Page 2 of 3 FirstFirst 123 LastLast
Results 11 to 20 of 21

Thread: Tuning...from a different perspective

  1. #11
    Join Date
    Jun 2010
    Location
    Australia
    Posts
    3,733

    Default

    Quote Originally Posted by Solarfire View Post
    I just posted this yesterday on the EMS7000 thread, seems to fit well here too.

    Being the way I am, questioning things that don’t totally make sense to me, I researched a little more on the subject of the kpps/ILDA pattern question. Well it turns out that my speaker analogy was spot on and the ILDA pattern at best is a compromise made at a time when 60k scanning was somewhere in the far future. The bottom line of what I’ve found is that tuning scanners to reproduce the ILDA pattern as intended, just means that the scanner is tuned to reproduce the ILDA pattern, it therefore says nothing as to how well the scanners will reproduce other complex graphics at different scan speeds.

    The main thing that bothers me about the ILDA pattern is the fact that the scanned image is supposed to display the circle inside the square. This is actually a distorted output of the input signal where the square is in the circle. IMO this makes no sense whatsoever. For my understanding of positioning systems or PID controls, output should equal input. Scanners tuned in this manner, must be prone to also distort other complex images.

    It seems to me that when the ILDA pattern was developed someone leaned towards the static test pattern of a television which has no moving mass, thus the static image. IMO for tuning a moving mass a dynamic signal, covering a broader bandwidth of a given frequency range would much better linearise output over a broader range of scan speeds. As with speakers, the goal is to be linear in output over a given frequency spectrum.

    I don’t know… just my 2 cents worth?
    I have been watching this discussion with a quiet mouth and open mind. Your post makes a lot of sense to me.
    P.S. If you're going too do a show with ILDA patterns then you're all set to go, great show dude...
    I think your on to a brilliant idea! Why didn't I think of that!!!
    This space for rent.

  2. #12
    Join Date
    Jun 2010
    Location
    Zweibrücken, Germany
    Posts
    605

    Default

    Quote Originally Posted by dnar View Post
    I think your on to a brilliant idea! Why didn't I think of that!!!
    Yeah.. I'm currently working on a multi projector ILDA pattern show and YES, we will be doing both, 30k and 12k so don't miss out folks, i'ts goining to be a doosie...
    The sound track is going to be a 1kHz sinwave signal as supplied with lokal television test pattern, enjoy folks!

  3. #13
    Join Date
    Jun 2010
    Location
    Australia
    Posts
    3,733

    Default

    Quote Originally Posted by Solarfire View Post
    Yeah.. I'm currently working on a multi projector ILDA pattern show and YES, we will be doing both, 30k and 12k so don't miss out folks, i'ts goining to be a doosie...
    Don't forget the invert-X and invert-Y variants of this show!
    The sound track is going to be a 1kHz sinwave signal as supplied with lokal television test pattern, enjoy folks!
    Do the full broadcast test tone, 1kHz on left, 1kHz broken on channel right.
    This space for rent.

  4. #14
    Join Date
    Mar 2010
    Location
    Raleigh, NC
    Posts
    2,314

    Default

    Quote Originally Posted by Solarfire View Post
    I just posted this yesterday on the EMS7000 thread, seems to fit well here too.

    Being the way I am, questioning things that don’t totally make sense to me, I researched a little more on the subject of the kpps/ILDA pattern question. Well it turns out that my speaker analogy was spot on and the ILDA pattern at best is a compromise made at a time when 60k scanning was somewhere in the far future. The bottom line of what I’ve found is that tuning scanners to reproduce the ILDA pattern as intended, just means that the scanner is tuned to reproduce the ILDA pattern, it therefore says nothing as to how well the scanners will reproduce other complex graphics at different scan speeds.

    The main thing that bothers me about the ILDA pattern is the fact that the scanned image is supposed to display the circle inside the square. This is actually a distorted output of the input signal where the square is in the circle. IMO this makes no sense whatsoever. For my understanding of positioning systems or PID controls, output should equal input. Scanners tuned in this manner, must be prone to also distort other complex images.

    It seems to me that when the ILDA pattern was developed someone leaned towards the static test pattern of a television which has no moving mass, thus the static image. IMO for tuning a moving mass a dynamic signal, covering a broader bandwidth of a given frequency range would much better linearise output over a broader range of scan speeds. As with speakers, the goal is to be linear in output over a given frequency spectrum.

    I don’t know… just my 2 cents worth?

