scanner options?

[absolom7691]

How can you make such a claim?

It is easy but if the handbag analogy doesn't resonate well, try this. Buy a Sunbeam and then buy a Kitchen-Aid. Buy Jensen and then buy Mark Levinson. There are certain things that just don't compare even though they "perform" the same task.

[kecked]

I had ems 2000 scanners to start. Thought not bad but could never really dial in a pattern. Got a set of 6800's with aging bearing and was blown away. Time goes by and they get worse so I get 6215's...HOLY CRAP what a difference. You get what you pay for. Now the point. I don't run any material over 30k and don't even own any. If you try to record 30k graphics at 60k they don't look right. Pangolin 506 scanners look to be the cats meow for me for my programming projector and I expect the wow of 6215's on 30k material will be there as well. As Bill said they are DESIGNED for the material. What more can you ask. AND when I do play them back on 6215's because the show was prepared on 30k DESIGNED scanners at 30k they will look as I intended them to look when I programmed. This means when one of you get a show I do it will look right regardless of the scanners you have as long as the are 30k DESIGNED scanners.

That's pretty damn important don't you think.

[buffo]

Do DT scanners give results whose quality is acceptable to most people? Sure. At least in the absence of a better-educated and more-discerning public.

This is the key right here that so many people are missing. Everyone is satisfied with the DT-40's, until they see what a better set can do. Then, suddenly, they realize that the DT-40's aren't nearly as good as they thought they were.

Does that mean they immediately scrap their DT-40's and go out and buy new scanners? Not always, no. Some people don't mind the image artifacts and feel they can live with them. And yes, some people also decide to tune them to 40K instead of 30K, even though they are giving up more performance by doing so. Then too, in general, some people are just dumb. (Or at least, uneducated.)

But the people who want (or need) the best graphics performance at the widest scan angles definitely do upgrade. (At least until they can't afford to anymore.) And they also understand that tuning to 30K has real benefits (beyond just a wider scan angle, it affects things like frames with pulled points), so they eschew non-standard scan speeds like 40K, 50K, etc. (To wit: I own a pair of ScanPro-50 scanners. When i tuned them to 50K, they could just barely manage the ILDA test patter at ~6 degrees. But when I tuned them back to 30K, I got 18.5 degrees out of them. Much better!)

When I came to my very first Photon Lexicon event, there were a handful of people who were using MediaLas software along with an EasyLase DAC. When I went to this event, everyone was pretty happy with their DAC. At least that was my impression. But when I pointed out to them that -- when projecting the ILDA test pattern, the circle inside the square was twitching, and then cranked up an oscilloscope and showed everybody why it was twitching (sample jitter, and pretty bad sample jitter at that), then this became unacceptable to everyone.

I can personally vouch for this event, as it happened at the very first LEM I ever organized. (FLEM Mark I, in January of 2007.) For those of you who are curious, you can dig up the original thread here.

As for the comment about the sample jitter not being acceptable, Bill is actually toning things down quite a bit. The Easylase DAC we tested belonged to Fred Blockland (Astroguy). And Fred's exact words after he saw the difference between the EasyLase and Pangolin were: "I want to take a f*cking hammer to my DAC now." He even posted this in the forms, see here. But we were all shocked by the difference. (Well, except for Paul, since he already had Pangolin and knew it was superior.)

At the next Photon Lexicon event I went to, not even a single person was using the EasyLase... The people remained the same.

Yup. Me, Mo, Fred, Aaron... We had all upgraded to Pangolin. I actually had to cancel my EasyLase order as soon as I got back home, since it was about to ship. (I was stuck with the Mamba Black software though, because I had purchased that earlier.) But it took me a while to decide on Pangolin for good. Fred was the first one to jump. He sold his Easylase controller and bought an LD-2000 pro board within 3 weeks of the first FLEM.

Education is really important, and going to LEMs to see the different builds, components and methods used.

