Page 11 of 11 FirstFirst ... 7891011
Results 101 to 110 of 110

Thread: ILDA Format BS

  1. #101
    Join Date
    Dec 2006
    Location
    Pflugerville, TX, USA
    Posts
    1,977

    Default

    Yes, sorrry. I saw it a day or so ago but was in a hurry at the time and forget to get back to it.

    I like your idea. I would like to see an implementation of it to see how well it work. You idea isn't unique because there are optimization routines in laser drawing tools that interpolate the points excactly how you have described. But, I am not sure if anyone does this real time from unoptimized frames right before sending them to the scanner. I have thought of doing that before. Somewhere on this forum I presented the same ideas of using a tuning wizard to find the characteristics of the scanners. The idea was to display a series of boxes, circles, dotted lines, etc and let the user twiddle things until they all looked right. Meanwhile the software would be determining the acceralation and blanking characteristics of the scanner. The next step would be to strip frames to the basics right before scanning. This means reducing lines to as many points as possible (2 for a straight line) and then interpolating the midpoints and adjusting the end points (using a formula based on results of the tuning wizard) to prevent slowing down too long or not enough at corners.

    The reason I never implemented the above is it would take a lot of experimentation that would take away from my laser show software work. And sense, things are OK as they are I haven't given it much weight. Some day I still intend to do it. I think there are some other folks on here who might have done some similar work (ex the 3d engine one of the guys on here is working on (zoof maybe?).

    Moving on to the 2nd portion of what you wrote. The dynamic frame idea that is interesting. I remember you presented it a long time ago before you went on your leave of absense from the board. Seems like a great way to draw abstracts. But, you don't have to abandon the concept of frames, you just have to have one frame end where the next one picks up. But I don't know how well it would work for drawing moving circles. If the end point of the circle never met the start point, due to what you explained, I wonder if it would end up looking like a moving circle or if the ends would overlap. I'd be interested in seeing that in real life to see what happens.

  2. #102
    Join Date
    Aug 2008
    Location
    UK
    Posts
    5,700

    Default

    I think what you're saying sounds interesting Gary but I also think it would need some kind of intelligent automation so a file put into the optimiser was automatically analysed and then interpolated and out putted. It would also need batch handling capabilities.

    The reason I say this is I doubt many professional users have the time and many amateur users the patience to go through their entire libraries and twiddle with the settings on each and every .ild file until they were optimised. Therefore some form of automation that could batch convert a folder by analysing, interpolating and then outputting an optimised version would be a must.

  3. #103
    Join Date
    Dec 2006
    Location
    Pflugerville, TX, USA
    Posts
    1,977

    Default

    I think you missed what I was saying. The exercise would be done in the software that you use to display laser shows. The exercise would be done once in the laser show software so that it could adjust to the characteristics of your galvos. I'm thinking maybe it would show a box on the wall and ask you to increase or decrease some setting until the corners look perfect. I'd expect there to be several differnent images that would need to be displayed in order to get it all right. But once that was done, the user would not be twiddling anything anymore because the information gathered in the proceeding exercise would apply mathematically to any frame. The laser show software could analyze each frame right before sending it to the scanner and optimize it so that it fits the characteristics of the scanner. The original file would not be touched. You would not want to share the optimized results with other people because their galvo characteristics would be different.

  4. #104
    Join Date
    Apr 2006
    Location
    Orlando, FL - USA
    Posts
    1,740

    Default

    Quote Originally Posted by carmangary View Post
    Some assumptions here are that laser display will stay analog. As soon as a smart person creates a better galvo driver (perhaps one that offers feedback information to the host) with a bidirectional digital interface any argument of using WAV falls flat on the floor.
    Hehe. See US Patent publication number 2005/0093818. It's the engine within this:
    http://www.pangolin.com/_Files/ProjView1.PNG

    Although my name doesn't appear in the list of inventors of that particular patent, that's only because of a clerical error. It is my software that runs in it, and my "moving miror scanner" pictured in the description which was later granted as US Patent 7,092,135.

    See, we have an even better idea than bi-directional digital communication. It's called -- the DSP does everyting. The DSP is the laser control engine, and also a complete two-axis servo system that runs moving mirror scanners.

    We showed this at the 2004 LDI conference in Las Vegas. Works great, and only consumes 4 watts of electrical power!

    Bill

  5. #105
    Join Date
    Dec 2006
    Location
    Pflugerville, TX, USA
    Posts
    1,977

    Default

    That's cool. Who sells them? I am curious to see the price tag and how they are marketed.

