3 replies to this topic

[konabrad]

I know that the "jittery video" issue has been discussed in various ways, but I can't seem to resolve my basic issue.

I recorded video in DVCAM (Sony's DV format) in .AVI format. I previewed the recorded video (tape still in the cam) on a monitor using "component out". The results were fine. Next, I captured (firewire) this same video segment using the Roxio "capture" feature. The captured file was saved as .AVI (uncompressed). Now I previewed the file using Windows Media Player. The video was generally ok except when the camera panned, then the panned area became jumpy or jittery. This did not happed when previewing the video before capture.

Since this is uncompressed video, I wouldn't expect to see this. I don't know what happens within the Roxio environment, but may be the ability of my computer and video card to handle the process. I don't think "rendering" (hardware or software) is an issue yet as I haven't gone beyond the capture stage. My video card is an ATI Radeon X 1950 PRO with 512MB memory and has the latest driver.

Anyone had a similar experience?

[ggrussell]

This has nothing to do with compression or even file format.  It has more to do with viewing interlaced video on a progressive device like your monitor.  Almost all camcorders record video using interlaced video.  Yes, there are exceptions like those that record in 720p and the more expensive 1080p.  In general, most are 1080i regardless of format (AVI, MPEG 2 or AVC, etc.).  The best solution is to not pan really fast while shooting.  Depending on how the final output will be viewed, you could deinterlace the output which may help some.  If the final output will be viewed on a TV which is also interlaced, you shouldn't need to do anything.

Edited by ggrussell, 05 September 2009 - 06:52 PM.

[konabrad]

The fog has lifited. I ran some quick tests to see what the interlaced vs. progressive issue is doing on my monitor and then on my TV, after rendering and burning. Depending on how I'm viewing the captured interlaced video on my monitor, it can be "jittery" or "smooth". However, the burned disc is "smooth" which of course is what matters. I also tried two burning tests; one with Hardware rendering and one with Software rendering. The results appear to be identical. Is there a preference of one over the other when both appear acceptable? For example; does one take longer to render than the other?

Thanks once again for your help Gary.

[ggrussell]

The fog has lifited. I ran some quick tests to see what the interlaced vs. progressive issue is doing on my monitor and then on my TV, after rendering and burning. Depending on how I'm viewing the captured interlaced video on my monitor, it can be "jittery" or "smooth". However, the burned disc is "smooth" which of course is what matters. I also tried two burning tests; one with Hardware rendering and one with Software rendering. The results appear to be identical. Is there a preference of one over the other when both appear acceptable? For example; does one take longer to render than the other?

Thanks once again for your help Gary.

Glad to hear the final output to disc plays smooth as expected.  Some DVD players can convert interlaced to progessive on the fly when connected using RGB and I'm sure the players with HDMI do the same thing.  Most camcorders shoot in interlaced so keeping everything interlaced is much easier for me. Like I said, the best solution is not to pan so fast while shooting.  Sometimes I just cut those out unless I'm doing it for FX reasons.

Hardware vs software render:  Hardware may be a bit faster depending on the transitions/effects you have used.  The 3D transitions/effects are only available with hardware render.  These use the 3D ability in DirectX so a good graphics card is needed.  The way I understand it, only the 3D transitions/effects will be rendered using the graphics card, but not the video.  This has change in Creator 2010 which now supports nVidia CUDA and ATI Streaming for rendering video which definitely speeds up video rendering.