[vade]

I don’t believe you can do both in and out simultaneously on that hardware.

[vade]

Additionally you could BUILD a max patch that records arbitrary Syphon Streams using jit.qt.record and a Syphon Client.

[vade]

The whole point of Syphon is that you don’t need to embed functionality into a host app?

Why not run a Syphon Server in your Max Patch and record it with Recorder?

[vade]

What settings are you using in Syphon Recorder? What compression are you recording to for 8 hours straight, at what resolution and frame rate?

Are you running out of disk space?

8 hours of Pro Res 422 at 1080p 60Hz is 1 TB. Photo JPEG is 1.8 TB, etc.

[vade]

Does this crash happen after a long time of running Recorder? Are you running a host app with vertical sync off? Just curious.

[vade]

Looks like the same crash. Interesting. Thanks.

[vade]

I can confirm this works for me with latest:

[vade]

In your example it will, assuming its set up to capture the OpenGL scene, output only layer 1.

Image 1
Syphon 2
Image 3
Syphon 4

Syphon 4 would output the entire scene. Syphon on layer 2 would output only Image 1.

This totally works. I just checked it.

[vade]

Wow thats nuts. Thanks for documenting, thats really super helpful Roger.

[vade]

I tested DP5 and 6 and things worked fine. When I get home I can update to DP7 and try it out.

[vade]

Max MSP is RTSP capable too.

[vade]

Thanks for following up. Obviously, you might want to ping the authors of Beautiful Chaos and let them know there are issues.

Good luck!

[vade]

Also what codec too.

[vade]

Can you record from other apps, or is this error just with Beautiful Chaos?

Are you running the most up to date Recorder? What frame rate, and size are you recording, and to an internal, or external disk?

[vade]

I haven’t done any direct work with VLCKit but I did look into it. If I recall correctly they have some lower level APIs and higher level APIs that give you a view to draw to.

Considering you need:

* An OpenGL context to initialize Syphon – and preferably one that is owned or ‘shared’ with a GL context used by the host app you are attempting to get frames from

* Either the OpenGL texture ID, or a handle on to the CPU side bytes of the data, where you make a texture yourself to pass to Syphon

I would suggest that you forego looking at higher level convenience API’s and try to find a VLCKit example that

* Draws to an OpenGL context
* Has an obvious Texture upload component, where VLCKit submits CPU side data to be drawn.

You could then more easily shim Syphon into that pipeline, than via Views and scraping, which is slower, less efficient and more prone to weird issues.

[vade]

Why not just de-interlace the video?

[vade]

Alternatively, you have to put in a delay to simulate the next field drawing.

Interlaced video the fields are drawn at framerate * 2.0 intervals – so if your framerate is 25hz, you need to wait 1/50th of a second before you send the second field to Syphon.

What Tom said is true, Syphon isn’t a series of frames, but a single image that is shared. You could update that image 10000nd times and if the other application is waiting, or lazily drawing, you’ll only see the last frame. We don’t keep a ‘queue’ of frames for many reason (performance and memory costs being one), and so theres only one surface.

Try the delay, it might be more natural, but you might also miss a frame or two depending on phasing of QC’s and your applications frame rates.

• This reply was modified 6 years, 7 months ago by vade.

[vade]

Syphon works on any Mac on Any supported Mac GPU regardless, as long as you are on 10.6 +

So yes, that config works.

[vade]

Not to my knowledge. If you are looking to contract a job you should … actually include your email?

[vade]

Use our Quartz Composer screen capture to a syphon server? Use Syphoner?

There are plenty of options.

[Author]

Posts