You got it. That's why I tell everyone to go to a LEM before you buy or build anything. You'll save a lot of time and money that way!

Very well. Then educate us please. Show us a comparison.

We have, several times. People have been posting their personal stories, and there are multiple threads here on PL dedicated to scanner testing. Have you bothered to look at any of them and read what they say? Or are you just assuming that everyone here is just lying when they say that DT-40 scanners are OK, but there are better ones available?

With your "chip on your shoulder" attitude, I suspect many people are unwilling to go out of their way to help you. I know I am. After all, if you've already made up your mind and can only tell others to "PROVE" what they're saying, rather than asking some more intelligent (and less accusatory) questions of them, why would anyone bother trying to convince you in the first place?

In your example the quality difference in real (not just test) animations might indeed have been noticeable for most people and enough reason to change to another product. But will it in this case? Maybe you think it will, but I'm sorry to say your opinion isn't enough for me.

He stated very clearly what the difference in performance will be. If you don't understand how that performance is measured, then you shouldn't be commenting on the issue at all, except to have the performance spec explained to you. And now you're asking him to predict what your perception of that performance will be? That's completely unreasonable. If you don't understand the terms used, ASK. But until you do, you've got no business making judgments, and in any case you really need to lose that chip on your shoulder.

Don't take this as rudeness, that is not my intention.

It absolutely sounds like it is, though. It's one thing to question a single person. It's another to question several people when they are all in agreement and you are the outlier. And it's even worse when you would do all this knowing that you don't have the knowledge to discuss the science of the issue. You willingness to disbelieve one of the experts in the field (Bill) even as he is explaining to you WHY things are the way they are is hugely arrogant, and very rude.

However, in the chance that this is all just a big misunderstanding, (and even though I've typed this countless times before), here is how you fairly test different sets of scanners to the same standard:

First, you have a fixed distance between the scanners and the projection surface (a wall or screen or whatever). Then you use the same controller for all tests. (Preferably something that doesn't add any artifacting into the image, like the flawed Easylase DAC that had the pronounced sample jitter.) Normally we use a Pangolin LD-2000 board.

Then you set the scan speed to exactly 30K. Now you TUNE each scanner until the performance is as good as you can get it. (Circle completely inside the central square and just touching the sides on the ILDA test pattern, and the diagonal re-trace lines on the LaserMedia test pattern equal and perfectly overlapping.) Remember, these are standard tuning parameters here. There's nothing subjective about it at all. (It's also important that you tune at or below 8 degrees scan angle, to be sure you are within the normal performance range for the scanners when you tune them.)

Now, once they are tuned, you increase the scan angle and watch the center circle on the ILDA test pattern. At some point, the circle will begin to pull away from the sides of the square and start to degenerate into a spiral shape. The point where the circle first stops getting bigger is the ballistic limit for the scanners, and that is your small-signal bandwidth limit. This is a fundamental limit of ALL scanners. At this point you measure the width (or the height, should be the same since it's a square) of the entire pattern, and knowing the distance from the projector to the wall, you calculate the scan angle. Write down that number.

Now repeat the above steps with the next set of scanners. Which ever set of scanners returns the largest scan angle number wins.

This is a long process, even if you have Bill Benner on hand to do the tuning. (Tuning is tedious work, but Bill had so much experience that he can dial in a set of scanners really quickly.) When we did this at a FLEM in March of 2012, it took several hours to work through all the different scanners we tested. Those results have been published several times. And almost immediately, people started questioning our results, despite the fact that at the time Pangolin didn't sell scanners, and thus had no axe to grind with anyone.

Now, given how much time (and work) it took to do that, can you start to see why some people might not want to repeat that effort, given the tone of some of your questions in this thread?

