Fade-Out Time Ignored

[novadatalabs]

I am trying to understand the effect of fade-out time on a chaser that uses the same scene in each of two steps. When I use two lights and two scenes, where the first scene sets the first light to full intensity, and the second scene sets the second light to full intensity, then create a chaser that uses those two scenes with a one-second fade-in, a one-second hold, and a zero-second fade-out, each light ramps up to full intensity in one second, stays at full for one second, then goes to zero intensity immediately and stays at zero intensity for two seconds while the other light ramps up to full intensity in one second, stays at full for one second, then goes to zero intensity immediately. The two alternate in this way for as long as the chaser continues looping. Here's a graph of what I see:

Now, when I replace the second scene in the chaser with a reference to the first scene, I would expect to see that one light ramp up to full intensity in one second, stay at full for one second, then go to zero intensity immediately, and then see that same light ramp up to full intensity in one second, stay at full for one second, then go to zero intensity immediately, continuing this repetition for as long as the chaser continues looping. Here's a graph of what I would expect to see:

But that's not what I see. Instead, that one light ramps up to full intensity in one second, and stays at full intensity for as long as the chaser continues looping. Here's a graph of what I see:

That doesn't make sense to me. Is it a bug, or am I misunderstanding something?

My workspace file is attached. Thanks in advance for any assistance others can offer.

[novadatalabs]

Here's my workspace file.

[novadatalabs]

Sorry if I messed something up in my first post. The two thumbnails of Figure 2 and Figure 3 seem to link to Figure 1, even though the links in the message are to the proper files. It appears the proper files are included as full-sized images at the end of my post. (If a moderator or someone else can correct for this problem, I'd be most grateful.)

[mlohrey]

Hi

I could be wrong here and I am sure someone will correct me if needed, but in chases, fading takes into account the DMX value of the light in the next scene. I guess this is the only way to work out the timing for a smooth fade.

So if you have L1 set to 256 in one scene and then 256 in the next, there will be no change if you fade between them

The effect you require means you need to set an intermediate step of setting L1 to DMX value 0.

In your two light example, because Scene 2 doesn't contain L1, then L1 will fade to zero as I am assuming the QLC+ realises that if the light is missing from the scene and should be set to zero.

Massimo will know the technical answers to exactly what the programming is doing here but hopefully I have given you a work around if you are in a hurry.

Cheers

Mark

[novadatalabs]

So if you have L1 set to 256 in one scene and then 256 in the next, there will be no change if you fade between them

That would be the case if the outgoing cue had a fade-out time equal to the fade-in time of the incoming cue. That would be a classic linear cross-fade.

But, imagine that you were manually operating a lighting board with two scene masters, one controlling the outgoing cue, and the other controlling the incoming cue. Pretty much all such boards have the outgoing cue master on the left, where full intensity is at the bottom of its run, and full dark is at the top of its run. Immediately next to that scene master, to its right, is an identical scene master that controls the incoming cue, where the bottom of its run shows the incoming cue at full dark, and the top of its run brings the incoming cue to full intensity.

If you slide those two scene masters together, keeping them side by side as you move them from the bottoms of their runs to the tops, then you get precisely the effect you describe: the lights smoothly blend from their outgoing values to their incoming values. That's what should happen in QLC+ if you set the fade-out time of the outgoing cue to equal the fade-in time of the incoming cue. But, imagine if you first moved the outgoing scene master quickly from bottom to top, before moving the incoming scene master. The lights would all go to dark. Then, as you move the incoming scene master from bottom to top, the lights would smoothly change from dark to the full intensity of each light as it is in that incoming cue.

In other words, the fade-out time should give the same result as moving the outgoing scene master from full to dark in that fade-out's period of time, and the fade-in time should give the same result as moving the incoming scene master from dark to full in that fade-in's period of time.

That's just not what it's doing, however. I think this is probably a bug.

[siegmund]

I agree with you that this is not what you would expect from a traditional lighting desk. Although, QLC+ is not intended to implement a traditional lighting desk in software. So most likely this is just a difference in basic concepts (mlohrey's explanation seems pretty good to me). Still it could be a bug but I'm pretty sure that there are good reasons for the actual behavior!

