I'm trying to acheive a rain effect with some movers. Therefore I need a two step chase going up and down, intensity should only be on when going down and thus off when going up again. Preferably, the upward movement should be as fast as possible, downward should be controllable (preferably by fader, but that's not nescessary).
Off course this isn't a problem with a couple of movers, but I want to spread the effect throughout 20 movers and keep it random. Random speeds would be super, but undoable I'm affraid. Is there any way of acheiving such an effect without having to program a chase with loads of steps? It would be like a spread tilt saw having 0 intensity up and 100% downwards.
Oh btw: Pearl 2k4
Rain effect
Moderator: Moderators
Well, you're just trying to be awkward now!!
So I see two 'potential' ways of achieving this.
1st is to overlay two shapes and match up speed and phase. Tilt shape would be some sort of saw as you say and dimmer shape would probably need a custom pattern to only shut down the light output briefly. You could fine tune both shapes to make things smoother - ie. you could do a new saw pattern that 'faded' into its descent and you could smooth out the fast fade in/out of the dimmer pattern. Make sure the patterns/shape params work correctly with the orientation/type of fixture, set both to the same offset (probably a high value) and the same speed. The Dimmer size would need to be 255 and Tilt to suit. Other than matching up speed/offsets, the most fiddly thing here will be to setup the custom dimmer pattern (or shape definition offset thereof) to ensure it shuts off for the 'rise' of the tilt. This is unlikely to be solely at the beggining or end, but more likely something that bridges both.
2nd method is to utilise just one shape. This avoids the problem of syncing up, but does introduce the main limitation of shapes - ie. for any given shape you can only use one pattern. So using a saw pattern and assigning to both tilt and dimmer channel (again assuming phase/pattern type works for fixture/orientation) you could have a shape that snapped to top and faded to bottom. Though this might work, I think there is a big risk of the dimmer snapping to full during the 'reset' of the tilt due to the physical limitations of the fixtures and this would be precisely the thing you want to avoid. It might be possible to cleverly avoid this by setting phase offsets with the shape definition, along with playing with other values such as amplitude.
I think I prefer option 1 along with neat custom patterns.
Note that IIRC after loading a new shape file you must reset the desk.
Good luck! Will be interesting to know if you succeed.
So I see two 'potential' ways of achieving this.
1st is to overlay two shapes and match up speed and phase. Tilt shape would be some sort of saw as you say and dimmer shape would probably need a custom pattern to only shut down the light output briefly. You could fine tune both shapes to make things smoother - ie. you could do a new saw pattern that 'faded' into its descent and you could smooth out the fast fade in/out of the dimmer pattern. Make sure the patterns/shape params work correctly with the orientation/type of fixture, set both to the same offset (probably a high value) and the same speed. The Dimmer size would need to be 255 and Tilt to suit. Other than matching up speed/offsets, the most fiddly thing here will be to setup the custom dimmer pattern (or shape definition offset thereof) to ensure it shuts off for the 'rise' of the tilt. This is unlikely to be solely at the beggining or end, but more likely something that bridges both.
2nd method is to utilise just one shape. This avoids the problem of syncing up, but does introduce the main limitation of shapes - ie. for any given shape you can only use one pattern. So using a saw pattern and assigning to both tilt and dimmer channel (again assuming phase/pattern type works for fixture/orientation) you could have a shape that snapped to top and faded to bottom. Though this might work, I think there is a big risk of the dimmer snapping to full during the 'reset' of the tilt due to the physical limitations of the fixtures and this would be precisely the thing you want to avoid. It might be possible to cleverly avoid this by setting phase offsets with the shape definition, along with playing with other values such as amplitude.
I think I prefer option 1 along with neat custom patterns.
Note that IIRC after loading a new shape file you must reset the desk.
Good luck! Will be interesting to know if you succeed.
Hmm I was thinking of creating a new shape for this, but i can't really figure out the shapefile for now.
The quick reset is optional, if that's a problem the upwards tilt movement could be slow too. Not that neat, but it'll do. I saw an other operator doing this on a gig, but they were using Hog 2, so I wondered if it would be possible at all because it looked quite fancy. After trying some options I couldn't get it to work on a Pearl so that's why I came here to ask
I'll try your options as soon as I've got an opportunity, thanks!
The quick reset is optional, if that's a problem the upwards tilt movement could be slow too. Not that neat, but it'll do. I saw an other operator doing this on a gig, but they were using Hog 2, so I wondered if it would be possible at all because it looked quite fancy. After trying some options I couldn't get it to work on a Pearl so that's why I came here to ask
I'll try your options as soon as I've got an opportunity, thanks!
Actually I couldn't resist the challenge, but I'm having some difficulty.
Seems Dimmer1:3 type shapes will be good with the correct phase. (long on, short off), but I'm having problems trying to achieve a 'saw' shape for the tilt.
The definition of 'saw' in pearl-shapefile speak is, IMO, not actually sawtooth, but triangular. So I tried to create a true sawtooth pattern, but it seems the Pearl insists on 'bouncing' the pattern. I can't get the 'ramp' option to do anything. Maybe this has been discarded as no longer used.
I'll play with it some more.
Seems Dimmer1:3 type shapes will be good with the correct phase. (long on, short off), but I'm having problems trying to achieve a 'saw' shape for the tilt.
The definition of 'saw' in pearl-shapefile speak is, IMO, not actually sawtooth, but triangular. So I tried to create a true sawtooth pattern, but it seems the Pearl insists on 'bouncing' the pattern. I can't get the 'ramp' option to do anything. Maybe this has been discarded as no longer used.
I'll play with it some more.
Ok. I was wrong about the 'bounce' thing, but there is something very odd about the handling of pattern values.
I have succeeded in making this work over the full tilt range and it looks just as you want. I edited Pulse1:3 to give a slight fade in/out. Tilt will fade in from top, travel down and fade at the bottom, then tilt will snap to top in blackout and repeat. (using full 'size', the tilt position should be 50%)
With offsets this works beautifully.
BUT, there is a problem with speed. Any alterations and it goes all wierd, even if defined in the shapefile itself! 64 seems only workable value. I believe this is due to the 'cheat' I have used in the pattern values, but was the only way I could get the full travel. These are undocumented and most confusing.
(following is mainly for Olie
)
The majority of patterns are defined in the range 0-127 and negative values thereof. But in a few patterns the range goes into the 128-255 area and this was the only way I could get it to work.
If the shape were plotted on a graph where vertical axis was amplitude (-127 -> +127) against a horizontal of an arbitary time then this makes perfect sense to work with a 'centered' origin. Yet it does not appear to work this way:
If I plot (hex) 80 thru to 0 and then FF00 to FF80 the effect is identical to plotting 80 thru 0 and back to 80 again.
If I use values outside this range, either positive or negative then I would expect the 'centered' value to need to be either 0 or 100%, but again this is not the case.
To achieve a full tilt and snap-back I plotted ff80 thru ff08 and followed with ff thru to 80 (in testing the same effect is achieved if the latter values are also negative - ie. ffff thru ff08), yet this works with the 'centre' value at 50%. More importantly it will only work with one speed!
So, questions: why the need for the negative values? should the 128-255 range work? why the issues with speed using the latter?
Or, to put it another way, HELP!
@(iCe): I will PM you a shapefile & instructions once I am sure this is the right way to go about it.
I have succeeded in making this work over the full tilt range and it looks just as you want. I edited Pulse1:3 to give a slight fade in/out. Tilt will fade in from top, travel down and fade at the bottom, then tilt will snap to top in blackout and repeat. (using full 'size', the tilt position should be 50%)
With offsets this works beautifully.
BUT, there is a problem with speed. Any alterations and it goes all wierd, even if defined in the shapefile itself! 64 seems only workable value. I believe this is due to the 'cheat' I have used in the pattern values, but was the only way I could get the full travel. These are undocumented and most confusing.
(following is mainly for Olie
)
The majority of patterns are defined in the range 0-127 and negative values thereof. But in a few patterns the range goes into the 128-255 area and this was the only way I could get it to work.
If the shape were plotted on a graph where vertical axis was amplitude (-127 -> +127) against a horizontal of an arbitary time then this makes perfect sense to work with a 'centered' origin. Yet it does not appear to work this way:
If I plot (hex) 80 thru to 0 and then FF00 to FF80 the effect is identical to plotting 80 thru 0 and back to 80 again.
If I use values outside this range, either positive or negative then I would expect the 'centered' value to need to be either 0 or 100%, but again this is not the case.
To achieve a full tilt and snap-back I plotted ff80 thru ff08 and followed with ff thru to 80 (in testing the same effect is achieved if the latter values are also negative - ie. ffff thru ff08), yet this works with the 'centre' value at 50%. More importantly it will only work with one speed!
So, questions: why the need for the negative values? should the 128-255 range work? why the issues with speed using the latter?
Or, to put it another way, HELP!
@(iCe): I will PM you a shapefile & instructions once I am sure this is the right way to go about it.
Oh well, here we go again with the SG.DAT file.
My findings:
1. Patterns: Positive amplitude is from 0 to 127 decimal. That is 0 to 80 Hexadecimal
2. Patterns: Negative amplitude is from 65535 to 65408. That is FFFF to FF80.
3. Patterns: Make sure to to have a number of steps equal to a factor of 8.
4. Patterns: Don't make patterns with an amplitude of more than 127 because it simply won't work.
5. Shapes: the first number indicates the attribute as set in the personality file. The numbers correspond to the letters in the personality (A=1;B=2;...). Usefull for making shapes on prism, focus, gobo etc...)
6. Shapes: Use only 1 shape for each pattern on each attribute and set the amplitude to 40 and the speed to 16. That way you won't have a super-swinging head like it's drunk.
6. Shapes: Don't enter any base value (set the second last column to 0) since you can't alter it afterwards in your show. Shortly said, only set column 6 to "1" and columns 4-5 and 7-8 to "0". You can alter everything afterwards and it's easier to start with zero spread, no repeat, no offset whatsoever.
7. Shapes: Rename your shapes in an ordered way, like "pan Sin", "Pan Saw", "Tilt Saw",...
So Ice, you still have the problem. What you need is a linear path from 100% to 0% and then a snap from 0 to 100%. Since the end of a pattern in the SG.DAT file is also the beginning this can be made by making a linear pattern from +127 to -127. It can be a daunting task to make a fully working pattern (trust me, been there, done that) so here is the code of my pattern (called pulse down):
PATTERN "PULSE DN"
80 7c 78 74 70 6c 68 64
60 5c 58 54 50 4c 48 44
40 3c 38 34 30 2c 28 24
20 1c 18 14 10 c 8 4
0 fffc fff8 fff4 fff0 ffec ffe8 ffe4
ffe0 ffdc ffd8 ffd4 ffd0 ffcc ffc8 ffc4
ffc0 ffbc ffb8 ffb4 ffb0 ffac ffa8 ffa4
ffa0 ff9c ff98 ff94 ff90 ff8c ff88 ff84
PEND
Instructions of use: copy this block in the first part of the SG.DAT file (before the shapes).
Now the shape.
SHAPE "Tilt PulseDwn" "PULSE DN"
6 40 16 0 0 1 0 0
SEND
Copy the shape in the shape part of your SG.DAT file. Be aware that you MAY NOT HAVE MORE THAN 200 shapes. I stress this because it can ruin the whole show in your pearl.
I am still thinking about how you can have the same effect with you dimmer because if you use the same pattern the light will be full on when going to the maximum point instead of being off. One solution would be making another pattern for the dimmer alone where there is some room for shutting off the beam. But let me just sleep on it now (notice the hour).
My findings:
1. Patterns: Positive amplitude is from 0 to 127 decimal. That is 0 to 80 Hexadecimal
2. Patterns: Negative amplitude is from 65535 to 65408. That is FFFF to FF80.
3. Patterns: Make sure to to have a number of steps equal to a factor of 8.
4. Patterns: Don't make patterns with an amplitude of more than 127 because it simply won't work.
5. Shapes: the first number indicates the attribute as set in the personality file. The numbers correspond to the letters in the personality (A=1;B=2;...). Usefull for making shapes on prism, focus, gobo etc...)
6. Shapes: Use only 1 shape for each pattern on each attribute and set the amplitude to 40 and the speed to 16. That way you won't have a super-swinging head like it's drunk.
6. Shapes: Don't enter any base value (set the second last column to 0) since you can't alter it afterwards in your show. Shortly said, only set column 6 to "1" and columns 4-5 and 7-8 to "0". You can alter everything afterwards and it's easier to start with zero spread, no repeat, no offset whatsoever.
