A few "almost there" shots of the new board.

[liteglow]

Yadda,
Let me know if you need any beta testers.

whaha! Yeah, I second that

We all do.............. we all do... :roll:

[Pangolin]

Hi Yaddatrance,

Interesting board you have there. Congratulations on your development. I always applaud the developments of laser enthusiasts since I was there myself in the past...

I hope you don't mind me commenting on a few things that you wrote above, especially where it concerns Pangolin.

First, as to your being intimately familiar with our product line, and our customer support, may I ask how you became so "intimately familiar" (as I see you have made several errors in your assumptions)? And may I also ask how many times you contacted our "support folks"? In your posting above, you make it sound as though you have had dozens of support contacts but I must say, I certainly don't remember too many, or any for that matter...

Regarding your comments about color, perhaps you are not as "intimately familiar" with our color system as you espouse to be. You wrote that "if you look at [Pangolin's] file formats, you'll note that the color is stored as a 24-bit value". First, our main file format is undocumented, encrypted and compressed, so it would be very hard to "look at" and decipher any individual data stored within our file formats. Second, the color data is absolutely NOT stored as a 24-bit value. It is stored using a method we developed after reading about the color theories of the famous test equipment manufacturer Tektronix. Our color format is specifically designed based on human vision, and is not based on a simple 24-bit RGB model. We do use 24-bit RGB at one point, and that is the point of our drawing program and conversion programs being based on Windows 24-bit color. However, once that gets into our system, it is expanded and augmented. Moreover, in our "Advanced Frame Properties" section, people have constant access to three separate interpretations of color -- a Fully Balanced color output, a High Saturation color output, and a High Power color output. Any of these can be selected at any time, on a per-frame basis, and can also be over-ridden on a per-track basis in Showtime.

Moreover, your discussion about color gamut must be qualified. The color gamut achievable depends on the laser wavelengths used. Someone with 610, 532, 440 wavelengths has a far different color gamut when compared to someone with 647, 530, 520, 514, 501, 488, 457 nm laser wavelengths. And of course, this whole discussion is nonsense for those who only have a red and green laser, or for those with only a single color laser. Our color system implementation is the only one which quickly and easily allows someone to setup and train their output colors regardless of the number of lasers or colors they have, and moreover the only one that actually takes into account non-linear correction, not only of the components you use (PCOAM, Laser diode driver, etc.) but also of the eye-brain system.

Above you wrote "TOO MUCH THINKING" and we agree. That's why we have a simple "Color setup wizard" that allows you to identify the laser wavelengths, and quickly setup a "palette" using a process that is not unlike setting up the palette in Adobe Photoshop. A few mouse clicks, and you are done, regardless of your laser setup or how you have it connected.

Regarding your point about "the actual laser modulation doesn't usually fall exactly on a point" -- this is true. But it doesn't matter and if there is a South Eastern laser enthusiasts meeting, I will show you all how, and why it doesn't matter. Also, you don't blow a lot of points waiting for color to catch up if done properly. Instead of debating this on-line, I would rather just show everyone what I mean in-person. Note that this will be a time where you can put your money where your mouth is. We can project images from each of our systems, side-by-side, and we can each comment to viewers about our respective position, and let the audience decide.

Regarding tuning each color so that the rise and fall times match up, we propose that this be done with simple analog hardware. I helped a few laser companies do this, and started to make up a web page about it, which can be found here:
http://www.pangolin.com/resguide03d.htm

(The page isn't complete, or finished to our normal degree of polish, but it does contain some useful information.)

Basically the concept is this. If you have multiple lasers, all of them should react (in hardware) in exactly the same way. That is, the rise time and fall time of all of your lasers should appear to be the same from outside of the projector. I would liken this to the speed of X-Y scanning. Everybody knows that, from outside the projector, the speed of the X scanner should appear to be the same as the speed of the Y scanner. If the speeds didn't match, you would tune the scanner amps so they did. If they could not be tuned to match, you would complain to the scanner manufacturer to fix their problem. Likewise, anybody building RGB lasers should implement (simple) circuitry so that the speed of the R, G and B lasers appears to be the same. Our position is that it is not a software problem if the blue laser does not modulate as fast as the red laser. This would not be considered a software problem if the X scanner was twice as fast as the Y scanner, so why would it be a software problem if the Red laser is twice as fast as the Blue laser? The answer is "it isn't a software problem, it's a laser manufacturer's problem" and this problem can be solved by a simple analog circuit. Doing so is the only complete solution. That way, no matter what the signal source, the lasers will always be performing consistently...