Regards,
siegmund

[novadatalabs]

I agree with you that this is not what you would expect from a traditional lighting desk. Although, QLC+ is not intended to implement a traditional lighting desk in software. So most likely this is just a difference in basic concepts (mlohrey's explanation seems pretty good to me). Still it could be a bug but I'm pretty sure that there are good reasons for the actual behavior!

If this is designed behavior, and not a bug, I would expect there to be some documentation describing it. Can you point me to any?

[siegmund]

I agree with you that this is not what you would expect from a traditional lighting desk. Although, QLC+ is not intended to implement a traditional lighting desk in software. So most likely this is just a difference in basic concepts (mlohrey's explanation seems pretty good to me). Still it could be a bug but I'm pretty sure that there are good reasons for the actual behavior!

If this is designed behavior, and not a bug, I would expect there to be some documentation describing it. Can you point me to any?

No, not every single bit of the QLC+ behavior is documented.

[mlohrey]

I have played around with your workspace a bit more, and agree something isn’t right here. I tried using the cross fade feature in the cue list widget to try an analyse the situation and it behaves very oddly. I will post some videos when I have time to test further.

[novadatalabs]

I have played around with your workspace a bit more, and agree something isn’t right here. I tried using the cross fade feature in the cue list widget to try an analyse the situation and it behaves very oddly. I will post some videos when I have time to test further.

Thanks, I am grateful for your assistance!

[mlohrey]

Finally got a chance to play a bit more with this problem.

I have created a workspace that seems to show some inconsistencies as to how chasers are handling fades in the scenario of a light set to the same value in two sequential steps.

In the workspace, to keep it very simple, I created two different scenes containing the same single dimmer channel set to 100%.

I created two different chasers containing repetitions of these scenes and then used cuelists to try and help work out what was going on.

The results were quite mixed.

1. Using the linked faders in a cuelist produces the expected results. Success!
2. Stepping through the cuelist when hold is set to infinite shows no fade in or out. Fail!
3. Using a chaser with hold, duration and fade sets doesn't show expect cross fades. Fail!

I resorted to using two 'different' scenes as using the same scene was producing very strange results. I need to experiment more this one.

Single Light Cross Fades.qxw

Single Dimmer Channel

(6.58 KiB) Downloaded 279 times

[novadatalabs]

Good work! That's precisely the behavior I am seeing (and I can reproduce your observations using your .qxw file).

Note also that, if you run either of your chasers while the monitor window shows your fixture, it does ramp up when the first scene is played in either chaser. The two slider widgets make it look like the fixture goes immediately to full intensity, but that's not the case with the actual fixture. But, as with the example I provided earlier, once the fixture reaches full intensity, it stays there.

I tried looking at the code, but I admit it was more than I can handle right now. I appreciate you confirming this as unexpected behavior, and am grateful for your efforts in dealing with it.

[kenact]

I was playing with your workspace, and there may be an easy way around what you're trying to do. In your "L1 Alone" chaser, replace the second "L1 Full" with a blackout, 0 fade in, 0 fade out, 0 duration. That should do what you're expecting.

[maximortal]

I got the same issue and this IMHO is a bug/issue.
also if you say that this software is not a replacement for traditional light desk (and i not agree) from a logical point of wiev this is a bug:
if a trigger action (next step command) triggers two different actions (fade out of scene A and fade in of scene B) booth actions must be done independently, otherwise the option to set fade in and fade out must be replaced by one single option that we can call "crossfade time between steps/scene".
Because the second described behaviour is a total nosense nowdays (it enloger a lot the programming time to create a simple theatrical play and according to what Massimo say about reducing programming on qlc+ 5) this must be considered a bug.
I've tested it on current stable release 4.11.1 and also in the last GIT 4.11.2.89 and booth are affected by this bug.

p.s.: as describe before by some workaround is possible ( dark scenes in between) but is really painful if you consider that a normal play in theater can be made by 60-100 steps