Results 1 to 9 of 9

Thread: RGB questions...

  1. #1
    Join Date
    Jul 2006
    Location
    Connecticut, USA
    Posts
    2,478

    Default RGB questions...

    hey guys (and girls if any ) got a few questions for ya...

    im in the process STILL of designing my SS RGB projector. Im getting ALOT of mixed info on an "appropriate" ratio of the colors for a good white balance.
    i am hearing more and more of 3:2:1
    (660:473:532) so, 300mw @660. 200mw @473. and 100mw @532. is this a good white balance?? or am i WAAAY off??

    also, CNI- good laser modules over all?? i am hearing nothing but good things. so, if anyone can shed a little more light (no pun intended, lol...) it would be greatly appreciated.

    TTL vs. analog- obviously analog you get infinite (or close to infinite) color possibilities, softwarware permitting. BUT, with TTL, is RGB still in the 10-20 or so color range?? anyone running an RGB strictly TTL and hace some various pictures theyd like to share?? oh, im running NML full auto by the way. (ISA version)

    Lastly-
    i have heard running 100' of ILDA off of a ISA full auto card is NOT reccomended. they have been known to burn out at these distances. is this a known issue??

    sorry for all the questions. hopefully i can return the favors with some help. im no super pro, but i do pretty good!
    thanks in advance!

  2. #2
    Join Date
    Oct 2006
    Location
    Central Florida
    Posts
    7,067

    Lightbulb

    Greetings and welcome!

    I'll take a stab at a couple of these.

    I have "heard" a good ratio is 4:2:1, but it is best to "see" what you feel is best. I am running 200:60:30 to 50 (150mw 532 dialed down), and I like the way my "white" looks. Mine is closer to 6:2:1... IT will also depend on your wavelengths also. My red is 660nm, if you are running 670nm you will want more, 635 less.

    As for CNI, from what I hear they are one of the best Chineese makers. My TTL Blue is a CNI and it is running over rated specs. My green is a Lasever and well, not running as well as the CNI. My red is a Maxyz module and probably runs and will run better and longer than the other two.

    TTL vs Analog... At first I was running a full TTL set up. You can only get 7... 8 with all blanked colors... R, G, B, Y, C, M, W, blanked. I feel you will only trully miss deep magenta/purple and orange, and to me, orange is the kicker. My Pango set up in an intro so I am only getting 12 or so colors in analog anyway. If you search around you will see some of my full TTL pics. I say full TTL, my red was analog but... you don't notice much. My Blue is still TTL, but I now have an analog Green and it makes a lot of difference to me. I got orange now.

    And I hav no idea about the cable length...

    Last edited by allthatwhichis; 09-20-2007 at 21:55.
    Love, peace, and grease,

    allthat... aka: aaron@pangolin

  3. #3
    Join Date
    Jan 2006
    Location
    Charleston, SC
    Posts
    2,147,488,509

    Cool

    Welcome to PhotonLexicon, Gottaluvlasers!

    To answer your question - a ratio of 3 : 2 : 1 will work, but you'll be low on red. Remember that you'll be using 650 nm red, which is a lot harder for the eye to see. Now, if you were using 635 nm for red, this ratio would be pretty good.

    The standard ratio that you hear quoted is 4.6 : 2.4 : 1, but that is based on 647 nm red, 488nm blue, and 512 nm green. Your red is lower in the band; it's beyond the point where the eye's sensitivity drops way off. Thus you need more red to compensate. Likewise, your blue is higher in the band, and it's about where the eye's sensitivity begins to drop off on the high end. So you also need a little more blue, though that's not nearly as crucial as the red. Green is fine because you're right at the peak of the eye's sensitivity. (Actually, the peak is in-between those two wavelengths...)

    Perfect color balance with your setup would probably be something like 5.5 : 2.5 : 1, though I'm too lazy to look up the chart and do the math to be exact.

    Having said all that, there are plenty of people running projectors with less than optimal color balance, and they look just fine. A good rule of thumb is to start with red, because it gets really expensive beyond a certain point. Buy as much red as you can afford, and then go from there. So with 300 mw of red, you probably would want about 130 mw of blue and around 60 mw of green. Personally, if it were me, I'd go up to 400 mw of red, then try to swing a 150 mw of blue. Then I'd get a 100 mw of green and call it a day. Yeah, your balance won't be perfect, but you'll be in the ball park. But you always want as much red as you can afford. (Better to be a bit shy on blue, actually.)

    When you get some time, use the forum's search function and enter "color balance" in the search field. You'll find *tons* of discussions on the topic.