  6. #106
    Join Date
    Oct 2006
    Location
    Cleveland, Ohio
    Posts
    2,333

    Default

    Quote Originally Posted by Pangolin View Post
    See, we have an even better idea than bi-directional digital communication. It's called -- the DSP does everyting. The DSP is the laser control engine, and also a complete two-axis servo system that runs moving mirror scanners.

    We showed this at the 2004 LDI conference in Las Vegas. Works great, and only consumes 4 watts of electrical power!

    Bill
    That's cool, it sounds like a mini pango board built right into the projector (ignoring the mirror-gimbal steering system which seemed like some of the purpose of the patent). This still uses bi-directional digital communication - USB is listed in the patent, presumably for uploading the show content, along with an IR remote control.

    This does take things a step beyond in complexity and a probably few more steps in price above what some of us were asking for - a basic ILDA digital point/frame feed standard that still leaves the smarts in or near the computer and only tasks the DSP with displaying the points fed as accurately as possible. This seems like the next logical step above the current analog standard.

    edit: can someone explain how claim 1 is different from a normal projector with a built-in DMX/audio responsive board with built-in frames for control?
    Last edited by drlava; 01-17-2009 at 21:05.

  7. #107
    Join Date
    Mar 2006
    Posts
    2,478

    Default

    Quote Originally Posted by carmangary View Post
    Yes, sorrry. I saw it a day or so ago but was in a hurry at the time and forget to get back to it.

    I like your idea. I would like to see an implementation of it to see how well it work. You idea isn't unique because there are optimization routines in laser drawing tools that interpolate the points excactly how you have described. But, I am not sure if anyone does this real time from unoptimized frames right before sending them to the scanner. I have thought of doing that before. Somewhere on this forum I presented the same ideas of using a tuning wizard to find the characteristics of the scanners. The idea was to display a series of boxes, circles, dotted lines, etc and let the user twiddle things until they all looked right. Meanwhile the software would be determining the acceralation and blanking characteristics of the scanner. The next step would be to strip frames to the basics right before scanning. This means reducing lines to as many points as possible (2 for a straight line) and then interpolating the midpoints and adjusting the end points (using a formula based on results of the tuning wizard) to prevent slowing down too long or not enough at corners.

    The reason I never implemented the above is it would take a lot of experimentation that would take away from my laser show software work. And sense, things are OK as they are I haven't given it much weight. Some day I still intend to do it. I think there are some other folks on here who might have done some similar work (ex the 3d engine one of the guys on here is working on (zoof maybe?).

    Moving on to the 2nd portion of what you wrote. The dynamic frame idea that is interesting. I remember you presented it a long time ago before you went on your leave of absense from the board. Seems like a great way to draw abstracts. But, you don't have to abandon the concept of frames, you just have to have one frame end where the next one picks up. But I don't know how well it would work for drawing moving circles. If the end point of the circle never met the start point, due to what you explained, I wonder if it would end up looking like a moving circle or if the ends would overlap. I'd be interested in seeing that in real life to see what happens.
    I thought I had an example posted here but nope, was caned off the old ISP web space when I left that ISP. Anyway, hosted on PL now. Good to be able to do that. It's a plot of 5 points in sequence, each in turn incrementally adjusted by sin and cosine for accurate rotation, and the interpolator fills in the draw. I used a constant step size, but I guess that as the co-ordinates are known by the control software and not by the scanner, it makes sense to adjust the step size at starts and ends of draws, taking into account angles to determine which scanner takes the harder load in that draw, to accelerate it for speed and line weight optimisations. Note that I did NONE of that, and the scan doesn't show the lack. All I used was the fastest sample rate that could draw straight lines, and the shortest dwell points that allowed sharp corners with no inordinate intensity there. (Short dwell also prevents visible wandering of the point because it hasn't got time to drift before being forced to move on).

    I realise it's not new. Far from it. I knew that any program that drew lines, or made waveforms, had to do interpolation, so I decided to do it too. Remember that this came out of realising I had sound outputs and no dedicated DAC's. The code that does this is very small, it just wrote raw data to file. To make a WAV I'd open the raw file with right parameters, then save as a WAV so other tools can play it.

    One thing about the sequence is the points can be changed in realtime, so morphing and rotating would be easy, you just feed new point co-ordinates to a lookup table array. Would be a great basis for an abstract generator that used ordinary frame images as one form of input. I wanted to fuse the two ideas in the same spirit that Yamaha fuse sampling and raw waveform generation in their SY99 and SY77 'RCM' synthesis concept.

