Test Pattern Question

Notice the double lines, what is causing that?

I can't seem to get a good picture of the standard ILDA test pattern. It looks OK, flickers even at 30k but I guess thats normal. All my pictures are missing some of it although in real life its all there. Tried longer exposure on my camera but that didn't work either :(

Circle touching the square and the "color" shift lines at the top look good but this pattern in that picture is the only thing I can find wrong in any of the patterns.


Attachment 43493

This is the Laser Media test pattern. What you're seeing is the fact that the test pattern is exposing a slight flaw in the tuning of your scanners. When those 2 lines don't touch, it means one scanner is running faster than the other. Not by much, though. With some additional fine tuning, you should be able to get that cleaned up. I typically flip back and forth between the ILDA test pattern and this one when tuning.

Quote:

---

Originally Posted by demolish50

Notice the double lines, what is causing that?

I can't seem to get a good picture of the standard ILDA test pattern. It looks OK, flickers even at 30k but I guess thats normal. All my pictures are missing some of it although in real life its all there. Tried longer exposure on my camera but that didn't work either :(

Circle touching the square and the "color" shift lines at the top look good but this pattern in that picture is the only thing I can find wrong in any of the patterns.


Attachment 43493

---

Is that also what causes O's to look like q when there are lots of O's being drawn? I thought it my scan angle but it makes no difference what angel I used.

I doubt it, but would probably need to see the frame to make a better guess. The tuning is pretty close, only slightly off from what I can see with the LM test pattern, and if you say the ilda test pattern also looks good then it may just be an issue with the frame itself.

Locate the low frequency damping and servo gain potentiometers on your scanner amps. Load the LM test pattern at a medium angle. Pick one amp and give each pot a fraction of a turn, say no more then one turn in either direction. Start with the damping, for it is more sensitive then the servo gain. See if gap between the lines expands or collapses. On one axis it should expand and on the other it should contract. You can then split the difference as needed between the two amps. If your amp has 20 turn instead of 10 turn potentiometers, you may need two full turns, in which case you should mark the screw position with a paint pen or permanent marker. If your scanners have two different mirror sizes, the larger mirror will be inherently slower, which is something to keep in mind. If things are way out, it may take more then one or two turns, but usually at the stage shown in your picture, the tuning is very close. The controls interact, which is why I'm suggesting small turns at first.

The LM test pattern is designed to show the difference in the velocity between the X and Y scanners, using the diagonal lines. It will also show how well the velocities and damping match at the corners, as the damping error grows, the corners show "overshoot".

http://lasershowparts.com/files/Arti...20Tutorial.pdf

http://www.laserfx.com/Backstage.Las...Scanning1.html

The ILDA pattern is scanned at specific angles, in order to standardize the tuning between users, this chart from Pangolin Inc. helps find the angle:

http://www.pangolin.com/userhelp/scanangles.htm

Steve

That is a wealth of info. Exactly what I was looking for. Thanks .

So I was able to use my wife's camera and get a good picture, long shutter speed and my hands tend to shake but turned out ok. It does appear the blue circle is slightly outside the square on the horizontal axis. I had not noticed this before. I made sure I had an 8 degree scan angle this time though.

Attachment 43511

green wall really makes the colors look funny =p

Hey guys. It's been awhile since I've posted here. I accidently broke off one of my scanner mirrors. I glued it back on pretty good but now I need to retune the scanners. I am scared to death to do this but I guess it's time to learn how. My problem is that I can't find the test pattern with the circle inside the square. The one I find has the circle outside the square and it's not a perfect circle. Is it ok to use this one for retuning? I'm using Spag...ti software on a home built projector for personal use only. Thanks for your help.

Quote:

---

Originally Posted by 2tall

Hey guys. It's been awhile since I've posted here. I accidently broke off one of my scanner mirrors. I glued it back on pretty good but now I need to retune the scanners. I am scared to death to do this but I guess it's time to learn how. My problem is that I can't find the test pattern with the circle inside the square. The one I find has the circle outside the square and it's not a perfect circle. Is it ok to use this one for retuning? I'm using Spag...ti software on a home built projector for personal use only. Thanks for your help.

---

You have the right test pattern. You are misreading the speed test portion of the pattern.
Set your scan angle to about 8 degrees on the screen. Increase your point rate (PPS) till the circle just collapses into the Square. At this point you;ll see whats going on. All this does is measure your upper scan speed.

