|
 |
Pangolin has studied laser
scanning at ILDA 50K and above extensively over the past few weeks. We
have reached some conclusions about issues related to 50K+ scanning. The conclusions and the
extensive research which supports them, are presented here. We hope this will help stimulate constructive debate in the laser community. We also hope it will help separate speculation from fact. Faster scanning: Current systems have limitations |
| 1: | 50K+
scanning as currently implemented by Cambridge and Pangolin is a
"trick". It is not due to any significant speed-related improvements in scanners themselves.
Faster scanning is accomplished
almost exclusively by retuning or slightly modifying the scanner amps used with Cambridge 6800 and 6210
scanners. Therefore, anyone with Cambridge 6800 scanners can run at 50K+ by modifying the amps (e.g., TrueK 50). It is not necessary to purchase 6210 scanners to get faster speeds. |
|
|
|
2: | 50K+ scanning improves small-step performance (when the scanners
move a small distance). Faster scanning does not improve medium- and large-step
performance. In fact, if you consider the 10% square wave overshoot of Cambridge factory-tuned 6210s, the large-step response is degraded. |
|
|
3: | As a result of item
2, 50K+ tuning has some drawbacks for
many existing laser shows. If you take existing 30K-tuned shows and speed
them
up proportionally (1.67 times the ILDA 30K speed), there will be obvious
flaws in parts of the show: A) 50K+ tuning means more frequent power limiting (where the scanner amp suddenly shrinks the image to prevent overheating of the scanner). B) 50K+ tuning means more rounded corners and a less-precise look on some frames. |
|
|
4: | The 50K+ drawbacks for existing laser shows can be solved simply by
scanning the show smaller. We estimate that the scan size needs to be
reduced by about one half (compared to the maximum size used for the same
shows displayed at 30K). |
|
|
5: | For new shows designed for 50K+, it is possible for laser software
or individual digitizers to minimize the 50K+ drawbacks. This is done by
adding points. The points slow down troublesome scan lines where the laser jumps
too far, and also help reduce power limiting. However, adding points has two drawbacks: A) Adding points reduces the benefits of 50K scanning. For example, if the software or digitizer adds one extra point for every five existing points, this is a 20% reduction in speed, and effectively gives only a 40K scan speed. B) Some types of laser frames require as-fast-as-possible long jumps. These include abstracts, rasters and star fields. Adding extra points to slow the jumps will make abstracts look different (example: less dense). It will also reduce the 50K performance gain, so you don't get 1.67 times as many raster lines or stars (we estimate the gain to be 1.3 to 1.5 times). |
|
|
6: | It is up to individual users and the laser industry as a whole to
decide whether the gains of faster small-angle performance, outweigh the
negatives of reduced large-angle performance. Cambridge recognizes this. In early July 1999, they wrote that "Cambridge Technology makes no recommendation for or against regular use of the 6210 at ILDA 60Kpps at this time". Cambridge 6800s and 6210s: Speed twins |
|
|
7: | The new Cambridge 6210
scanners (galvos) are essentially equivalent in speed to the older Cambridge 6800
scanners. Any speed increase is achieved by re-tuning and changing
components on the scanner amps. Both the 6800 and 6210 scanner amps can be modified to reach the same speeds. This is supported by three pieces of evidence: A) The published and stated specifications of the 6210 scanner (galvo) are equal in all speed-related aspects to the specifications of the 6800 scanner. B) The motors of the 6210 and 6800 scanners are near-twins. It is impossible for the 6210 to go faster than the 6800 due to anything inside the scanner. C) Pangolin's 6800 scanners go just as fast as the two sets of 6210 scanners Pangolin has seen operating. No 6210 thus far has gone faster than Pangolin-tuned TrueK 50 6800s. |
|
|
8: | At
angles of 6 to 7 degrees, the Cambridge 6210 scanner/amp sets demonstrated so far
have only achieved an ILDA speed of 50K. This is supported by demonstrations and tests of the two sets of 6210 scanners which have been delivered to laser show companies as of June 28 1999. A) New Method Lasers demonstrated their 6210 scanners at the Laser FX 99 conference. The ILDA speed at about 7 degrees was about 50K. (There was some speculation that an inadequate power supply caused the 10K shortfall, but this could not affect the ILDA Test Pattern at a small 7 degree size.) B) Pangolin is publishing photographs of our 6210 scanners. We found a maximum ILDA speed of 50K at 6 degrees. |
