Forum

Writer (dx11 texture 2d NRT Advanced) - video playback FPS

Hi everyone!

I need your help :)

I’m trying to export a high quality video installation using Writer (dx11 texture 2d NRT Advanced). I’ve set the framerate that I need: the pause mode on the mainloop works for all animations that I’ve made on vvvv, but the video playback inside vvvv (both with filestream and HAPreader node) doesn’t work.

This way the final exported video has the right speed, but the video playback (mp4 and HAP format) inside vvvv is faster!

In the render preview, when the writer node is on, all the animations are slower and the videos have the standard velocity (they don’t go as slow as the rest of the animations).

Do you know why? There is a way to avoid this?

Thank you very much!!!

Marti

Hi not really.
You can only provide videos as image sequences, since video playback happens usually on separate thread then mainloop

D:

So no way exporting from vvvv?
I have to convert all videos in a sequences of frame and playback them with an LFO using filetexture? D:

If I understood correct, you need to synchronize your hap video with NRT writer. It’s not trivial but with a modification that detect your “write” status you can directly trigger next frame or choose frame with the seeking.

Otherwise, it’s like Anton said. It’s a typical situation.

Hi Yar!
So you suggest to trigger the HapAudioClock pin “next frame”… very good idea!
I have tried like this: I’ve created a lfo with a period=1/n frame per second, then I’ve sent the change output of lfo to trigger the next frame of hap palyback (so a bang for each new frame). The problem is that the change bang is so fast that is always on, and the “next frame” input doesn’t work…

PS With the framerate setted as 60 frame per second (as I would like) it doesn’t work, but with 25 frame per second it works…

This is already a good solution, but if there is a way to maintain 60 frame for second, it’s better :D

Use NRT-Writer output pins, for example, the frame number. You can do your routine on the change.

Also, you can write whole your video with OBS. Use flv container for smoother framerate and convert it later.

Do you mean using the “current frame output” of the writer node?
Also in this way the “next frame” trigger doesn’t work, it is always on (too fast changes).

I need to use vvvv to export… OBS can’t help me :(

Something like this could work for you
изображение

изображение

This is exactly what i have done, but it doesn’t work, the bang of the change node it’s always on.
The problem can be that my Graphic card? Maybe it is not powerful enough?Cattura

As you have an Incremental Mainloop with Writer NRT you would actually want to have a true toggle (“always on bang”) connected to your HAP Player, as you want it to serve a new frame for every new frame you render in vvvv … so that’s exactly what you need. If that pin doesn’t do anything, then this is an issue with HAP Player. I haven’t used writing imagestacks to capture a vvvv project in ages though, is it common practice still?
OBS can work, or for more high quality captures maybe use a capture card.

S advanced also a frame delay internally

The issue it that in the Hap video player, the next frame pin works just with a bang, not with a toggle.

and if you try a simple patch like the one that i’m posting, you can see that the togedge works just one time, becayse the period is too little…Cattura

I won’t to use obs because I automate the writing process, to have different videos of the different part of the flow of the patch, and i don’t want to do it by hand

antokhio can you explain better? I haven’t understood :)

I solved setting (in writing mode) the speed pin of hap node = 1/frame rate!

Thanks to all :)

Yea that’s also a solution think i did quite same when i needed to record mp4’s, about, why TogEdge did’t work, well, NRT mods vvvv main loop, so one frame you write becomes vvvv 1 frame for vvvv, so basically it just writes every frame so basically toggle or bang when it writes it always 1. And TogEdge reacts on 1 to 0 changes.

Everything send with S/R Advanced will be received on next frame for vvvv, so when you do that means it will react one frame behind.