    Now, about blanking: Since you're running Full Auto, you only have TTL outputs. So even if you buy analog lasers you're still only going to get 7 colors. This makes color balance a bit more tricky, since you can't easily dial back the power on the lasers. (If you buy lasers that support analog blanking you *could* build a bias circuit into each blanking line to limit the power output of the laser, but that's a good bit of work...) So with your current controller, analog blanking isn't all that important.

    Ask youself if you plan to replace the lasers or the controller first. If the controller will be out-lasting the lasers, then go with TTL and save 10% on the cost of the lasers. But if you were thinking of upgrading to a new controller (one that supported analog blanking), then go ahead and buy the lasers with analog blanking now.

    As far as running long lengths of ilda cable, several people here have run 30 meters (98 ft) with no problems. In any case, you won't "burn out" the cable or the controller. The resistance rises as the cable length increases, and eventually the loss of signal through the cable means that your projector doesn't get a clean signal. Then your scanned image quality goes to hell and you loose your blanking signals. But nothing will "burn out". It just won't work worth a damn once you get over a certain cable length. Buy *good* cables and connectors and you should be OK up to 100 ft. (Though you are pushing it at that length.)

    Adam

  4. #4
    Join Date
    Jul 2006
    Location
    Connecticut, USA
    Posts
    2,478

    Default

    Thank you VERY MUCH for your in depth and precise explanations.

    i chose to go with the 660 for my red, well #1 for obvious reasons- its A LOT cheaper!! im looking at 1+ WATT for the red. and this power in the 635 or DPSS 671 package was just too much. Also, 660, although still a diode, is it true that it proveds a MUCH better beam quality over the 635? i have heard TERRIBLE beam specs for most 635 reds. 671, obviously would be a nice beam, but WAY to up there in the red spectrum and being DPSS, WAAAYY too much $$. i am at the point where i think 660 is going to be my happy medium. i know not a perfect beam, but a pretty good beam. i know not as visible as 635, but still pretty bright. im on the right track???

    and after reading my original post, i think i worded my main objective wrong. LOL...oops..
    i am looking for COLOR BALANCE. a nice pure white, yeah that would be great. BUT, my main concern would be all modules fully on, lasing together i dont want one washing out the other. i want red to be as bright as blue. green as bright as red...so on and so forth...

    "allthatwhichis"-
    what is your blue? 473? or 457?

    which now brings me to this point and question-

    if i decide to combine 2 reds (im looking at about a 3-4 W RGB) i am trying to understand this "polarization" aspect. obviously i cant just get 2- 1W 660's, shine em through a beam cube (combiner) and get 2W out. this is correct right?? what does polarization play into combining 2 same wavelegth beams to increase power??

    buffo, and allthatwhichis-
    THANK YOU very much for your previous replies!!!

    -GLL

  5. #5
    Join Date
    Oct 2006
    Location
    Central Florida
    Posts
    7,067

    Lightbulb

    Quote Originally Posted by gottaluvlasers View Post

    "allthatwhichis"-
    what is your blue? 473? or 457?
    Mine is 473. I dream of the day I can get 457nm...


  6. #6
    Join Date
    Jun 2007
    Location
    northern maine.
    Posts
    329

    Default

    from what i've read here,the polarization issue as regards combining beams with cubes,is that you shift one laser so it's polarization is 90 degrees to the other.
    i'm sure you can get more exact instructions for this process from others here.
    combing two 1watt reds, very cool idea.

  7. #7
    Join Date
    Jan 2006
    Location
    Charleston, SC
    Posts
    2,147,488,509

    Default

    Quote Originally Posted by gottaluvlasers View Post
    I chose to go with the 660 for my red, well #1 for obvious reasons- its A LOT cheaper!! im looking at 1+ WATT for the red.
    I'm confused. In your original post you said you were looking at having 300 mw of red at 660 nm. Now you're looking for over a watt? That's going to be very difficult. (And expensive) You can combine two red diodes using a polarizing beam-splitting cube, but that still won't get you anywhere near a watt. (More like 500 mw.) I don't know of any 660 nm diode that can make more than about 300 mw while still keeping the small, tight beam profile...

