Original LXSeries Forum

Hi,

i saw since i'm using Lx consol 1.3.38 some thing very strange:

Some time, but not every time, and in a very confusing way, i have some fade problem when i play a cue. I try to explain: in the live window, i saw the channel running correctly, but on stage, light don't go, and when the time up and down are finish, the new cue is coming like if the time up and down is 0. The stranger thing is: i can play the cue a lot of time and every thing ok, and once a time, it's not running well, and it's always in the same cue, but very rarely, it could happen in a another cue.

What i changed since i saw this problem: I bought a new mac, i'm using Lx consol 1.3.38, and i work in big theater. So i'm very confuse but i think the problem could come from the Mac, or from the power of the DMX signal of the enttec DMXpro because the cable is like 200 meters long, or lx consol. But the truth is i really don't know, and this problem is very confusing me.

Did you know about that? Am i the only one to repport this bug?

Well i'm trying tonight to put an DMX booster. Will see.

Thanks
Cordialement

Tonight i thought about something: i'm unsing 3 lx document during the live. 1 is for the main sequence, with all the cue, and no programme, no submaster, and only one midi command for the GO. One other is for the chase, and i use midi control change in Qlab who controlled channel, and the third is use for a shutter and a ventilator, who are controlled by manual midi controller. I did that to have no problem with marked channel in the main document.
So, i have a chase who is using only one channel, going up and down, with a midi curve in Qlab in the second document. And i think that can give some conflit with the main document, because it's always during the chase is playing that i have this problem.

Is it possible that one lx document can makes some problem with another? Open only one document and program submaster could be better?

I'm doing the show on thusday, i'll try with a new way, with submaster...

I hope it will be fine.

Cordialement

So Alrik How did it work?

Was a programmed submaster within only one sequence to be preferred?

I'm curious to know!



Greetings Johan S.

I'll tell you wenesday,

i'll programme today but i'll test in real monday and if it works, i'll run the show thusday, and i post here my answer.

Hope you're fine.

Cordialement

Does the problem with the fade jumping to completion happen with LXConsole v2.0?

It almost certainly does not have to do with the length of DMX cable.

It may have to do with running more than one fade at the same time. This would be a threading issue which would be why it doesn't happen all the time. It also makes it very difficult to figure out.

It is possible that one thread is blocking the thread that actually sends the output to the interface from running until the fade is complete.

Use of multiple threads makes it possible for different processes to happen using the multiple cores of most new computers. But, it introduces a lot of unpredictable complexity.

I was downloading lx version 2 just now. I'll try that on monday. And yes i'm now quite sure it's nothing to do with the lenght of DMX cable, because i used a booster and the same problem happened...

I'm using the Enttec DMX USB pro.

Try 2.0.1 which is available now as the latest build. It changes the way it prevents multiple threads from reading and writing the same data at the same time. This may prevent a locked thread from failing to update the output during a fade.

Helllo,
i found what make the fade problem, but i don't understand why?










This 4 pictures are to explain what i'm doing, i programmed a chase in Qlab with midi curve, when i push "W" it start, what ever i am in the cue list of Qlab, it's an independant module. With the "go to" cue, it run with out stop. And at this time i can run a cue in Lx, using a note on in Qlab to say GO to Lx, for that every thing works well. (But i can't use other midi control for Lx, but i already explain that in another post). My point here is the strange fade.

Well, with this programmation in Qlab, i push "X" it desarm the "GO TO" cue, so it goes on the "chase down 29", who is a fade out for this channel.
What is very strange and i can't explain is: If i push "X" during a cue is running, it stop the fade (not in the main windows, but on stage), and the cue is jumping to the good value when the time fade is finish.

Well, maybe it's a conflict between Qlab and Lx, i really don't know... Maybe if you need some more information, i can send you my folder of Lx and Qlab...

Thanks again for the work, i hope this feedback can help you to fixe this problem.

It would probably help to send the LXConsole and QLab files. This is a pretty complicated set of circumstances to reproduce.

In LXConsole does the control change set a channel level or a submaster level? You may have better results using a submaster even if it is programmed with just a single channel. Submasters in LXConsole v2 are independent of cues with the output taking the highest level.

Ok, i'm working tomorrow at a theater with a internet connection, i'll send you by email the configuration i use to the technique, Lx consol, Qlab, and Live.

Hope you will find some thing. Because it's a very strange thing...

I had too much work to take the time to send you the document with explication. And now i don't have with me my computer. So i'll send you the document next week.

By

It appears that MIDI control changes can overwhelm the main user interface. When you move a fader or scroll a level using LXConsole, it gets a message every time the mouse moves. LXConsole has responded to MIDI messages as if it was being controlled by the mouse/keyboard. MIDI, however, can send many more control change messages than you would get by moving the mouse or by typing commands into the command line. Hence the backlog where the interface lags behind and/or seems unresponsive when receiving continuous control change messages. Also, some MIDI messages may even have been skipped while others are being processed.

LXConsole v2.0.14 provides some of the possible MIDI action commands with a delayed update of the user interface. This allows quick processing of MIDI control changes without attempting to update the interface each time. Updates are instead deferred and several messages may update the interface at a single time. This makes LXConsole much more responsive to multiple control changes messages being streamed at it.

MIDI actions can execute command line commands. Continuous MIDI control changes can cause choking of the command input. So, using command lines with control changes may not have the best results.

To allow smooth and quick MIDI level control, LXConsole v2.0.14 adds a new MIDI action command of the form "SUB-nn" (where nn is the submaster number). These commands set the level of a submaster from the velocity or control change value of the MIDI message. The result is the same as a command line action, "sub: nn@%v". However, the "SUB-" action bypasses the command line and uses deferred updating so that it is much quicker than the command line version.

LXConsole v2.0.14 also has updated processing for master@, manual@, manup@, and mandown@. This makes the use of a midi controller to run a split manual fade much more responsive than prior versions.

LXConsole 2.0.14 can handle at least 2 simultaneous chases/continuous control changes from QLab that run "SUB-" actions. At the same time, a separate cue can be run using the Go button on a G5.

A G4 can handle two simultaneously moving sliders on a MIDI controller that trigger "SUB-" actions while also running a cue. (Can't run QLab on the G4 to try out the chase but it would be similar to rapidly moving the sliders).

I could not test this new version because i have no DMX for the moment. And for the "strange fade" i need to be connect to a DMX and see the real effect on the light. And i didn't send you the files because i went directly after my last work. I come back the 20 of march and i go to work the 25, so i can try and send you every thing even if it's too late.

Sorry again