I do hope that we can get together a South Eastern meeting of laser enthusiasts so that people can see for themselves how all of this works, and the benefits and drawbacks of various color schemes and systems. As the saying goes, the proof of the puding is in the eating, and I invite anyone interested to "taste" the output from our system.

Best regards,

William Benner
Pangolin Laser Systems

[eric343]

Regarding your comments about color, perhaps you are not as "intimately familiar" with our color system as you espouse to be. You wrote that "if you look at [Pangolin's] file formats, you'll note that the color is stored as a 24-bit value". First, our main file format is undocumented, encrypted and compressed, so it would be very hard to "look at" and decipher any individual data stored within our file formats. Second, the color data is absolutely NOT stored as a 24-bit value. It is stored using a method we developed after reading about the color theories of the famous test equipment manufacturer Tektronix. Our color format is specifically designed based on human vision, and is not based on a simple 24-bit RGB model. We do use 24-bit RGB at one point, and that is the point of our drawing program and conversion programs being based on Windows 24-bit color.

Would you mind discussing how this improves on Yaddatrace's idea of storing the color information as 8x8bit? I assume you aren't willing to say how exactly you store it.

lets store all 8 color channels
as 8-bit values in both software and hardware. Hardware engineering wise,
it was slightly painful due to the bandwidth requirements, but we did it and
it works...

[Pangolin]

Hi Eric,

A human's ability to perceive color depends on a few things. For example, a human's ability to discern individual shades depends on the saturation level. Humans can perceive more shades if they are highly saturated, and not so many shades when they are washed out. Also, more shades can be perceived between red and green, than between green and blue. A paper put out many years ago by Tektronix discussed the various aspects of human vision, and we used the concepts in that paper as the basis of one part of our color system.

Our internal implementation might be thought of as a 3-dimensional system, where two out of the three dimensions are non-linear, taking into account how humans perceive color. A lot of "resolution" is put into areas where color change perception is more sensitive.

In addition to this though, there are further non-linearities that exist within laser equipment (for example, the voltage to power, or moreso voltage to perceived-brightness for laser diodes, PCAOMs, scanner-blanking systems, etc.).

Then you have something even more important (or more prevalent), which is the variations of equipment that each user has. I would be willing to bet that no two users on this forum have exactly the same lasers, power levels, drivers, controllers, etc. So how do you reconcile all of these possibilities?

Our system takes all of this into account while at the same time offering the user an easy way to "train" the system so that it understands what equipment you have, how it is connected, and compensates for all of the non-linearities that exist everywhere between our output connector all the way into the picture you perceive in your brain.

Within the walls of Pangolin we have a saying. The implementation of X-Y-Z geometric coordinates is baby stuff when compared to the implementation of color. Color is a very complex topic, especially to do it right. Over the years, many people have imitated how we do things one way or another, but nobody else does color the way we do.

Best regards,

William Benner
Pangolin Laser Systems

[buffo]

Bill,

I'm not sure that you've really addressed the issues that Yadda raised about the way Pangolin handles color. Regardless of how the data is stored, and whether it's encrypted or not, do you have anything to say about Yadda's comments regarding color saturation? For example, the difference between "boosted" cyan, as Yadda's software handles it, vs the way Pangolin does it?

Also, your comments on point-based color look-ahead have me confused. You state that timing differences are a problem for the laser manufacturer, not the controller. Yet you also admit that galvos and PCAOM's have different timing/latency values. Likewise, solid-state lasers have their own timing limits with regard to analog/TTL blanking. Now how is the laserist supposed to remedy this? Only buy lasers from a single manufacturer? Seems to me that Yadda is on the right track here by including support for the timing variances in the hardware and software of the controller, allowing for greater flexibility as to what equipment is used in the projector.

You also made the comment that "done correctly, you won't blow a lot of points waiting for color to catch up..." I was under the impression that Yadda was trying to prevent you from having to waste ANY points, since you'd never need to wait for color to catch up?