|
|
9: | Cambridge has frequently stated and supported claims of ILDA 60K to
66K. This is true according to the ILDA scanner tuning standard, which
specifies the tuning can be done at any scan angle 8 degrees or below. Thus,
for small angles, the 6210 can run at ILDA 60K and above. Pangolin has documented this performance with photos of ILDA 60K at 4 degrees and ILDA 72K at 3 degrees. Cambridge has also documented this performance with oscilloscope photographs, available at the Laser F/X "Scanner Wars" web page. |
|
|
10: | Based
on item 9 above and on Pangolin's tests, Cambridge 6800 scanners with
TrueK 50 modifications can also run at ILDA 60K, at small angles. Again,
the performance of 6800s and 6210s is essentially equivalent. |
|
|
11: | Cambridge's factory-tuning of the 6210s is not optimal for laser
show applications. The Pangolin TrueK 50 modification gives demonstrably
better performance. This is supported by the photos of square wave performance in the Cambridge-tuned 6210s, which overshoots 10% at 30 degrees, versus the Pangolin TrueK 50, which has no overshoot at 30 degrees. It is important to have as little overshoot as possible, for important laser show images such as abstracts, raster patterns and star fields. |
|
|
12: | Cambridge
has repeatedly stated that they do not tune 6210s at the factory to any particular
standard, and they do not tune to the ILDA Test Pattern. When you
order from Cambridge, you cannot specify
ILDA 30K, or ILDA 50K, or any other tuning. Cambridge will
supply the tuning they think is best for you. Therefore, you should test your Cambridge-supplied 6210 system to see how it performs on the ILDA Test Pattern, and on any other tests or shows that are important to you. If it does not meet your desired tuning, (ILDA 30K, ILDA 50K with no square wave overshoots, etc.) you then should re-tune the system yourself. |
|
|
13: | A
test pattern, or set of test patterns, is a necessity for the laser
industry. Sometimes it is said that "What really matters is how existing and new laser shows will look when projected for an audience." But it would be impractical to run all existing shows or create new shows, merely to evaluate the performance of a scanning system. Therefore, test patterns must be used to single out the most important aspects of relative scanner performance. |
|
|
14: | At
this time, the ILDA Test Pattern is the best single industry-wide indicator of
relative scanner performance. |
|
|
15: | Ideally, companies marketing
scanners and projectors to laserists should be able to tune, test and
demonstrate using the ILDA Test Pattern. This does not replace other
tests, whether using lasers, oscilloscopes or other techniques. But it
does provide the most common point of reference for laserists. |
|
|
16: | Companies
should recognize that the ILDA Test Pattern is the best industry-wide
standard. Companies should provide information which is meaningful for
"typical" laserists who may not be scanner or electronics
experts. A) Laserists should not have to interpret oscilloscope photographs, instead of seeing laser-projected test patterns or familiar sample frames. B) Companies should not claim that they "do not tune or specify to the ILDA tuning standard" and then turn around and repeatedly quote performance in ILDA-tuning equivalents. C) Companies with a better system than the ILDA standard (whether laser-based or oscilloscope-based) should propose this for ILDA consideration. |
|
|
17: | While
it is the best widespread standard, laserists should recognize that the ILDA Test Pattern is not perfect. A) The circle-in-the-square test measures performance only at a single point on the scanner's performance curve. B) It is easy to develop a scanner or tuning which passes the ILDA circle-in-the-square test at very high speeds, and yet has terrible performance on other test patterns or real-world sample frames. |
|
|
18: | The
ILDA Technical Committee is currently discussing revisions or expansions
of the ILDA Test Pattern. These could include tests for accuracy, overshoot, power
limiting, and other effects which are not directly shown in the single
ILDA Test Pattern frame. A) Laserists should determine the most important characteristics for today's compatibility and for tomorrow's desired capabilities. B) Laserists and suppliers should work together to develop and validate new tests. The tests should give reliable, repeatable indications of relative scanner performance in the areas deemed most important by laserists. |
This page last updated: Tuesday, June 19, 2007 08:54:20 PM
Tech Support