    The rotating pentacle is very basic, but it shows how smoothly it works, even with a low rate wide scan angle. That one was at least 40 degrees.

    Again, I make no claim for any relation to the way others do it. I didn't go there. At the time I made this I didn't know any of that, I just bought soem WideMoves and tried to figure out how to do what I wanted without seeing how others did it. That was the point of the excercise.

    One thing bothers me about this Pangolin thing of putting all the DSP in the scanner end. What is NOT in that end in that scheme? I'm assuming there is still a need for a person programming a show to tell the business end stuff like "turn right, travel X distance in Y units of time, wait, repeat for new co-ordinate..." This is the stuff that defines shapes of graphics, stuff the scanner can't 'know'. Does it disregard drawn line sequences and instead make its own optimisations? If so it amounts to sending a vector language to the scanner end, which then translates it for its own use. Clever, but it seems to defeat TWO things, one being the use of a general purpose PC to do that work, and two being to make a system exclusive to link only with stuff that feeds it high-level contructs it can understand. Better it might be in a self-contained form, but context is everything.

    I think a lot of people will still like an analog/digital rendering of an actual traced path, for the same reason a lot of people still prefer MIDI more than a decade after its first obituaries were written. And if the analog noise is lower than one lsb's worth of jitter, it doesn't matter if it's analog OR digital, so I'm not going to split hairs over that one, we've covered that issue enough...


    Anyway, the rotating pentacle is here:

  8. #108
    Join Date
    Dec 2006
    Location
    Pflugerville, TX, USA
    Posts
    1,977

    Default

    I'm sure different people can come up with many good ideas of how to optimize data before it gets to the scanner. But, it takes a lot of time and thought in order to do that. The point of putting the brain in the scanner is so that each person who wants to control the device doesn't need to know the details in order to make their pictures pretty. Most of these people don't want to know what goes on inside the box, they just want their McDonalds logo to look like Golden Arches instead of a couple of psychedelic tacos or whatever.

    The sooner that the process becomes easier and Joe the Art student can make laser shows without having to really know anything the sooner this industry takes off. I think that there are probably a lot of people here who are afraid of that day, though. And yes, it is true that we don't want laser projectors in the hands of people who don't know what they are doing. But, I think we definitely do want laser show software in the hands of as many people that can get it. At least I do. But, as long as laser show software involves all kinds of complex processing, it will remain high priced and exclusive.

    I don't have much more to say about it I guess.

  9. #109
    mixedgas's Avatar
    mixedgas is offline Creaky Old Award Winning Bastard Technologist
    Infinitus Excellentia Ion Laser Dominatus
    Join Date
    May 2007
    Location
    A lab with some dripping water on the floor.
    Posts
    9,164

    Default

    . But, as long as laser show software involves all kinds of complex processing, it will remain high priced and exclusive.


    High priced and exclusive, you gotta be kidding me! When I started in 1988, they threw tarps over the projectors until it was dark, and put them back on before the end of the show. Tighter security then a Fast Attack Sub base. Finding out where to get a galvo took years of digging. I know, I dug for about 4 years to get my hands on a open loop pair. I had to get a job at Sea World just to get my hands on Ion.

    12K open loop scanning 1990s dollars, cost me 225 for two used G124s, 99$ each for two accel 124 amps, 500$ for a used Amiga 500 with a expansion interface and 50$ for LSD1000, laser was a 5 mW hene. No blanking, no sound sync. And a couple of trips to Canada to meet LMR to get introduced to the community at 300-500$ a trip.

    (PS, thank you Greg at LSDI for the accelerators and Page Hare, Pat Murphy, LMR at L-FX, for turning a light on in the darkness.)

    The world is a lot more open place now, my friend. Trust me.

    Steve Roberts

  10. #110
    Join Date
    May 2008
    Location
    nerdtown, USA
    Posts
    1,165

    Default

    Quote Originally Posted by mixedgas View Post
    . But, as long as laser show software involves all kinds of complex processing, it will remain high priced and exclusive.


    High priced and exclusive, you gotta be kidding me!
    It's better, but it's not good.

    In 1990 a hard drive would have set you back maybe $600. Nowadays you can get 1 Tbyte for $80.

    Lasers are mostly made of the same sort of stuff and, these days, can be mass produced. But the same scaling does not apply, at least partly because the market incumbents like their relatively good margins.

Posting Permissions

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