Neat Video & Neat Image community

Hello.

As i may have a problem, i would like to know the following about the programmation of the VDub filter.

As parameter of temporal is a radius, i assume, if set to 2, you use, if you are in frame N, frame N-2 to N+2.

For doing this, before there was only one way, now there are two :

1 : Use the "old" way, setting a LAG in the filter, and using/handling frames in an internal buffer.
2 : Use the new way, using prefetch2 function (with prefetcher->PrefetchFrame() inside) and getting frames in the run through *fa->mpSourceFrames[]->mpPixmap.

As i encounter unexpected and strange behavior between some of my filters and method 2, i would like to know what method you use, according your answer, i may have or not have some troubles...

We use the method 1 for compatibility with old versions.

Vlad

Warning, in case you want to change, take a look at the problems i encounter here : http://forums.virtualdub.org/index.php? ... 7c41297536

Things evolved and are solved. My problem was somehow tricky and very specific to my cases.
I suggest you (or developpers) to take a look at the forum link i've posted before.

Apparently, using new method should produce some improvement.

I think it could be interesting to update to new process. Compatibility with old versions... VDdub is a free program, and last version can be downloaded without problem, so i personaly thinks, it's not realy a viable reason...
And... i have felling, that LAG support may be deprecated in the, not near, but not far future... Pure speculation from me, but it seems to be the way VDub appear to evolve (now, no internal filter use LAG anymore).

So, i'm also seeing this post being in "Improvements, new features suggest a way to improve Neat Video"...

Thank you for the links and comments. We will certainly keep that in mind when working on updated versions of the plug-in. Your findings will help us.

Thank you,
Vlad