Google Search for "Buffo's scanner tuning tutorial"

Steve

Here's the link: http://lasershowparts.com/files/Arti...20Tutorial.pdf

Adam

Thanks for the reply. I am still having tuning problems. I've tried tuning to the test pattern with the circle outside of the square and it looks ok but not good. I'm not sure I have everything setup right(scan angle, scan speed and distance)and I would like to tune to the test pattern with the circle inside the square but every time I download the pattern it's the one with the circle outside the square. My projector consists of PT20 scanners with I-Trust show card with a 400mw RGB China brick module. I'm using Spaghetti software with an Ishow DAC. This is my first laser project so I'm using it to experiment and learn on before investing in more/better hardware so please go easy on me. Thanks for any help you can give me.

The test pattern is designed in such a way that when the software conditions are correct, and when the amps are tuned correctly, the circle will collapse into the square.
The pattern is designed to measure speed using the complex frequency response of the galvos. Aka the "Transfer Function" or Equation of Motion. The Circle only has a handful of points, and thus will collapse at high speeds.

Steve

Quote:

---

Originally Posted by 2tall

I've tried tuning to the test pattern with the circle outside of the square and it looks ok but not good.

---

You need to re-read the tuning tutorial I linked to. The circle is supposed to be INSIDE the square. (As Steve mentioned above.) You will not get your tuning correct with the circle outside the square. Period.

Quote:

---

every time I download the pattern it's the one with the circle outside the square.

---

Don't look at the frame in your software, look at the displayed pattern on the wall. That's what matters. Ignore what you see on the computer screen. You are using the correct pattern.

Note that there was a special setting in Spaghetti for tuning scanners. Contact Gary Harper (JohnYayas here on PL) and make sure you have the software set correctly before you try to tune. Otherwise, you won't be able to get the tuning correct. (Spaghetti has some frame optimization built-in, and this will interfere with how the standard ILDA test pattern is designed to function.)

I don't remember the details on how to do this, but I thought that Spaghetti had a special version of the test pattern that you could call up from a menu, and doing this also disabled the frame optimization. But check with Gary to be sure.

Adam

Wait a minute.... now, I don't remember what's built into Spaghetti or what may be different about it but....

My understanding of this is there are 2 ILDA test patterns, and I'll lay money I've seen both. One that says 12k in the lower left hand corner and one that is 30k. The 12k pattern always shows the circle as being outside the square with the four corners of the square touching the circle. Nobody ever discusses that one, as we all refer to the 30k pattern. The 30k pattern has the circle inside the square with the circle touching each of the four sides in the middle.

If he's got 20k scanners, he would want to use the 12k pattern correct? I can see how he'd be confused if he's got the same patterns I know I've seen. Why would you tune 20k scanners to 30k?

The issue I regularly encounter is that I can never find the ILDA 30k test pattern without the huge white square around it when I'm trying to locate it. Everyone (including buffo) says I have it but I'll be damned if I ever find it. . When we've tuned at LEM's, it's always been really easy to see the circle and what's going on as you adjust the pots. When I get home and try to do it, I bring up the pattern and it's the one where everything is tiny with a huge white square around it.

I just fired up Quickshow a second ago to double check and sure enough, it's only the one with the big white box.

I also just fired up BEYOND and in the list of test frames it has both the 12k and 30k patterns that I mentioned. The 12k has the circle outside the square, the 30k has the circle inside the square but it's really tiny with the big white box.

I just checked LD2000 and it's the same as BEYOND. A 12k pattern that's big enough to see (with circle outsode the box) and the 30k pattern which even says "Reduced-size ILDA test pattern. Best suited for tuning to 30,000 pts/sec standard." And has the huge white box.

(The 12k pattern says - "Full size ILDA test pattern. Best suited for tuning to 12,000 pts/sec. standard." )

So.... I understand his confusion. As even I have been, after being around for 4 years and been through several tuning seminars including a hands on one where I've turned the pots.

WTF?

Quote:

---

Originally Posted by Bradfo69

My understanding of this is there are 2 ILDA test patterns. One that says 12k in the lower left hand corner and one that is 30k.

---

There are a couple versions of the 12K pattern, but they are all the same. Some say "12K", but most say "12/24K" or "12/30K". They all have 12 dots in a dodecagon that lies completely outside the center square. When displayed, those dots are "rounded off" to make the center circle, which is supposed to lie just inside (and touching the sides) of the center square.

