equivalent 532 green mw your program gives you for that wavelength?

That's a great post..

Everyone treats Chroma's output like it's the gospel when it comes to color sensitivity, but in reality no matter what math we use the differences between people's eye shape, sensitivity and chemistry means that at best we get a general idea. There is no real precision possible when discussing relative visibility of various wavelengths across a whole population.

 

[LazyBeam]

I thought the real-world consensus was that 300mW of green was about equal in beam intensity to 1W 445nm?

 

[Moptsp]

That's a great post..

Everyone treats Chroma's output like it's the gospel when it comes to color sensitivity, but in reality no matter what math we use the differences between people's eye shape, sensitivity and chemistry means that at best we get a general idea. There is no real precision possible when discussing relative visibility of various wavelengths across a whole population.

Yep. +1

 

[qumefox]

I don't think chroma was ever intended for more than generalities. Otherwise it'd take Rayleigh scattering in to account. Those equations aren't hard after all. I think chroma was written solely for people doing color mixing of lasers.. i.e, building projectors, to be able to ballpark what power lasers they need to achieve whatever balance they wanted.

However I listed chroma numbers because, as stated, people are different. It's hard to describe something subjective as how you see something to someone else when comparing them. I used chroma because it's output ISN'T subjective.. It's output is correct regarding the input you give it regardless of how anyone perceives a color themselves. I also bet that if you averaged everyone's opinion on a certain mix, the averaged opinion would be close to what the equations dictate, at least for what the application outputs. Which is a diffuse reflection terminating on a white surface.

 

[seoguy]

Grix & davidgdg - Thanks for the +Rep! :thanks:

You'll have to wait until I get back to office the for 660 numbers...

Did you get a chance to get back yet?

BTW, I took a look at those links you provided - thanks!

I always had trouble with the cChroma package not working right, so if you use a 32bit OS, you'll have better luck downloading the matlab runtime straight from MCR and just downloading the chroma executable from the link.

Unfortunately I am running 32bit, so his version D/L would be problematic for me.

Also worried that a 1/4 gigabyte runtime might be more than my poor laptop could handle!

Noticed f/the PL thread that the latest version (V1.1) of his pgm ONLY works on 64bit :cryyy: -

An updated version is finished. Unfortunately, since I'm running W7 x64 with 64-bit MATLAB on this computer the new Chroma is 64-bit (and for no good reason)...

This program will only run on 64-bit Windows!

He's had offers to help compile the new version for 32bit and/or to port it to something more efficient, but has apparently declined.

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.

Yes, it is still slow to start and uses a lot of memory. That's the MCR... a bit like java. The only way to get rid of that is to port it to some other language. It might happen some day, but it's not a high priority.

Given that, and the expressed desire of other PL members to help...I am wondering if he could be convinced to make this already free pgm an open-source project? :thinking:

 

[qumefox]

Sorry. Just didn't get around to it.

Chroma 1.1 says 200mw of 660 = 13.788mw of 532.

I just realized something though. When comparing beams, we're forgetting to account for Rayleigh scattering of 532 as well. So I don't think any of the 'beam' numbers here are accurate vs a 532 'beam'. The only way they would be is if 532 was close to the 1.0 multiplier.

And the 1.0 version of chroma should still work fine on 32bit systems. I'm honestly not sure what 'improvements' there were. Nothing that I can really spot. Last time I used chroma before this time it was the 1.0 version.

 

[seoguy]

I just realized something though. When comparing beams, we're forgetting to account for Rayleigh scattering of 523 as well. So I don't think any of the 'beam' numbers here are accurate vs a 532 'beam'.

Actually, I did NOT forget - that is already taken into account.

The Rayleigh scattering #'s I posted above are not absolute values, they are the relative difference in actual beam brightness between the listed wavelengths, and a 532nm beam.

And the 1.0 version of chroma should still work fine on 32bit systems.

I thought you said that the 32bit version of the cChroma file from his D/L link here -

tocket.mine.nu - /Chroma/distrib/

Had some problems?

But after reading about the slow performance and high memory usage in that PL thread (likely at least partially caused from having to load that humongous 1/4gb runtime file! ), to be honest I was more worried that this monstrosity might by too much for my poor laptop to handle! :undecided:

Sorry. Just didn't get around to it.

Chroma 1.1 says 200mw of 660 = 13.788mw of 532.

Great! Taking Rayleigh scattering into account, we then get...

200mw 660nm red = 5.721mw green

That's what I was afraid of - an LPC beam is going to take a double-hit, due to reduced visual sensitivity and reduced Raleigh scattering at 660nm. From a visual perspective (rather than burning), even a lower-power (~80-100mw) 635nm might be a better way to go?

 

[Kevlar]

This is really useful info that I've seen MANY people ask for. Great posts guys!!!

 

[qumefox]

I thought you said that the 32bit version of the cChroma file from his D/L link here -

tocket.mine.nu - /Chroma/distrib/

Had some problems?

No, the cchroma package has the matlab runtime installer built in. that's what has the problem. the chroma.exe in the dir is all you need. But since the full package has a problem, you'll need to download the 32bit matlab runtime from MCR and install it before chroma will run.

the problem I was having was the matlab part wasn't installing right using cchroma_pkg. chroma itself works fine once matlab is actually installed correctly.

But who knows. It might work ok for you. no harm in trying to install it. This particular old xp machine i'm using might just have issues. I was using the x64 version on my laptop which has 64bit windows 7 on it to run 1.1.