[Pangolin]

Hi Buffo,

I don’t want to make assumptions about the knowledge level of Yadda, or his techniques. But if I take what I understand about Yadda's "boosted" color palette, using his cyan example, he says that he is adding some red (and blue) to the cyan. (First, we all must know that the definition of Cyan is 100% Green plus 100% Blue.) Well, if you understand how color works, by adding red to cyan, you increase the power alright (because you have more lasers contributing to the color -- not only green and blue, but also red). BUT, you also DECREASE saturation!!

Try it for yourself. Go into Adobe Photoshop and bring up the color dialog. The color dialog allows you to adjust red green and blue and see the resulting color chip. Likewise, you can see the resulting "Hue", "Saturation" and "Brightness" if you switch modes. Adding red to cyan will only DECREASE saturation, while not increasing Brightness (as defined in the HSV or HSB model) nor changing hue. So, given only this one explanation alone, I personally would not want my color automatically being "boosted" any more than I would want my car to be “boosted”. If you wanted to de-saturate cyan while simultaneously increasing laser power just a little bit, go right ahead. But this can be done much more directly, and in two separate ways using our color system...

Now, perhaps Yadda mis-wrote, and he really meant to write that when the RGB input was 0%, 100%, 100% (RGB value of CYAN), that he would activate both the Cyan line *and also* activate the *green* and blue lines (not the red and blue). This would have been technically more correct. BUT, it assumes that there is a mixed gas laser that has a cyan line, a green line, and a blue line – a phenomenon that is becoming less common as solid state lasers take hold.

Nevertheless, to this specific example, the color system within LD2000 would do something very similar – depending on which of the three color interpretations you have selected for your frame. If you have the “fully balanced” palette selected, it would provide amounts of green, blue and cyan to be balanced in the context of other colors that might be within the same image. If you had the “high saturation” palette selected, it would probably reduce the cyan line a bit, which (depending on the shade of cyan) would give you more saturation, and if you had the “high power” palette selected it would give full power to all green-related, cyan-related, and blue-related lines do give you maximum power at the expense of color balance and saturation.

Regarding the timing differences, yes there are differences between PCAOM and Scanner and Laser Diodes, etc. Differences are OK, and to be expected between the scanning system (both X-Y scanners as a system) and the color modulation system (all lasers as a system). What is bad is for there to be differences *within* the scanning system (i.e. one scanner faster than another) or *within* the color modulation system (i.e. one laser faster than another). Moreover, what I am saying is that there is a right place and a wrong place to put the compensation. The right place is in hardware, in the laser diode driver itself. The wrong place is in software. Again, you must take a close look at my argument about the differences in scanner speed. Absolutely nobody in the world would accept a situation where the X scanner is twice the speed as the Y scanner. Everyone would say "hey scanner manufacturer, fix your scanner -- make them the same speed". That's what I am saying about the laser. For an RGB laser system, all lasers should perform at the same speed -- just as for an X-Y scanning system, both scanners must perform at the same speed. Its the only *completely right* way of doing it. There are other ways, but they are what we call "band-aids". If you get a cut on your arm, and you cover it up with a "band-aid", you can't see the cut any more, but it is still there. Likewise, if you try to cover up the differences of an RGB laser in software, the problem will still be there (and it isn't a total solution anyway -- I can prove all of this in person). Doing it in hardware is the only *total solution*.

As for how is a hobbyist supposed to handle this, I suggest one of two ways. First, for the laser manufacturers from whom you buy, demand (or at least request) that they spend the extra dollar’s worth of components to make sure that all lasers perform at the same speed. The second method is for the hobbyist to make the circuit themselves. This is easy, and it isn't anything that hobbyists aren't already doing anyway...

My bottom line will always be that the proof of the pudding is in the eating. All of this discussion is pretty pointless if it doesn't result in superior projected imagery. And you won't know the superior projected imagery until you see it with your own eyes.

I will be happy to engage in a discussion and prove everything that I am saying in an in-person setting. It is hard to talk about a visual phenomenon using only this text-based forum as a communications means... As John Tilp once said "talking about art is like dancing about architecture". It just doesn't work... So I suggest the only real way to debate about the results of a visual phenomenon is to start by viewing the visual phenomenon with your own eyes

Best regards,

William Benner
Pangolin Laser Systems

[yaddatrance]

Hi Bill! Great to see you here... Thanks for your compliments on this board.