    Now, if you want to get creative, you can use a knife-edge mirror to position two beams very close together. Do that twice, and then use a single PBS cube to combine the two pairs of beams, and you could approach a watt of power. But your beam will be twice as large. (This is how the arctos units get so much power, btw.)
    is it true that it proveds a MUCH better beam quality over the 635? I have heard TERRIBLE beam specs for most 635 reds.
    At lower power levels (200 mw and less) you can get pretty much the same beam quality from a 635 nm diode as you can from a 660 nm diode. However, once you get up into the higher power levels (300-500 mw), you are correct. The beam on the 635 nm diodes at those power levels is more like a flashlight than a laser. Often the beam is so large that it won't fit on the scanner mirrors, so right away you loose a good bit of power because it "spills" off the mirrors and is lost. Then too, the divergence is a lot higher, so the beam spreads out fast... You end up with a big blob on the wall instead of a small dot. You can imagine what this makes your graphics look like...
    i think 660 is going to be my happy medium. i know not a perfect beam, but a pretty good beam. i know not as visible as 635, but still pretty bright. im on the right track???
    Well, you are on the same track as a lot of the hobbyists here anyway. 660 nm red will also give you a wider color pallette, BTW. (Oranges and pinks look different with the deeper red.) However, an alternative to a pair of 200 mw 660 nm diodes is a single 200 mw 635 diode. You'll end up with slightly more brightness (because of the eye's increased sensitivity to the 635 nm red) and end up spending about the same amount of money. Beam quality should be close, so long as you don't exceed 200 mw of 635.
    i am looking for COLOR BALANCE. a nice pure white, yeah that would be great. BUT, my main concern would be all modules fully on, lasing together i dont want one washing out the other. i want red to be as bright as blue. green as bright as red...so on and so forth...
    In other words, you are looking for color balance! The power ratio's in the posts above are derived from the human eye's sensitivity to different wavelengths. So if you have a projector with perfect color balance, the red beam will look as bright as the blue beam, which will look as bright as the green beam. And when you put them together, you'll get white. (Though truthfully, there are a lot of different shades of white too, from a warm redish white to a cool blue white.)

    With regard to one color "washing out" another color, this *is* a problem. Since you only have TTL modulation, you need to be more careful about selecting your lasers, because you can't dial back the power to adjust your color balance. (Well, at least not in software... You could build a biasing circuit to adjust it in hardware, assuming you had lasers with analog blanking, but that's more work than you need. Better to get the color balance right at the begining.)

    One other thing I should mention... Power density plays a role in how bright the beam looks. So a nice, tight beam will look brighter than a fat, divergent one. Also, beyond a certain intensity, the human eye will begin to saturate, and thus the beam won't look much brighter even if you keep increasing the power. (This is more of a factor when you're doing graphics though. When you're doing beam shows, you're spreading that power out all across the room, so more power almost always looks brighter.) Anyway, the net effect is that at really high power levels, the ratio of red to blue to green doesn't need to be quite so far apart. For example, 3 : 1 : 1 might be a perfectly acceptable color balance if you're runnning 5 + watts of power in medimum-sized indoor venue, because the lasers are going to appear massively bright...
    if i decide to combine 2 reds (im looking at about a 3-4 W RGB) i am trying to understand this "polarization" aspect. obviously i cant just get 2- 1W 660's, shine em through a beam cube (combiner) and get 2W out. this is correct right?? what does polarization play into combining 2 same wavelegth beams to increase power??
    No, you can't just use a standard beam-splitting cube to combine the beams. If you tried this, you would see that you would end up with two beams exiting the cube on the other two faces. They would be roughtly the same power level as the incident beams (minus optical losses in the cube), and would consist of 50% power from each of the two incident beams.

    The only way to combine two beams of the exact same wavelength is to use a polarizing beam-splitting cube. This cube is normally used to separate a randomly polarized incident beam into two component beams that have their polarization angles at 90 degrees to one another. (IE: one beam is horizontally polarized and one is vertically polarized.) So what we do is reverse the process. We start with two polarized lasers (aligned so that the beam from one is horizontally polarized and the other is vertically polarized) and send those two beams into the PBS cube. What comes out the other end is a randomly polarized beam that has roughtly twice the power of a single laser. (Again, minus the optical losses through the cube.)

    Now, lots of people ask, "Well, why can't I just do the same thing again? Use another PBS cube and combine this new beam with another beam from another pair of lasers, and then I'll have 4 times the power..." Ahhh, but you forget that once the two beams have been combined, you now have a randomly polarized beam. So if you send it into a PBS cube, half the beam will go straight through, and the other half will be reflected. You've lost the ability to descriminate between beams now, because the polarization is random. So you can only use the PBS cube as a combiner ONCE. If you need more power than that, you either buy bigger lasers or you use the knife-edge mirror trick to position multiple beams *really* close together...

    Adam

  8. #8
    Join Date
    Jul 2006
    Location
    Connecticut, USA
    Posts
    2,478

    Default

    "buffo-"

    "I'm confused. In your original post you said you were looking at having 300 mw of red at 660 nm. Now you're looking for over a watt? That's going to be very difficult. (And expensive) You can combine two red diodes using a polarizing beam-splitting cube, but that still won't get you anywhere near a watt. (More like 500 mw.) I don't know of any 660 nm diode that can make more than about 300 mw while still keeping the small, tight beam profile..."

