Caching from current position

[pechart]

Hi again !

One last thing for today:

Is it possible to do the RAM caching from the current frameslider position forward? pdplayer always caches a sequnce from the start, but often you want to jump to a specific time in the framerange and play from there.

[pdimov]

Hi,

If you set the work area before preloading (with shift+space) only the part of the layer in the work area will be preloaded.

[pechart]

Thanks for the suggestion.
The way we usually review footage is by just dropping it into the player from the explorer. And then often we just want to jump to a specific position. so there's no need for the player to cache from the start. Other RAM players (like RV) do the caching always from the frme indictator and i think this is really the best behaviour.

[pdimov]

It's easy enough to add a "preload from current position" command that you can then bind to a key, allowing the manual open-jump to position-preload workflow; but if you're using automatic preload and want it to intelligently follow the time indicator as you are scrubbing... this would be a bit harder to implement.

[pechart]

That would be better than nothing, of course.

[pechart]

Hmm, as I work more with pdplayer, i really think, the caching needs some work. It's just a little bit to basic, in my opinion. The problem with the simple 'cache-from-start-to-endOfRAM' method is not only the one i mentioned below, but also another one. If I have a sequence of say 1000Frames and run out of RAM at frame 900 it's not possible to plaback the last 100 frames from cahce, as pdplayer doesn't dynamically free up the cache, where it's not needed by now. And this is really a big issue, because it makes pdplayer a little bit useless on longer sequences for me .
Don't get me wrong, we're evaluating it now and like most of the features and things like the pretty fast startup. But this is a big point, don't you think?

[pdimov]

Not sure that I understand; the preload does indeed stop when it runs out of RAM, but if you do play the last 100 frames once, they will be loaded into the cache, and you can then replay them from RAM. You don't have to preload - it's just a convenience. Playback loads into RAM as needed.

[pechart]

Ok, you're right. Of course it loads on playback. Sorry for that!
I think my main problem is, that the precaching doesn't happen on the actual position in the timeline, as RV and framecyler does.

[pechart]

Hi again!

Would it be possible to implement this? I mean, we really don't want to Play back the sequence stuttering when a Client sits beside us. All the other Players (not djv_view) like framecycler, RV and even After Effects in CS6 do this and it is really helpfull.

Thank you!

[pechart]

Hmm, I'm getting a feeling, that the support isn't really listening to this forum? It's a little bit strange, because on the vray side the support is really great and all.
I think support is a keyfeature nowadays and this is one reason why so many people are happy with vray. I thought it would be the same for pdplayer.

[pechart]

Ok, this is the last time I'll write to the pdplayer section of the chaosgroup boards. Maybe someone cares...

So, will there ever be an update that adresses the caching? When the RAM is full the sequence doesn't play realtime anymore. I really can't understand why there is no smart caching implemented, that tries to ALWAYS keep something like 10% of the sequence ahead of the frameslider in RAM.

If this doesn't get fixed we will switch to another player/tool for sure, most likely RV. Or Hieroplayer. Or Scratch Play. Or anything else. So. Anyone? You can contact me via email or PM, of course.

[Svetlozar Draganov]

We are sorry about the delay of our answer - we do our best to monitor all the forums but unfortunately sometimes we miss some threads and it is a good idea to remind us about them.
Feel free to contact us also at our support emails: pdplayer@chaosgroup.com or support@chaosgroup.com

I'll forward your request to our developers and I'll inform you as soon as we have some news about it.