dvangennip / web_remote_for_obs Goto Github PK

• Would require some way of selecting an item from a list.
• Edit pane UI (see GetSceneItemProperties for a range of options to edit)
• Set and get settings from OBS

Example: Camera Control Unit elements are not useful to most users. So instead of always showing this, let users enable CCU. When enabled, use a pane below the main UI dedicated to CCU elements.

For other additions, a similar approach may work. Scene switching and audio could be considered examples of such modules. Settings panes are perhaps another, always-there instance as well.

In terms of overall UI, there must be a way to make managing and moving between them easy. Perhaps status bar has to become a fixed bar to the side for quick access. Needs visual mock-ups.

As a solution, push label and value text over the side edges?

• once done, remove/pause/deprecate update_on_interval

use .source-active css class for one to avoid conflict?

Depends on that functionality becoming available in obs-websocket plugin (planned for v5)

• Create UI that incorporates actual volume next to volume fader.
• Register to listen for input volume events
• Show volume in UI

Not updating properly as it's tied to per-scene visibility, and some (audio) elements can go across scenes.

• Needs refresh on any scene changed event
• active state may adjust after visibility changes, so needs to be checked as well

• Wait for first stable release of obs-websocket-js built for v5
• Check changed events and requests

These classes are very, very similar so should have one parent class, or let two inherit from one to reduce duplication in code.

currently assumes a standard 16/9 aspect ratio in .scene class

• Enable use of Fader, Slider, Knob elements (see #11 )
• Improve layout of status bar on smaller screens
• Make scene preview size adjustable (see #35 )

• Adopt modern JS import and module export approach
• Split classes into their own module files
• Makes it easier to integrate CCU branch into main as it can be a separate module (see #15 )
• See also #22 (would benefit from doing this)
• Update README.md to reflect need for running a (localhost) webserver as modules bring CORS requirements

I'm connecting to my OBS instance through an NGINX server and don't see anything in the VU section of the audio devices.
I tried multiple sources on different volumes but it always stays blank.

• Update SourceFilter class to work with v3
• Include default settings into SourceFilter class
• Filters are added when this happens in OBS (event-based)
• Filters are removed when this happens in OBS (event-based)
• Filters are reordered when this happens in OBS (event-based)
• Filter visibility changes when this happens in OBS (event-based)
• Ensure filter settings are updated whenever OBS updates (no suitable events, so must do polling)

Not needed by default so best separated.

• Consider CCU/preset interface for NDI cameras

Related projects and thoughts

• use https://packetsender.com/ for testing commands
• Using VISCA over IP:
• see: https://github.com/misterhay/VISCA-IP-Controller
• see: https://help.ptzoptics.com/support/solutions/articles/13000077734-an-introduction-to-ip-control-scripting-for-ptzoptics-cameras

PTZoptics python examples look really simple:

 s = socket.socket(socket.AF_INET, socket.SOCK_DGRAM, 0)
 s.connect(('192.168.100.81', 1259))
 data = bytes.fromhex('8101043F0201FF')
 s.send(data)
 s.close()

Tasks

• build proof of concept UI
• build working code around websocket server connection in JS because VISCA can't be used over websocket directly, so it needs a proxy

Currently, audio inputs deleted or created in OBS are not handled properly, so UI goes out of sync.

• Subscribe to events
• Remove audio elements if destroyed
• Add audio elements if added

Currently, there is no way to see or adjust transition settings.

• Add a slider to this effect in the status bar?
• Work out reasonable min and max sizes.
• Double-click to reset to default?
• Store this so it's available on page refresh

Also, implement reasonable defaults on smaller devices:

• on mobile
• on tablet

Currently, no indication is given about connection and authentication success or failure, which makes it hard to troubleshoot for users.

Needs better styling

• Status bar edit buttons
• Filter settings select elements
• Source items

Check for wakelock availability in navigator and also listen for visibilitychange events to re-request.

Leads to undefined behaviour and very likely wrong outcomes, so obs-websocket recommends halting updates until changeover is complete.

• Listen for collection changing and changed events
• Listen for profile changing and changed events
• Implement variable to track state, like this.is_changing_x
• Ensure requests and polling are halted until tracking variables are all false

Sliders are blue by default but better to match the overall colour scheme that matches the state of that audio input. Unless we map the colour to the filter enabled state.

Get data from ListOutputs command.

Typical outputs:

1. name: adv_stream, type: rtmp_output => Streaming output (may only be listed when streaming is set up)
2. name: adv_file_output, type: ffmpeg_muxer` => Recording output (always available)
3. name: virtualcam_output, type: virtualcam_output => Virtual camera
4. name: NDI Preview Output, type: ndi_output => NDI output
5. name: NDI Program Output, type: ndi_output => NDI output

• First two outputs are already covered in status bar (using dedicated info requests and events).
• Add virtual cam to status bar as well (using its own info requests and events).
• NDI outputs cannot be started reliably from a remote connection, so only list their status?

Currently, failure to start a stream isn't made clear in any way.

It looks like obs-websocket v4.9.1 does not return any info from which errors can be deduced (other than knowing if we're steaming or not), so this one will have to wait until more info is available. v5 will give a outputActive response on ToggleStream request.

add vectorscope based on https://github.com/mikkab/vectorscopeJS?

Will go into outputs status pane.

In the webinterface you can only set a max of 0dB gain while in the advanced audio properties it goes way beyond would be great to implement this. Also if a gain is above 0 the sliders actually go out of their box while still showing 0dB as their value.

Currently, there are some elements that are tied to a specific OBS scene setup. This wouldn't transfer to any other scene setup, so it would be good to think of ways to make this more generic.

Current behaviour sets the old Program scene as Preview and desired scene as Program, so the old Preview assignment is lost.

Proposed implementation is to send the SetCurrentScene request wrapped in a batch:

1. Set desired scene as Program
2. Wait for a bit
3. Set old Preview scene as Preview once again