The 30K pattern looks the same, with the exception that it has a LARGE white box around the outside of everything, and all the features of the 12K pattern are shrunk down really small at the center. While the center features are identical, they are so small that on the screen it's hard to see that those 12 dots are actually outside the center square. But they are. It's just that on the screen it looks like a circle.

So the only real difference between the two is the scan angle that you project them at. The original pattern (the 12/24 pattern) is designed to be projected at 8 degrees or less for tuning. The larger pattern (the 30K one with the large box around the outside) can be displayed at or near full scan angle for tuning. The idea is that the large pattern has all the relevant parts shrunk down really small, so those parts will only be scanned at 8 degrees or less even if the whole pattern is being scanned much wider.

Quote:

---

If he's got 20k scanners, he would want to use the 12k pattern correct?

---

No. You use the same pattern no matter how fast your scanners are. The circle is always supposed to be just inside the square (and touching the sides).

Quote:

---

I can never find the ILDA 30k test pattern without the huge white square around it when I'm trying to locate it.

---

You don't need to use this one. You can tune using the pattern without the huge square around everything. In fact, I think it's easier to tune with the small pattern. I never use the large one. However, both are included on the LD-2000 disk. Don't know about Quickshow though.

Adam

Quote:

---

Originally Posted by buffo

So the only real difference between the two is the scan angle that you project them at. The original pattern (the 12/24 pattern) is designed to be projected at 8 degrees or less for tuning. The larger pattern (the 30K one with the large box around the outside) can be displayed at or near full scan angle for tuning. The idea is that the large pattern has all the relevant parts shrunk down really small, so those parts will only be scanned at 8 degrees or less even if the whole pattern is being scanned much wider.

---

This still doesm't intellectually make sense to me. I understand what your saying but, if you use the 30k pattern and blow it up real big so you can see the test patterm enough to refernce it for tuning... a) aren't the scanners now working their ass off drawing a big useless white box that's being displayed on the floor, ceiling, adjacent walls etc? Shouldn't that effect their tuning ability? b) don't you now have an eye hazard in nearly say... 150 degrees in every direction because of the big white box??

Quote:

---

You don't need to use this one. You can tune using the pattern without the huge square around everything. In fact, I think it's easier to tune with the small pattern. I never use the large one.

---

This seems like a contradiction. I don't need to use the 30k pattern with the offending big white box. I can use the 12k large one that's easier to see. But then you folowed that with: "In fact, I think it's easier to tune with the small pattern. I never use the large one."

Trust me, I know you know what you're talking about and I'm just being a little (or a lot) dense. We'll clear it up in about three minutes at SELEM I'm sure but, (and I'll double check again when I get to work) this kinda means every projector I've ever had is no where close.

I think you and Adam are just using the words 'small' and 'large' to talk about the opposite patterns to each other ;)

When Adam says 'easier to tune with the small pattern' I'm pretty sure he means the one without the big white box, which you are referring to as the 'large' pattern.

Quote:

---

Originally Posted by Bradfo69

if you use the 30k pattern and blow it up real big so you can see the test patterm enough to refernce it for tuning... a) aren't the scanners now working their ass off drawing a big useless white box that's being displayed on the floor, ceiling, adjacent walls etc?

---

There are lots more points that make up that large outer box than there are in the center circle. So no, the scanners aren't working all that hard, although the total point count for the frame is a bit higher, which leads to a little bit more flicker.

Quote:

---

b) don't you now have an eye hazard in nearly say... 150 degrees in every direction because of the big white box??

---

Yes, assuming you have a long throw. This is another reason why I don't like using that pattern.

Quote:

---

I don't need to use the 30k pattern with the offending big white box. I can use the 12k large one that's easier to see. But then you folowed that with: "In fact, I think it's easier to tune with the small pattern. I never use the large one."

---

As Norty pointed out, we have a confusion in terms. The 12K one is the one I was referring to when I said "the small one". What I meant was that it was small because it doesn't have the huge box around the outside.

But yes, I can see how that could be confusing, because the pattern with the huge outer box actually has a "small" version of the test pattern in the middle. I should have been more specific.

Quote:

---

Originally Posted by norty303

I think you and Adam are just using the words 'small' and 'large' to talk about the opposite patterns to each other ;)

---

Good catch, Norty! You are correct. I should have been more clear in my terminology. My bad...

Adam

May I refer you gentlemen to this series of PDfs attached to my post #297116 in the "506 Scanner Review Thread"

http://www.photonlexicon.com/forums/...1&d=1406430084

There are four of them.

As long as the circle in the square is the correct size, the rest of the full box drops out of the equation.

Steve