But rest assured, I'm reasonably certain that SOMEONE will re-do those tests (or at least a portion of them) in the future. However, going by what has already been posted (in this thread and others) about details like torque constants and rotor inertia, it's clear that there is a science behind scanner design, and the numbers don't lie. For the vast majority of the chinese scanners, the science shows that it would be impossible for them to reach some of the performance targets they claim. Our tests in 2012 merely proved that the science was accurate.

Adam

[dnar]

I can foresee were this site is heading.... When do the pro-Pangolin/anti-Pangolin badges appear with our Avatars?

[buffo]

You're already unwittingly giving up a lot, by tuning 30K scanners to 40K.

By "tuning" do you mean using DT40 at the default 40K speed?

Yes, that is exactly what he is saying. (Others have said it too.) 40K is not a standard. 30K is.

A video like that would be nice.

You need a video to see that tuning at a faster speed costs you scan angle?

You can demonstrate this yourself with your own scanners. There's no reason to post a video of it.

What I think is Bill is also an idealist.

He is a professional. There's a difference. Bill understands that not everyone can afford the Cambridge 6215 scanners. When he discovered a way to offer superior performance to those scanners, he launched the Pangolin Saturn 5 series. But because he knows that not everyone is looking for the absolute best performance at any cost, he also released the Scannermax 506 units.

Adam

[White-Light]

@ Cipher0, following on from what Adam said, there's a very good page on the ILDA test pattern on Pangolin's website:

http://www.pangolin.com/ILDAtest.html

Everyone usually focuses on the square and the circle because these indicate speed and Adam is entirely correct in drawing your attention to this as the primary concern.

However, if you look at the white annotated diagram on Pangolin's page, you'll find that every part of the test pattern has a function and other issues with scanners can be seen in other parts of the pattern eg lines with slight ripples instead of being straight, or double lining or rounded corners, or the small circles that aren't round. Everything in there has a place. Only when everything is absolutely 100% perfect do you have a scanner that is perfectly tuned or performing perfectly at the tuned speed. (For some reason it has the circle drawn outside of the square although the notes of it being inside are correct!)



Now I pulled this off Google, its been posted on here, so to be clear, I'm not wanting to pick on anyone, but this is an example of how other artefacts affect the test pattern. I'm sure Bill or Adam can give you a much better analysis than me, but one immediately obvious fault is the straight line through the centre which is rippled. Another is the turned down / turned up ends on the blanking speed / gain lines, the main outer box isn't square, the corners aren't square and sharp but rounded nor the lines straight, the inner circle isn't round. I could actually go on and on for quite some time. The point is, the ILDA pattern shows many aspects of the quality of the display so what you should be asking is not if the scanners can make 30K but if they can display a perfect ILDA test pattern at 30K and 8 degrees. Even then, you can't necessarily trust the answer from a manufacturer, the real answer is to go to a LEM and see the results for yourself.





The perfect test pattern is below (courtesy of the Pangolin website):




Does it matter if they can't do a perfect test pattern? Well for beams, Joe Average won't notice until, as others and myself have pointed out, you see a perfect set. Then you''l notice the difference! That's not perfectionism but merely the fact that most people have only seen cheaper scanners and its like a car, you're satisfied with the performance of your GTI until you've tried a Ferrari. The somehow the GTI doesn't quite cut it any more!

@Dream, regarding 40K tuning, as Kecked pointed out, most content is designed to be displayed at 30K so unless you have a specialised usage, you want your scanners tuned to 30K. As Adam said, the advantage of faster scanners is not really the speed but the ability of the scanners to be tuned to perform wider at 30K.

Now going full circle, this is one of the advantages of the 506's. They go ridiculously wide at 30K even though they are only rated for 30K. In fact from the video Bill posted, its very difficult to think of any scanner I've seen that's going to go anywhere as wide whilst maintaining such a perfect ILDA pattern. Maybe 60K Cambridge. Even then, when the pattern started to break down at around 50 degrees, there was only slight degradation. Remember that's a pattern that's supposed to take scanners to their limits at 8 degrees!!!! That is an excellent result in anyone's book.

[Pangolin]

I don't think that Dream or Cipher0 are *trying* to be rude. But I do think that everything comes across that way on the forum, perhaps especially from me, as I have a very straightforward approach. I've also stopped expecting people to simply "believe the expert" even though that would be easiest (and perhaps best) for everyone

I'm still trying to understand the psychology of all of this. It could very well be that if you own a Ford Probe, you'll have a natural tendency to like it, and not want to hear about the flaws when presented on a forum, but if someone opens the hood of this car and a neighboring car and shows you the difference, then you'll believe it. The experience with EasyLase in the 2007 forum showed exactly that behavior.

So I'll make a video. That's no problem. In fact I've already made three -- none of which are "public" on our YouTube channel, but all available with unlisted links (two for which I have given the unlisted links here on PL, with the third being a specific test for a projector manufacturer). Seeing is believing and I'll be able to communicate more details and even crank up the scanners on our advanced test equipment (like an X-ray for scanners) and show you the differences.

I remember him telling me that to him projections with Pangolin software and hardware always look smoother and not jittery compared to others. Correct me if I'm wrong Bill.

Yes that's true, and I will go even further. To understand how all of these other sofwtares and hardwares work would be understand why they CAN'T POSSIBLY BE SMOOTH!! But nobody really understands this the way we do, and also I don't want to teach people about the differences because the differences serve as a competitive advantage for Pangolin...

I had a a RIYA DAC and LaserShow Xpress and I managed to get a copy of Quickshow for a day from someone here and we made comparisons. We tried different animations and different level of complexity. I honestly couldn't tell the difference.

As I wrote above, I'm still trying to understand the psychology here... To me, stuff produced on other people's software is so jittery, I can't even watch it.

Here's something you can try and I really hope you'll see the difference, because at the last PL meeting this JUMPED RIGHT OUT AT ME. Try scrolling text. To me, the scrolling text produced by other systems is like xxxx xxx x x xxxx xxxxxxxx x x xxxxxx xxxx xx (with the spaces representing "spaces in time"). The scrolling text is simply not smooth.

One reason why scrolling text is a good test is because of the way the eye-brain system works. When something is moving, the visual cortex will expect it to be moving at a constant, and non-changing rate. But with other systems the rate will always be changing (and to understand this would be to understand how and why)!! Definitely! No question about it! (Remember folks, we've actually purchased two separate competitors, both of which had this problem, and one of which as based on LDS -- the source code of which all of the other LDS variants are based).

Try it, and see if you can see it. Then try scrolling text on any Pangolin system. Scrolling text (and everything else) is smooth as butter on Pangolin. Really jittery on all of the rest except for Lasergraph DSP who solved the problem in exactly the same way we did.

Now, the degree of jittery-ness will depend on exactly what kind of text is being scrolled, how quickly it's being scrolled, for how long it's being scrolled, etc. When you're flailing beams all over the place, it's quite likely that you won't be able to notice the jitteriness (because the eye-brain system doesn't know what to expect). But scrolling text is just a great test for things. It was also something we tested our LC-Max with, and had to tune our LC-Max to produce well.

I'm not saying Bill was wrong. I'm just saying that my eyes couldn't notice difference.

If you're flailing beams all around a room or running a single, non-animated image, or running an animation where the eye-brain system doesn't know what to expect, then you *might* not see the difference. But if you're running an animation that requires time-consistency, then I hope people will notice.

(By the way, there is one PL member who's stuff is SOOOO jittery, but who is also SOOOO insistent that it's perfectly smooth, it makes me wonder if he really can't see it, or simply doesn't want to admit it... To me the problem is super-duper obvious. Like I said, I'm still trying to understand the psychology in all of this...)