    P.S. If you're going too do a show with ILDA patterns then you're all set to go, great show dude...
    The idea behind the test pattern is to so that shows can be created and played on different scanners. There is a requirement behind this and it is that the shows must be created and optimized to look good on the scanners while running them at the exact same angle and speed at which they were tuned. If you tune your scanners at 8 degrees and 30K and create a show at 15 degrees and 15K then all bets are off when you play them on some other set. However, if you create the show at 8 and 30 and then play it back on some other scanners that are running at 30K and at the same angle that they were tuned then the show should look correct, including pulled points. That made sense back in the ADAT days where optimizations were recorded along with the shows but it doesn't apply so much when shows are computer generated and optimizations are built into the playback software. For what the test pattern was intended it makes perfect sense. Input NEVER equals output when it comes to scanners. There is always compensation and it is either BUILT IN to the show (which is what the ILDA test pattern is for) or it is ADDED DURING PLAYBACK, which is what all modern software does. Some people attempt to combine the two or expect that the ILDA pattern applies to shows that were created without optimizations built in. The end result is typically confusion and disapointment.

  5. #15
    Join Date
    Jun 2010
    Location
    Zweibrücken, Germany
    Posts
    605

    Default

    Quote Originally Posted by JohnYayas View Post
    The idea behind the test pattern is to so that shows can be created and played on different scanners. There is a requirement behind this and it is that the shows must be created and optimized to look good on the scanners while running them at the exact same angle and speed at which they were tuned. If you tune your scanners at 8 degrees and 30K and create a show at 15 degrees and 15K then all bets are off when you play them on some other set. However, if you create the show at 8 and 30 and then play it back on some other scanners that are running at 30K and at the same angle that they were tuned then the show should look correct, including pulled points. That made sense back in the ADAT days where optimizations were recorded along with the shows but it doesn't apply so much when shows are computer generated and optimizations are built into the playback software. For what the test pattern was intended it makes perfect sense. Input NEVER equals output when it comes to scanners. There is always compensation and it is either BUILT IN to the show (which is what the ILDA test pattern is for) or it is ADDED DURING PLAYBACK, which is what all modern software does. Some people attempt to combine the two or expect that the ILDA pattern applies to shows that were created without optimizations built in. The end result is typically confusion and disapointment.
    Thanks for the input and I agree to a point long past in time where open loop- and torsion based systems were in and the scanners had no clue of where the hell they were. I am fully aware of the physics involved in electro-mechanical, torsion- and non-torsion based, open and closed loop positioning systems and exactly for this reason I question this out of date method of system tuning. Tuning a system in this manner drastically reduces bandwidth which results in the necessity of compensation via software or hardware. I use an analogy to speakers to clarify the problem at hand. Before I can optimally tune such a system I have to determine parameters such as resonance, the area of linear bandwidth, acceleration, deceleration, just to mention a few. IMO the ILDA pattern is not laid out to accommodate tuning to the best possible linearity and efficiency of modern day scanning systems. I believe each system should be tuned to optimally reproduce a bandwidth within its capabilities which will make the system less static and allow greater dynamics resulting in a broader ability to scan signals in less dependence of their origin. Reproduction of the ILDA pattern via scanners, where a requirement is that the circle be inside the square, even though the input signal says outside of the square is a distortion of the image and requires that the scanner not reproduce what it is fed. Absolutely absurd! Sorry, that is crack pipe physics and a revision is long overdue.

  6. #16
    Join Date
    Jun 2010
    Location
    Zweibrücken, Germany
    Posts
    605

    Default

    Quote Originally Posted by dnar View Post
    Don't forget the invert-X and invert-Y variants of this show!

    Do the full broadcast test tone, 1kHz on left, 1kHz broken on channel right.
    Hey Wayne,
    great idea but we have decided to go all out and not shy any costs so we acquired the original 1969 CBS sound track to their television-test-pattern and have it digitally remastered in absolutely best mono SACD quality.
    Yeah.. now your lookin dude... a real jaw dropper huu?

    Ah.. don’t miss out folks.