Maybe I'm just forgetable, but I've spoken to you personally practically
once a week for about 6 months in the mid-90's ironing out some
problems with my QM32 system... In fact, a lot of my PCAOM knowledge
came from you personally. I just haven't needed any support in the past
8 years. I currently own one Pangolin QM2k Pro systems and one
QM2k Intro... but I also buy pangolin systems for install for client sites...
You forgot to mention that you have an excellent support forum.

I also consider myself good personal friends with many of your top distributors.
There's no call to be competitive, I'm not out to compete with Pangolin in
any sense of the word. As I've mentioned before I have no interest in the laser
field as a source of income. The thought of selling and supporting the
number of boards required to even "break" even is frankly unappealing.
My chosen target for my engineering team is the heavy manufacturing field,
a multi-billion dollar industry, however, I was unfortunately bitten by the
laser bug so I do spend a lot of time thinking about what is possible.

To answer your question... we DO handle the color timing in hardware.
Our "trick" is cache the data and use picosecond timing with sample and hold
to find a balance point so that all of the color lines and all of the laserlines
line up on an actual point boundary which allows us to implement a standard
point delay but with near perfect precision based on the actual slew rate
of the color itself.

Not to disparage your obviously extensive research in human color vision by
disagreeing somewhat, but we've also done research and worked directly with
PHd's who specialize in modern color models as well as biology. For example,
noting that the color peaking shifts based on total light, our software can
compensate for ambient lighting conditions in realtime using a simple VfW
camera (which we normally used to tune linearity and used in the preview
window of our play software) to perceptually boost or color-match the
individual laser lines based on the background. Using color math,
by specifying the exact color line, the power per line, the slew rate and
compensating for linearity using a cubic curve, we can choose with a higher
level of confidence an exact color. It's silly, mostly unnoticable, but
I am a firm believer of not compromising and doing things "right".

In fact, Boeing Phantomworks currently uses the code we wrote for this
for use with an advanced optical colormetric barcoding system for precise
and identification of densely placed parts.

I also respectfully disagree that it is the role of the laser manufacturer to
make sure the color modulation slew rate matches... Having purchased
the Belgian 2W whitelight (terrible speed on the red lines) and working
with even the $100,000 JenLas units, I've come to the conclusion that there
is no system which does that. The request for such is just wishful thinking.
On our newest boards, I've fixed requiring such by adding about $30 worth
CPU and offboard components.

We also have a optional hardware FFT which can interpolate the proper
modulation if the requested color frequency modulation is faster than
the physical modulation capabilities of the laser itself.

Also in terms of saturation, you are not entirely correct on how we handle
supersaturated colors... Keep in mind we actually store true 64-bit color
per point so we do not have to make any compromise on color whatsoever.
When performing beamshows, where you need power, not just color
accuracy, there is no substitute for turning on as many color lines as you
can as long as the grand total approximate the color frequency you are
trying to simulate. The fine granularity of 64-bit color allows us to pinpoint
multiple color triads to boost color. I'm not "panning" or putting down what
pangolin has done with color but the very existance of the palette wizard
suggests that pangolin does not use and store and transmit 64-bit color
per point. And color models aside, you cannot argue that a full 64-bit
color pallete somehow has less saturation... The palette wizard is very friendly
and usable, but to use a specific example, it just didn't behave the way I
needed when building a show for a realtime demo for Cirque du Soleil which
required more palettes than would even be considered feasible. Also note
that I think 16W whitelight lasers are terribly dim, especially when compared
to 2000W intelligent lights.

As we all well know, there is nothing wrong with Pangolin, it is an excellent
package with considerable thought put into it's design and architecture.
I'm a paying customer who will no doubt buy newer versions as they come out.
I have a profound respect for you and your company. However, I do not think it
is out of line for me to say it does not do everything I want it to do and just
the way that you addressed my concerns in your post shows to me that you
clearly intend to stay with the "Pangolin Way"... A fine way, it is, and one I
use often, but it is not the "One True Way"...