    -SORRY adam, my apologies. my original post, when i used 300:200:100 mW's, i was just using those figures for simplistic math as an example of powers to be used for a good balance. sorry for the confusion.

    i searched and searched. CNI offers 1W 660's. yes, a little pricey. but i am building this for my business, not simply hobbyist. (which, by the way...some of your guys pics on here of your "home brew" projectors, WOW!! GO PRO!!! LOL...)
    anyway...
    i am thinking more along the lines of

    2W- 660 (doing this beam polarizing combining thing. ill get back to that in a sec)
    1W- 532
    500mW- 473 (i know its gonna be blue deprived, but price on 473's- OUCH!!)

    the "knife edge" combining you talked about- COMPLETELY lost on that! lol. i think ill stick to the polarizing beam bube for the 2 reds.

    which brings me to this-
    what should i be looking for for the "polarizations" specs for both the laser and the beam cube? and stupid question, but- why cant i just use a red dichro on a 45 deg AOI to one of the 660s??
    Attached Thumbnails Attached Thumbnails 660.jpg  


  9. #9
    Join Date
    Jan 2006
    Location
    Charleston, SC
    Posts
    2,147,488,509

    Cool

    Quote Originally Posted by gottaluvlasers
    my original post, when i used 300:200:100 mW's, i was just using those figures for simplistic math as an example of powers to be used for a good balance.
    Ahh... Ok - gotcha!
    i searched and searched. CNI offers 1W 660's. yes, a little pricey. but i am building this for my business, not simply hobbyist.
    Yeah, they do. If you're doing this for a production projector, then you've no doubt got the $$$ to spend on it. It's a bit out of the range of the average hobbyist though! So yeah, two of those units - combined via a PBS cube - should give you around 2 watts of red.
    2W- 660 1W- 532 500mW- 473
    Well, you are going to be very green heavy then. Remember that Full Auto only supports TTL outputs. You won't be able to dial back the green to get good color balance, and with a full watt of green your "white" is going to look greenish-white. Likewise, your yellow will be more yellow-green - not yellow-orange at all. Your color balance will be something like 2 : .5 : 1 in that example. (Remember that 4 : 2 : 1 is a lot closer to optimal.) It will make some very bright beams, however!
    the "knife edge" combining you talked about- COMPLETELY lost on that!
    It's pretty simple, really. You use the very edge of a mirror to reflect one beam while letting another beam pass just to the side of the edge of the mirror. This places the two beams *really* close together so that it looks like a single beam. Have a look at the pictures in this thread for an idea as to what I'm talking about.

    It's a pain in the ass to align, and the beam will be larger than it would be from a single laser (obviously), but it does work...
    what should i be looking for for the "polarizations" specs for both the laser and the beam cube?
    You need one laser to have a vertically polarized output and the other to have a horizontally polarized output. Normally this is accomplished by rotating the diode in the housing, but if CNI won't adjust it for you, you can always mount one laser flat and mount the other on a vertical plate. (Assuming the polarization axis for each laser is either parallel or perpendicular to the baseplate.) Note that each laser needs a polarization ratio of at least 100:1 for the PBS cube trick to work.

    The cube needs to be a polarizing beam-splitting cube rated for red wavelengths. It will have one face labeled for the incident beam, one face labeled horizontal, and one face labeled vertical. The three faces will be in a T shape as you are looking down on the cube from above.

    Assuming the left side of the T is the incident face and the bottom side is the horizontal, then normal operation would be for a beam to enter the cube on the left. The cube would split the beam into a horizontally polarized beam coming out the bottom and a vertically polarized beam exiting the right side of the T.

    When you set the cube up in your projector, you will reverse this process. One laser (the one with the vertically polarized beam) will send it's beam into the right side of the T. The other laser (the horizontally polarized one) will send it's beam into the bottom of the T. The two beams will combine in the cube and you'll get a randomly polarized beam out the left side of the T. Make sense?
    why cant i just use a red dichro on a 45 deg AOI to one of the 660s??
    Because the dichro will reflect 660 nm light no matter where it's coming from. In your picture, the beam from the left would be reflected to the top of the picture, while the beam coming up from the bottom would be reflected to the right. So you'd loose the power from the beam on the left, since it would be sent off into the wall of your projector housing rather than continuing to the right where the scanners are.

    Remember, a dichro is nothing more than a wavelength-selective mirror. It reflects certain wavelengths while passing others. If you hit it with a wavelength that it's designed to reflect, it *will* reflect it - no matter which way the beam is traveling through the dichro.

    Adam

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •