Keeper of the Keys Posted December 31, 2008 Share Posted December 31, 2008 So as far as I understand the DMX standard describes 8bit slots of data per channel, but when reading descriptions of desks they are very often talking about how they fully support a 16bit "control resolution", and more recently (the grandMA2) even 24bit so I was wondering are they actually changing the standard or do they just mean that the desk is aware/capable of using multiple (2 for 16, 3 for 24) channels to control a single aspect of a fixture? Thanks and happy new year. Link to comment Share on other sites More sharing options...
johnhuson Posted December 31, 2008 Share Posted December 31, 2008 or do they just mean that the desk is aware/capable of using multiple (2 for 16, 3 for 24) channels to control a single aspect of a fixture? Correct. Link to comment Share on other sites More sharing options...
Keeper of the Keys Posted December 31, 2008 Author Share Posted December 31, 2008 In that case shouldn't that just be a question of the identities/profiles that get loaded for smart lights and not be such a big deal? Link to comment Share on other sites More sharing options...
gareth Posted December 31, 2008 Share Posted December 31, 2008 It's not that great a deal these days (although 24-bit control is a pretty new thing). It's just to do with a desk's ability to 'intelligently' map two (or three) DMX control channels onto one logical attribute/parameter/channel/insert-your-terminology-here. They're not changing the DMX standard in any way - just adding a "feature" to the console that allows increased control resolution within the confines of what the DMX512 standard provides. To be honest, I think DMX's days are numbered in many ways - back when it was first conceived, I don't think anyone had any idea that we'd now be at a stage when a single small RGB LED starcloth could eat up a whole universe of DMX all on its own ... Link to comment Share on other sites More sharing options...
Ynot Posted December 31, 2008 Share Posted December 31, 2008 In that case shouldn't that just be a question of the identities/profiles that get loaded for smart lights and not be such a big deal?Several of the more pro-range moving lanterns have options where you can change the profile to a 16 bit setting - eg many of the modern Martin fixtures can be set to '16 bit extended' which gives a finer control over, for example, the pan/tilt settings, via 2 additional channels (ie 3x and 4x). Thus you set the fixture to the extended mode, then tell the desk (by using the correct fixture option) to use that lantern in that mde - and hey presto. Link to comment Share on other sites More sharing options...
Tomo Posted December 31, 2008 Share Posted December 31, 2008 Actually, the Martin '16 bit Extended' personality/option adds 16 bit control to the Intensity, CMY and/or Gobo Indexing settings. All 'professional' quality moving lights have 16-bit pan & tilt now - it's only disco kit that doesn't. To be honest, 24-bit sounds distinctly like a marketing gimmick with no purpose whatsoever.There's no use for it now, and probably never will be - very few fixtures actually use all 16 bits anyway, most are really closer to 12 or 14 bit. 16-bit is 65535 steps, around 570 deg (moving head) is only 0.0087 degrees of arc per step, which works out as ~15cm at 1km range.If anybody has a fixture bright enough and accurate enough for that limit to be an issue, I want to see it! Link to comment Share on other sites More sharing options...
obsoperator Posted January 1, 2009 Share Posted January 1, 2009 Perhaps it would be useful to the OP to point out that some early consoles with specific software support for moving lights allocated two channels for, say tilt. But while they presented a 16-bit number to the operator (as well as two eight-bit numbers in other views of the data), they actually performed only the same 8-bit crossfades they were designed to do before the moving light software was built onto the older hardware platform. That is, if the light was at 50% coarse tilt, and 50% fine tilt, and a fade to 25%C/50%F was executed on a 25-second count, the fine channel would not vary at all (in the DMX output of the console), while the coarse channel would move 1% each second. Needless to say, this would usually produce a very unattractive, "steppy" motion. One such console was the one I ran for ten years, the Obsession. Of course the quality of the user interface and other factors determine the "real" suitability of any given console. But my reply is meant to address the quoted advertising claim regarding "resolution". Note also that because of the limited DMX refresh rate, and the inertia of the moving light itself, a "full resolution" fade does not actually provide every continuous integral value in the fine DMX channel during the average crossfade. It's physically impossible for most large-distance fades. And it's not necessary either. But sometimes a 16-bit fade can show stepping anyway. Link to comment Share on other sites More sharing options...
Keeper of the Keys Posted January 3, 2009 Author Share Posted January 3, 2009 That does explain the significance of the desk being well aware of the amount of channels, though it still seems that it would be something that is solvable 100% in minor software updates... How come 16bit fades are still choppy, 65535 values for steps seems quite a lot? (24bit would even give 16,7 million steps) Link to comment Share on other sites More sharing options...