7. Shapes: Rename your shapes in an ordered way, like "pan Sin", "Pan Saw", "Tilt Saw",...
So Ice, you still have the problem. What you need is a linear path from 100% to 0% and then a snap from 0 to 100%. Since the end of a pattern in the SG.DAT file is also the beginning this can be made by making a linear pattern from +127 to -127. It can be a daunting task to make a fully working pattern (trust me, been there, done that) so here is the code of my pattern (called pulse down):
PATTERN "PULSE DN"
80 7c 78 74 70 6c 68 64
60 5c 58 54 50 4c 48 44
40 3c 38 34 30 2c 28 24
20 1c 18 14 10 c 8 4
0 fffc fff8 fff4 fff0 ffec ffe8 ffe4
ffe0 ffdc ffd8 ffd4 ffd0 ffcc ffc8 ffc4
ffc0 ffbc ffb8 ffb4 ffb0 ffac ffa8 ffa4
ffa0 ff9c ff98 ff94 ff90 ff8c ff88 ff84
PEND
Instructions of use: copy this block in the first part of the SG.DAT file (before the shapes).
Now the shape.
SHAPE "Tilt PulseDwn" "PULSE DN"
6 40 16 0 0 1 0 0
SEND
Copy the shape in the shape part of your SG.DAT file. Be aware that you MAY NOT HAVE MORE THAN 200 shapes. I stress this because it can ruin the whole show in your pearl.
I am still thinking about how you can have the same effect with you dimmer because if you use the same pattern the light will be full on when going to the maximum point instead of being off. One solution would be making another pattern for the dimmer alone where there is some room for shutting off the beam. But let me just sleep on it now (notice the hour).
Good post!
Damn. Your code is exactly what I tried originally. Now I try again it works, albeit a simplified 32-step:
When I try this in the simulator some strange things still happen - ie. occasional flicks, just as I was experiencing with the values that shouldn't have worked(!) so I am now thinking it is just the simulator.
As for dimming, I also can't see how you can make it work from just one shape due to the limitations of one pattern.
Using a dimmer shape overlayed does the job and here is slightly modified pattern to give fade-in/out: (...could be better!)
and associated shape:
For an improved 'random' effect you could run two sets of these paired shapes at different speeds along with offset, ie. all odd lamps at one speed and all even lamps at another.
HTH
Damn. Your code is exactly what I tried originally. Now I try again it works, albeit a simplified 32-step:
Code: Select all
7f 77 6f 67 5f 57 4f 47
3f 37 2f 27 1f 17 0f 07
0 fff7 ffef ffe7 ffdf ffd7 ffcf ffc7
fbf ffb7 ffaf ffa7 ff9f ff97 ff8f ff87When I try this in the simulator some strange things still happen - ie. occasional flicks, just as I was experiencing with the values that shouldn't have worked(!) so I am now thinking it is just the simulator.
As for dimming, I also can't see how you can make it work from just one shape due to the limitations of one pattern.
Using a dimmer shape overlayed does the job and here is slightly modified pattern to give fade-in/out: (...could be better!)
Code: Select all
PATTERN "PULSFADE"
ff80 ff80 ff80 ff80 ff80 ff80 ff80 ff80
00 30 50 7f 7f 7f 7f 7f
7f 7f 7f 7f 7f 7f 7f 7f
7f 60 50 40 30 20 10 00
PENDand associated shape:
Code: Select all
SHAPE "Dimmer PulsFade" "PULSFADE"
8 255 64 0 0 0 0 0
SENDFor an improved 'random' effect you could run two sets of these paired shapes at different speeds along with offset, ie. all odd lamps at one speed and all even lamps at another.
HTH
Guys, I can't get it to work...
I've copied both patterns and shape definitions into the SG.dat file and placed it on fixture disk 1. Inserted the disk and loaded the shape file via system mode. The tilt shape appears but the movinglights don't work. The dimmer shape doesn't appear at all, what am I missing here?
I've copied both patterns and shape definitions into the SG.dat file and placed it on fixture disk 1. Inserted the disk and loaded the shape file via system mode. The tilt shape appears but the movinglights don't work. The dimmer shape doesn't appear at all, what am I missing here?
I don't know what I'm doing wrong here, but it's just not working. Maybe one of you guys could provide me with a correct sg.dat so I can check what I'm doing wrong here...Who is online
Users browsing this forum: Google [Bot] and 29 guests