matkeane

I've never actually tried this from a timeline, but 'standby' is listed in the manual (p164 in the 6.1 docs) as one of the available network commands, so it should be possible to send a TCP/UDP output string from a timeline with something like 'standBy true 1000' to enter Standby.

Assuming the forum moderators have no objection to blatant self-promotion, I built an app to help with display layouts like this. Enter the screen resolution and dimensions, the offset between screens (use a negative value for the overlap), and the app works out the stage size calculations for you, ready to paste into Watchout. You can download it from my website: https://matkeane.com/project/watchout-display-builder-softedge-calculator The version currently online only supports display rotations in 90 degree increments, so there's one display in your layout you'd have to tweak by hand. Also working on supporting virtual displays as well as 2D displays.

It's not listed on the VLC features page but, according to recent posts on other forums, version 3 of VLC player does now natively support HAP playback. It seems to work, but with a HAP-alpha file, the edge of the alpha channel looks a bit nasty (flashing pixels) on my machine. Standard HAP files seem to play fine though.

Hi, In the past, my solution to the port closing issue was to create a looped task in Watchout which fired a blank - literally "$0D" - command to the projector every 10 seconds to avoid the connection timing out. My real commands for the shutter could then be sent once when needed and worked reliably and on-cue.

Watchout applies edge blending when Displays overlap on the same Stage Tier. If you move the stacked/overlapping Displays to another Tier things should probably behave as you expect.

I think the comment by Dimi (http://forum.dataton.com/topic/794-feature-requests-post-here/?p=11737) about finding another way to manage feature requests makes sense. There are also some handy tips hidden away in this thread, but without paging through all the comments people are unlikely to find them. Maybe another sub-forum could be opened so that each request could be in its own thread without making this main 'Watchout' forum too noisy. A thread-per-request might make it easier to gauge user interest, and it would also be nice to get an idea of Dataton's feeling about a request - whether it's "yeah, we're thinking about that" or "won't-fix, deal with it". Funnily enough, the idea of a layout tool for Displays was mooted on page 1, didn't get much support, and has now been implemented in v6.2. Meanwhile, the request for a way to organise the Task window has received several positive comments, starting on page 2 of this thread back in 2012...

OK, I'm trying to be positive here, but I apologise in advance if this turns into a rant as I've just spent the day dealing with this and related issues. Can somebody from Dataton please show this thread to the Watchout developers and see if they can come up with a solution to what is a conceptually quite simple problem. The client just wants to transition into a looping animation, then transition seamlessly out again, and they don't want a long explanation of why that's complicated and how we should just avoid it or simplify the animation instead. Like others on this thread, I've tried the various, increasingly convoluted workarounds (DMX commands which trigger MIDI commands which trigger Aux timelines... really?!), but none are particularly satisfactory. A quick opacity tween out of a looping video works as long as the animation is simple enough for it not to show (try using a quick cross-fade from a looping globe animation currently showing the American continent to the exit-animation which starts in Europe and see if anybody thinks that is seamless). Or we can add a cue to a task to jump back to the loop point... and hope that it doesn't flicker - sometimes it works fine, mostly it doesn't and I haven't managed to pin down exactly what the reason is. Conditional Layers actually seem to work the best, but better hope you don't need to manage several conditions simultaneously as there isn't a way to toggle just one condition, you need to set the binary value taking all of the other conditions into account. So a Task which wants to unset a condition needs to know the current state of all the other conditions, and there doesn't seem to be a way to do that within Watchout (OK, there probably is, using an enormous number of tasks evaluating all the possible combinations of layer conditions, but I mean something simple). Maybe I'm abusing layer conditions by using them as boolean variables to track the state of various things in the show. Generic inputs are great, but then they can't be used to enable/disable control cues, so we're stuck messing with layer conditions. So what would be really helpful, would be a new 'set_layer_condition x true/false' command, so that we can just toggle the current value of an individual condition without changing any others. That would already simplify a bunch of stuff, and I could stop writing my helper app just to deal with it. I assume that, somewhere deep in the Watchout code, it's keeping track of each looping media cue and knows when it hits the loop point - it's just not currently exposed to the user at all. Maybe we need a way to get the current state of looping media (0-100%) and, even better, be able to evaluate that in an expression in a control cue. Each time the media reaches the loop point, the pause cue holding the timeline at that media cue would evaluate a condition ('my_generic_input == 1', or 'layer_condition_8 == true') and either stay put and keep looping, or no longer be considered an active 'pause' cue and would allow the timeline to run seamlessly into the exit animation. I dunno, something - anything - other than the current situation would be good! Don't get me wrong - I'm impressed with the major functions being added to Watchout (Virtual Displays, HAP codec, 3D mapping, NDI support) but it more mundane tasks can still be surprisingly frustrating (which reminds me, I have a bunch more comments to add to the never-ending feature request thread!).