I remember him telling me that

Another thing I said was that our hardware was safer than others' hardware too. I think this too was dis-believed here on PL but finally I incorporated this into a safety-related video that we made and also started doing these demonstrations in public. You can see the video here: http://www.youtube.com/watch?v=pohPyDRsCRk

The demonstration is in the beginning of the video. Although we show only a single competitors' laser controller here, I have a whole box of them at Pangolin, and NONE of them will pass this test. I suspect even laser controllers selling for 100,000 Euro (and yes, there are some) will also not pass this test.

I mentioned this on PL and everyone scoffed. Now I've started doing this demonstration at trade shows and such. It could very well be that "seeing is believing" is the real key here. And I really think the smoothness thing will become important, once people know what to look for...

Bill

By the way Dream, we get the components that make our hardware safer in Russia -- one of only two places that they are made...

[Pangolin]

The wavy lines you see on the picture shown above by Whitelight is caused by "torsional resonance". In this case, pretty bad torsional resonance, and the torsional resonant frequency must be pretty close to the point output rate to be stimulating the resonance that badly. You'd be hard pressed to see any torsional resonance in our scanners, because of the materials we use in the rotor.

But can they display 40K at lower angle like my DT40?

Hehe. Dream, any 30K scanner can be tuned to faster-than-30K. Any! But you'll have the same kind of problems with all of them (to some extent or another):

1. Any resonant behavior will be exacerbated (won't happen in ours because there largely aren't any resonances)
2. Large-signal performance will be sacrificed. Not only will you absolutely have to run images smaller, but...
3. Lines that are supposed to be straight may be curved
4. Heat inside the scanner grows dramatically as speed is increased. This will shorten lifetimes.
4. ... etc...

When I make the video I hope I will be able to convince everyone that TUNING 30K SCANNERS TO FASTER-THAN-30K WON'T DO YOU ANY FAVORS!

By the way, I also did a study more than 20 years ago, and wrote an article for ILDA about the findings. The findings were that you really have to make a 50% difference in scanner tuning before the result in projected images even starts to be noticeable. 30K scanners tuned to 40K will produce images that are just A LITTLE BIT LESS rounded, but that is all. I'll try to demonstrate that on the video too.

Bill

[masterpj]

You might "if it's not your bearings causing this or your shaft flexing attempt to artificially harden your scanner mirrors with epoxy.
The wobbly line in the middle is caused by resonance of your scanset (likely the mirrors here) (surf plank effect?).
The epoxy will increase inertia though so you will be sacrificing the max scanspeed you can obtain and it will require you to do a retune.

[cipher0]

Does that mean they immediately scrap their DT-40's and go out and buy new scanners? Not always, no. Some people don't mind the image artifacts and feel they can live with them.

Exactly! Thats one of the reasons I want a comparison.
And then you think I want a test because Im just thinking you guys are lying to me? Dont you see how ironic you are by calling me rude?

With your "chip on your shoulder" attitude, I suspect many people are unwilling to go out of their way to help you. I know I am. After all, if you've already made up your mind and can only tell others to "PROVE" what they're saying, rather than asking some more intelligent (and less accusatory) questions of them, why would anyone bother trying to convince you in the first place?

So the person who makes a product is "helping" the customer by showing evidence why their product is superior and making it more probable for them to by their product? Kind of. But not helping as in "doing me a favor", but "helping make a choice". They have a gain in linking to a test like that.

It absolutely sounds like it is, though. It's one thing to question a single person. It's another to question several people when they are all in agreement and you are the outlier.

Firstly Im not an outlier. Secondly, Im also not questioning anyone. Im not going to waste my time rephrasing what Ive already said. I dont know why you'e getting so defensive. I want to know if a product suits my needs. Bill showed evidence (complex phyiscs which many customers won't understand, instead of something simple like video comparison) why his upcoming scanners will be better. But as you said,

Some people don't mind the image artifacts and feel they can live with them.

I want to know if I'm one of those people here. I want to *see* how those artifacts look like. Thats all. Chill out and stop making false accusations about me and calling me arrogant.