Dmills Posted January 3, 2009 Share Posted January 3, 2009 Well, DMX only (assuming a full universe) refreshes at 44Hz or so, and there is no guarantee that the stepper driver logic in the lanterns can actually use that resolution (or move at the speed required, or even work out what speed is required to get where it is going without overshoot due to simple torque limits in the drives). Pan and tilt are **Tricky** control engineering problems, made worse by the rather slow discrete updates in where the lantern needs to be moving to. Consider a very slow pan or tilt, the lantern if running in full step mode may be able to get to the new position sent in each frame before the next frame arrives at the receiver! All it can do is guess based on past history as to how fast it needs to move, and if it is looking at say 8 frames of history to establish speed then it will be 0.4 seconds behind the action. This is why a lot of early movers (even 16 bit ones) had a speed channel so the desks could pass the information that only they had about how fast the movement really was to the lantern. Also consider that a lot of the early control programs did not calculate the values for the next frame based on when it would be transmitted, but instead just updated a transmit buffer periodically and let the DMX hardware pick it up from there, the result? **LOTS** of jitter in the transmitted values during a move as the time the computer was using was not related to the timing of the actual frames hitting the wire. It is not a simple thing to implement (mainly from the lantern side). Regards, Dan. Link to comment Share on other sites More sharing options...
Beachhouse13 Posted January 5, 2009 Share Posted January 5, 2009 All 'professional' quality moving lights have 16-bit pan & tilt now - it's only disco kit that doesn't. Just and FYI, the Mac 2000 Performance II and Profile II both have 8-bit pan and tilt in 8-bit mode. Scary!!! :) Link to comment Share on other sites More sharing options...
gareth Posted January 5, 2009 Share Posted January 5, 2009 Just and FYI, the Mac 2000 Performance II and Profile II both have 8-bit pan and tilt in 8-bit mode. Scary!!! :)Well, yes, but they also have 16-bit mode (unlike the cheap disco tat that Tomo was referring to), and you'd only put them into 8-bit P&T mode if you were either a) seriously short of DMX channels, or b) a total numpty ;). Link to comment Share on other sites More sharing options...
Dirk Posted January 6, 2009 Share Posted January 6, 2009 Actually, the Martin '16 bit Extended' personality/option adds 16 bit control to the Intensity, CMY and/or Gobo Indexing settings. All 'professional' quality moving lights have 16-bit pan & tilt now - it's only disco kit that doesn't. To be honest, 24-bit sounds distinctly like a marketing gimmick with no purpose whatsoever.There's no use for it now, and probably never will be - very few fixtures actually use all 16 bits anyway, most are really closer to 12 or 14 bit. 16-bit is 65535 steps, around 570 deg (moving head) is only 0.0087 degrees of arc per step, which works out as ~15cm at 1km range.If anybody has a fixture bright enough and accurate enough for that limit to be an issue, I want to see it! H'mmm all sentiments of all posts generally agreed with. The ONLY time one might HAVE to have ultra high desk resolution is when driving media servers. But thats another can of worms. 16 Bit pan and tilt is nice only really because it makes fine setting of pan and tilt easier, ie altering "fine" pan or tilt just jiggles the fine channel. And thats presuming, as already pbserved, that the desk processes crossfades properly. And even then.... 16 bit control for CYM etc? Do me a favour! You're telling me the stepper / mechanical resolution is that good, and that repeatable? Well, its not. Old fashioned folk would of course say (as one major manufacturer did too) that 16 bit p/t is not necessary and would never appear on their lamps, and use of timing channels would give a far superior p/t performance, however, I digress... "All 'professional' quality moving lights have 16-bit pan & tilt now - it's only disco kit that doesn't." Oooh - are you saying my friends in the "Disco" world, grossing £1000s every night aren't professional?No, I didn't think you were!!!! I enjoyed that. Carry on! Dirk Link to comment Share on other sites More sharing options...
Tomo Posted January 7, 2009 Share Posted January 7, 2009 16 Bit pan and tilt is nice only really because it makes fine setting of pan and tilt easier, ie altering "fine" pan or tilt just jiggles the fine channel.Nope. 8-bit is 255 steps. For the same 570 deg moving head, each step would be ~2.2 degrees of arc.At 10 metres, that's a step size of ~40cm. In the world of theatre and rock'n'roll, that's simply not acceptable. Arenas quite often have grid heights of 20m or so, then each step moves 80cm to 1m on stage. Try hitting the lead guitarist that way. In a disco, it might not matter - you are often more interested in the movement rather than the absolute positions. -With regards to 16-bit CMY, I agree with you. While the movement system could easily hit 14-bit accuracy, I'm unconvinced that the actual change in colour is useful. Link to comment Share on other sites More sharing options...
Dirk Posted January 8, 2009 Share Posted January 8, 2009 16 Bit pan and tilt is nice only really because it makes fine setting of pan and tilt easier, ie altering "fine" pan or tilt just jiggles the fine channel.Nope. 8-bit is 255 steps. Tomo - I think we are singing from the same hymn sheet. I'm more than aware of the significance of an 8 bit byte. What I was trying to say was, when I hit the "fine" button on my desk, it drives only the fine value (255 times) before incrementing coarse by one... If that makes any sense! Its been a long time since I've had my lil DMX'ter on the output. I must dig it out. Trouble is, I discover horrid things about manufacturers DMX implementations. I shan't give examples to spare the guilt of those concerned! RegardsDirk Link to comment Share on other sites More sharing options...
Beachhouse13 Posted January 8, 2009 Share Posted January 8, 2009 A little OT but since people are talking about the necessity of 16b CMY I have to ask does anybody actually want/need 16b for intensity? Link to comment Share on other sites More sharing options...
Recommended Posts
Archived
This topic is now archived and is closed to further replies.