Hi Erik, it was a quick test using the Newtek NDI plugin for Adobe software (Premiere and After Effects) and it's quite possible that there is no video output when a timeline is not active (After Effects seems to only activate the external preview monitor when actually previewing). Also I was looking at the NDI input in the Watchout Production software, not the output from a Display machine, which may also make a difference... so I should probably have done more rigorous tests before jumping to conclusions. Will try and test things more thoroughly soon...

Hi, Not a full tutorial... but I downloaded and installed the Newtek NDI tools (https://www.newtek.com/ndi/tools/#download-tools) on my laptop, which allowed me to choose any window as a source for an NDI stream. In Watchout I added an NDI live source, and it detected the active stream on the local network. I dropped the media cue on the timeline and voila... As somebody pointed out in another thread, there are differences with capture card sources - no settings to tweak (that I can see in the NDI software) and, if you pause the source stream, Watchout seems to see it as an inactive stream, and displays nothing - no 'hold on last frame' like the Datapath cards. There are other NDI streaming packages out there though - I haven't tried the Vmix software (https://www.vmix.com/ndi/) yet to see if there are more settings - but the Newtek tools make it ridiculously simple to set something up - I can even stream directly from my After Effects timeline (but I'm not telling clients about that!)

Salut Benoit, But if I rotate my laptop by 90 degrees, it falls off my lap! To be fair, there maybe isn't an easy solution to this... what should the arrow keys do if you rotate a display by 45 degrees? I noticed in Photoshop, if I rotate the working canvas, the arrow keys still move layers along the original, unrotated, image axes, so Adobe side-stepped this issue too.

I'm sure Walter, Benoit and JFK are correct in their interpretation of the original question, but I did just wonder... After a Display has been rotated in the settings, the geometry window remains stubbornly in landscape mode, meaning some mental gymnastics when dealing with warping on a rotated display (hit the left arrow to move a point down, unless the projector is rotated the other way, in which case it's up...!). It's just possible that Alvaro was looking for a way to rotate the geometry view to match the display settings... (maybe I'll add it to the never-ending feature request thread).

Hi, the screenshots aren't showing up for me, but I'm guessing that you are looking at the 3D Displays in the Watchout production software, and not the output from the Display machine/projector itself. If the preview in the production software looks blurry, it's probably the 'quality' setting in the Virtual Display settings, which only affects the preview quality in the Display preview window. Changing the setting to 'High quality preview' should be what you want. The only drawback I can see to activating the high quality previews is, if you have a number of large Virtual Displays, it could affect performance of the production software... I tend to switch them back to low quality after 3D calibration, just in case.

@Scabious: There isn't a lock mechanism as such, but you could move (cut/paste) the displays to a new Stage Tier, and then go back to working on the main Tier. It should make it much harder to move things by accident.

Sorry for hijacking the thread but, having done exactly what JFK suggests in the past - put a bunch of cues in a composition in order to fade them all off at the same time - am I the only one seeing problems when the cues don't all have the same blend mode? The problem I saw was with a composition containing an image in 'normal' blend mode and another layer for the shadow in 'multiply' blend mode. When the composition was placed in the main timeline with an opacity tween, the layer with the 'multiply' blend mode would pop on and off instead of fading. Not sure if it was a limitation of the graphics card in the client's machine or expected behaviour... In the end, I reverted to a bunch of layers in the main timeline and copying the opacity tween between them (which is a pain) but avoided the unsightly behaviour, and I haven't had a chance to test on another machine yet to see if I can replicate the problem.

Hi Erik, I am a little surprised by the assumption that r601 should be used by default by the ProRes decoder. I haven't encoded anything in the r601 colour space since the days of PAL production. Since HD production took over, and content is generally being produced for on-screen or video projection instead of broadcast TV, r709 is - for me at least - the default. Maybe other Watchout users have use-cases that require r601, but I would think that modifying the colour space of user-generated content should be the exception, not the rule, especially if other video formats are being handled in a different way.

An afterthought: Some of the Atomos recorders (can't remember which models) have a built-in waveform monitor. If the Flame has this function, you add some colour bars to your Watchout project and see whether the output levels are actually 'hot' or whether it's more a monitor calibration issue.