[Pangolin]

Hi Yadda,

Thanks for the warm welcome. Maybe I would remember you if I knew your real name. Right now, all I know is your nickname Sorry that I can't read between the lines and figure out who you are by only the nickname, but you will have to forgive me, as I am an old dude compared to most of the young whipper-snappers here

While I will reserve a quite defense about most of what you wrote until later, for now I will say that I don't agree with what you wrote about laser manufacturers not putting a simple less-than-one-dollar circuit into their laser to correct their own problems. In fact, recently laser manufacturers are doing this, including NEO-NEON and soon RGB Lasers in Hungary. Others will follow suit as well, now that they know there is a simple and low cost way of doing it, and also they will have to because competition will demand it.

Regarding your color system, I guess I will just have to see for myself how well it works in real life. As I have often said, the proof of the puding is in the eating. Picosecond and VFW blah blah blah, let's see how it looks

If the South Eastern Laser Enthusiasts Meeting happens, I look forward to seeing you and your wazmo system there and we can all judge for ourselves how well it works.

By the way, I told you (and everyone here) that it was the famous test equipment manufacturer Tektronix where we got much of the information to craft our color system. You mentioned "PHds who specialize in modern color models as well as biology". Would you please name these PHds (actually PhD) so that we might do a bit of peer review, and get their take? Maybe they would like to write an article?

Best regards,

William Benner
Pangolin Laser Systems

[yaddatrance]

Hi Bill, My name is Ken Son Sorry, I should have thought of mentioning my
name earlier Now that I recall, you also stepped me through walking
an I-90 which came misaligned during transit over the phone .

I am a director at Wireless Facilities, Inc (http://www.wfinet.com)
and I am also a managing associate at Promedia Telecom...
On the side I'm forming up a company we humorously refer to as
"Ghetto Heavy Industries"... Photon Dream is our "fun" company
where we spend our free time working with side technologies which
amuse us... such as solenoid triggered laminar flow tubes, etc.

I've done my stint working in Government Services as a Senior Research
Scientist, I've also done government sponsored research on the Internet 2,
and RS:NG as the Senior Research Programmer III at MERIT Networks, Inc.
I've worked as a manager at both Ford and Chrysler (before the merge),
so I'm not terribly new around the gills.

My point on the laser color matching is that it'll only be fixed when China
decides to follow suit, and so far from my discussions with them this is
not in the near future, and a horde of "non-compliant" lasers already exist
now...

I'll personally hold judgement on RGB Laser of Hungary's ability to implement
your $1 circuit until I see it as on the last unit I had the distinct displeasure of
fixing (the RGB1800 sold to Mobius 8 by Dean), the slew rates were so
unbelievably broken and slow and their optic alignment system was so prone
to failure that I question their ability. The RGB analog system behaved more
like biasing a broken TTL modulation system to trick it into doing analog. Oh
and all this for a nice cozy $25k+ price.

I forwarded a request to my professor (from my alma mater) to personally
discuss his research into color theory on this site. It would definitely be
interesting to merge the two philosophies. I am assuming your colormetric
study is based on the Tektronix HSV system?

The very reason I did everything I did on the hardware board was precisely to
address all the issues that all the local SoCal old-timers assured me would
come bite me once I finished... we laboriously added features until our friends
ran out of complaints.

I would be pleased to show you the guts of the board at SELEM, and if I end up
down your way earlier than that, I'll let you know so you can see what I'm
talking about. It's not special or extravagent, it's a subtle philosophy change
which lets me do things which can be difficult to do with conventional
systems... I encourage you to hold your judgement until you see it in action.
Some of the differences are subtle, some are quite dramatic. And
of course never fear, Pangolin has two decades of bulletpoints which I
haven't begun to (or even thought of) addressing... What I implemented
was what I wished pangolin had. So while at first glance I have some
wonderful features, pangolin is of course a far more mature and well rounded
product.

[Pangolin]

Sorry Ken, I still don't remember you. And I have to say, I don't remember ever talking you through the walking of an I-90. That sounds more like Bill or Bob Arkin than myself. But my memory is getting foggy in my old age, so anything is possible...

I already wrote above, that I would reserve a quiet defense until seeing the results with my own eyes...

Regarding the cheap Chinese systems, they will have my solution in their lasers sooner than later. NEO already has it, and others will follow suit, if nothing else, out of necessity to remain competitive. We work with the Chinese companies a lot at Pangolin, and that is one thing that impresses me -- they will absolutely add to their systems what is necessary to fill out a "list of features".

More later...

Best regards,

William Benner
Pangolin Laser Systems