  7. #17
    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
    10,114

    Default

    Quote Originally Posted by Solarfire View Post
    Thanks for the input and I agree to a point long past in time where open loop- and torsion based systems were in and the scanners had no clue of where the hell they were. I am fully aware of the physics involved in electro-mechanical, torsion- and non-torsion based, open and closed loop positioning systems and exactly for this reason I question this out of date method of system tuning. Tuning a system in this manner drastically reduces bandwidth which results in the necessity of compensation via software or hardware. I use an analogy to speakers to clarify the problem at hand. Before I can optimally tune such a system I have to determine parameters such as resonance, the area of linear bandwidth, acceleration, deceleration, just to mention a few. IMO the ILDA pattern is not laid out to accommodate tuning to the best possible linearity and efficiency of modern day scanning systems. I believe each system should be tuned to optimally reproduce a bandwidth within its capabilities which will make the system less static and allow greater dynamics resulting in a broader ability to scan signals in less dependence of their origin. Reproduction of the ILDA pattern via scanners, where a requirement is that the circle be inside the square, even though the input signal says outside of the square is a distortion of the image and requires that the scanner not reproduce what it is fed. Absolutely absurd! Sorry, that is crack pipe physics and a revision is long overdue.
    There is MORE to the pattern then the circle in the square, which is designed to measure the 3dB down frequency response at small angle. You have to increase speed to the point that the circle collapses into the square, and it is impossible to build a mechanical angle based scanner that will not do that. There is a reason I have 12 test patterns to tune with, Laser Media clone for linearity and speed matching, the grids for linearity, the circles for color, and a zig-zaq and sqaure wave pattern for initial tuning with a oscilloscope.

    Lets not shoot down the ILDA pattern just yet, at least until we have a replacement that lets you bring galvos up from scratch.

    In fact one of the reasons I still own a "point dumping" system is for tuning, some of the more modern vector based software just might not let you get max speed out of your system, and I have friends that complain about "matching" for their multi-track shows on this software,

    Steve

    Steve

  8. #18
    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
    10,114

    Default

    Quote Originally Posted by Solarfire View Post
    Thanks for the input and I agree to a point long past in time where open loop- and torsion based systems were in and the scanners had no clue of where the hell they were. I am fully aware of the physics involved in electro-mechanical, torsion- and non-torsion based, open and closed loop positioning systems and exactly for this reason I question this out of date method of system tuning. Tuning a system in this manner drastically reduces bandwidth which results in the necessity of compensation via software or hardware. I use an analogy to speakers to clarify the problem at hand. Before I can optimally tune such a system I have to determine parameters such as resonance, the area of linear bandwidth, acceleration, deceleration, just to mention a few. IMO the ILDA pattern is not laid out to accommodate tuning to the best possible linearity and efficiency of modern day scanning systems. I believe each system should be tuned to optimally reproduce a bandwidth within its capabilities which will make the system less static and allow greater dynamics resulting in a broader ability to scan signals in less dependence of their origin. Reproduction of the ILDA pattern via scanners, where a requirement is that the circle be inside the square, even though the input signal says outside of the square is a distortion of the image and requires that the scanner not reproduce what it is fed. Absolutely absurd! Sorry, that is crack pipe physics and a revision is long overdue.
    There is MORE to the pattern then the circle in the square, which is designed to measure the 3dB down frequency response at small angle. You have to increase speed to the point that the circle collapses into the square, and it is impossible to build a mechanical angle based scanner that will not do that. There is a reason I have 12 test patterns to tune with, Laser Media clone for linearity and speed matching, the grids for linearity, the circles for color, and a zig-zaq and sqaure wave pattern for initial tuning with a oscilloscope.

    It would be difficult to get all those issues into one pattern.

    Lets not shoot down the ILDA pattern just yet, at least until we have a replacement that lets you bring galvos up from scratch.

    In fact one of the reasons I still own a "point dumping" system is for tuning, some of the more modern vector based software just might not let you get max speed out of your system, and I have friends that complain about "matching" for their multi-track shows on this software,


    If we need ANYTHING right away, it is a means to match diode based projectors for color response. Note I am hesitant to say color temperature, for that does not cut it with these beasts.

    Steve

  9. #19
    Join Date
    Aug 2008
    Location
    UK
    Posts
    5,704

    Default

    Yeah that's why Pangolin software comes with 10 different test patterns.

    Some just test colour and modulation but others test other geometric aspects.

    That said, there maybe other patterns out there that test other aspects as Steve has found.

    For anyone who's never checked the standard patterns, here they are (2nd one in is ILDA but small in the screenshot):

    (BTW I imagine these are copyright to Pangolin)


  10. #20
    Join Date
    Feb 2011
    Location
    New Hampshire
    Posts
    3,513

    Default

    I have a set of EMS 4000s that I have not attempted to tune since receiving them simply because I was told they were tuned at the factory and did not require further adjustment. I thought "fine, leave well enough alone". But, when running through the Pangolin test screens, they are far from perfect and as the scan speed is increased from a minimum of 8K there is a steady and progressive deterioration of the images. There is no sweet spot, no optimal scan rate, unless I want to set that at 8K. These are equipped with the big mirrors, but even so 8K seems REALLY slow.

    Are the EMS in some way, special and do not require tuning? If not, are they in some way different and require special tuning? If not, what was that reference to a tuning tutorial/algorithm, because I better get to work.

Posting Permissions

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