My first thought was that you could just drop a black solid over the top of all your Displays and play with the opacity, but that's perhaps not a very elegant solution. An alternative would be to adjust the RGB sliders in the advanced settings of the Display configuration. I think this is normally for adjusting the white balance of different outputs to match, but dropping all the RGB slider to, say, 90% should achieve what you want. And, although it's not a global setting, as it's set on each display, it would be quite easy to adjust all your Displays to the same settings. Matthew

This thread got me curious, so I opened a very simple test show - two image files on the main timeline, one with opacity and scale tweens - and copied the cues into Notepad from Watchout 6.1.4 and 6.0.2. With a text editor that supports Diff, it's easy to find all the differences between the files. The version number at the top is one, but there are also structural differences in the data which probably explain why 6.0.2 won't accept data pasted from 6.1.4 - There is an extra property in 6.1.4 for the Interaction Name feature, for example, as well as an extra data object saving the state of a dialogue box. After removing these extra lines in the text file, the cues paste successfully into WO 6.0.2. So, for simple media cues at least, some search/replace could work, but other media types, compositions or tasks could be more tricky. Is this something people need to do on a regular basis?

If you're using Midi inputs to trigger tasks, you could combine those with a Generic input to make the task trigger conditional. You could add a new Generic input called "midi_active" and then change the task triggers to something like "my_midi_input > 0 && midi_active > 0". You can update the value of the midi_active generic input using a TCP command from your control system or from within Watchout.

I seem to remember somebody asking about this not long after the Stage Tiers feature was first added to Watchout. As a result, the ability to rename existing tiers was added, so at least you can rename them to something like 'unused', 'spare' or just '---'.

I just noticed that, in the chapter on Display configuration, the section on 'Display Resolution' on p77, says: "IMPORTANT: When using multiple outputs from a single display computer, all outputs must use the same resolution". I thought that limitation was removed in 6.1?

Thanks for the insight Mike. I think that when copying-pasting tasks between shows, I wouldn't really expect links to keep working, so assigning a new ID at that point would be OK. Another thing I would find really helpful would be more feedback about a Task timeline. In the tasks window we can see the current time position, but memorising what that corresponds to across dozens of tasks is a bit tricky. I was wondering if, when a Task hits a pause cue - and if that pause cue is named - the cue name could be displayed instead of the current time. That way, glancing at the Task window, I would be able to see that, for example, my "Logo" Task is paused at "Fade-out Logo", and my "Live Camera" Task at "Fade-in Cam", etc. If a task is playing, or paused at a cue without a name, the current time would be shown as it is at the moment.

That's exactly what I do too... but every time I do the 'double-click, ctrl+J, ctrl+C, Esc, ctrl-P, click, ctrl-V' sequence, I think to myself "In the 21st century, shouldn't the software be doing this instead of me?" - plus my pinky finger is getting worn out from using the ctrl key all day! I appreciate all the work that has gone into developing Watchout 6 and solving difficult problems like 3D camera mapping; now it would be great to see some solutions for little UX/UI problems that would make everyday life more efficient while encoding... Just off the top of my head: - A 'stop all Tasks' function, instead of having to create a task with control cues to stop all my tasks (and doing the task name copy-paste manoeuvre for every... single.. task!), for when the client interrupts a rehearsal and you have umpteen different tasks running which aren't all readily visible in the tasks window. Or the ability to add active tasks to the status window? - The ability to relink more than 1 media item at a time when something gets moved. Other software recognises automatically when other media which need re-linking are in the same folder, but with Watchout there's a lot more clicking, searching and waiting involved. - A better way to add easing to keyframes for scale/rotation/etc - A curve editor, or inspector window so that it's easier to get a smooth animation curve, or maybe just a keyboard shortcut to expand the selected tween track in the timeline window to manipulate the values, then shrink it back to normal size so the other timeline layers are still visible.

RoccoDNYC's drag-and-drop suggestion would be great for adding control cues for Tasks. An alternative would be to replace the 'enter the cue name without making any typos if you want it to work' text box with either a drop-down menu with the list of Task names (although this might get unwieldy or slow when you have a couple of hundred tasks), or a text box with a auto-complete function, so you can just start typing the name of a Task and select the correct one from the list. Better still, when selecting a Task in a control cue, link to it using an internal unique ID that doesn't change so that, if a task name is changed later on, it doesn't break all the links.

Hi, In the second part of your question you ask about imports, so I assume you are asking about importing keyframe data while preparing a show. In some ways, it's actually easier to control a media cue's position externally by updating position data live over the network. If you want to paste in keyframes from another application I'm afraid that, at the moment, there's no direct way to do this that I know of. Since you're not afraid of complicated solutions, then try setting up a few position keyframes on a cue, then select them in the timeline and copy-paste into a text document. You'll see a bunch of keyframe data - a bit of scripting should make it possible to convert from your source application's format to the Watchout format. Gathering dust on my to-do list is a script to create & import easing curves to Watchout keyframes but, until I get around to that, maybe you should add your idea to the feature request thread.