Resolume Forum

Hey

I'm a new resolume user and I'm converting my footage with the DXV-codec at the moment. I have some selfmade footage (shoot with a Canon 7d and compressed with h264) in 720p and with 50 fps. When I convert this footage with the DXV-codec I get huge filesizes (300Mb for 15s) and huge bitrates (up to 150 Mbits/s). The Quicktime player isn't able to play this files anymore (after 1-2 seconds the image freezes), resolume can still play them. But I guess not many of them. I tried to compress the files with h264 and other codecs to really bad quality and re-encode them with the DXV, but the files still get that big (and look like shit).

I wonder if this is normal with HD material? Is it bad for the performance of resolume, to have this kind of bitrates and filesizes?

Cheers

Luke

Hey Luke

Without going to deep into what makes DXV different from H264 when it comes to VJ'ing, the big difference is that DXV stores each frame in its entirety. (H264 instead saves file size by copying pixelblocks from other frames.) This makes DXV ideal for VJ'ing, where quickly shuttling, jumping to cuepoints and scratching are often used techniques.

This performance boost comes at the expense of a larger filesize and higher bitrate, especially at the reosultion and framerate you're using. When playing back these files in Avenue, you should get the best performance possible though.

I Hope that makes some sense.

Joris

This makes sense, thanks for the fast response.

I'm quite happy with DXV for my short clips, but I'm finding that they are probably not ideal for longer, linear run style clips. These clips don't get shuttled, they just play for a minute or more. In this case, it sounds like photoJPEG, or even some of the temporally compressed codecs may work better, even if they occupy some CPU.

Comments?

also take a look on the alpha settings.
dxv with alpha makes bigger file sizes!

goto10 wrote:Hey Luke

Without going to deep into what makes DXV different from H264 when it comes to VJ'ing, the big difference is that DXV stores each frame in its entirety. (H264 instead saves file size by copying pixelblocks from other frames.) This makes DXV ideal for VJ'ing, where quickly shuttling, jumping to cuepoints and scratching are often used techniques.

Joris

Just to clarify, while h264 uses predictive interframe compression, if you set the keyframe to 1 while encoding then it will in effect make it an intraframe codec, like dvx, lagarith, dv, huffyyuv, image seq etc... Though i believe the main advantage dvx has is that its made to be run off the gpu, therefore bitrate should only be an issue of drive speed.