How to Compress TV/VIDEO NOISE without Artifacts??

[Jbertain]

I am exporting self contained DVPRO50 as we speak. But forgive me, I don’t know what you mean by blue channel? And how do I blur it?

j

[Russell Lasson]

In the channels folder, use the channel blur.

-Russ

[Jbertain]

got it. I’ll try burning. Thanks.

[Rob Moreno]

> Exporting a self-contained file at current settings then taking that file to compressor is the > same as sending it directly from FCP to compressor.

Actually this is not necessarily true. When you have several video layers, especially text and graphics, exporting from the FCP timeline will give you much better results because these layers are compressed directly to MPEG-2 from their native format. This is why compressing from directly from the FCP timeline takes longer than compressing a stand-alone movie file; because it’s calculating all text and graphics layers as uncompressed. With a self-contained movie, all text and graphics are converted to DV (if the sequence setting is DV). So when you take this file to Compressor you are actually compressing text and graphics twice. I will admit, however, that the difference in quality is not always enough to justify the increased rendering time. Depending on the complexity of the sequence and the format of the video the difference in rendering time can be drastic, especially when working with uncompressed video.

MediaKobo
Tokyo, Japan

[Gordon Gurley]

Also try tweaking the settings in Compressor. The “Highest Quality” mpeg2 preset is not even close to highest quality. But, of course, when you crank those settings up, it takes MUCH longer to compress.

Gordon Gurley
Director of Operations
Stanford Video

[Rafael Amador]

[rob moreno] “> Exporting a self-contained file at current settings then taking that file to compressor is the > same as sending it directly from FCP to compressor.

Actually this is not necessarily true. When you have several video layers, especially text and graphics, exporting from the FCP timeline will give you much better results because these layers are compressed directly to MPEG-2 from their native format.”
That’s right and this is the reason that if you don’t send it from your time-line, you better export in an Uncompressed codec (8/10b Unc).
As Russel says:

It’s an uphill battle when you shoot on DV to make a great looking MPEG2 file, but you might try applying a 4:1:1 Color Smoothing filter (in the Key folder of Video Filters) before you compress the section. That should help cut down the artifacts.

This is right too but this works if your set your time-line in with a codec better than DV. If keeps in DV all the Chroma Smoothing goes away.
I would recommend you edit your fim in DV and when ready change the setting of your sequence to 8/10b Unc. Then you have the three options:
– Export to Compressor from the time-line. This will take long time because all the movie will be rendered even if you have already render all and you will have working the two applications.
– Once you have rendered all your sequence export a Refference moovie and send it to Compressor .
– Export a “Self-contained movie” and send to Compressor.
Each one got his favarite method.
If I’m not too pressed I preffer the first one. Whatever you have set and how you have set your sequence you are sending a full 444 movie to Compressor, as Rob Moreno says.
Cheers,
rafael

[Rafael Amador]

Sorry I forgot to tell you I reccommend you to try “Chroma Sharpening” of Nattresse. Drop it in your DV footage in a 8/10b Unc time-line and you will get shock seeing how all the blockyness disappear. Make a test with somebody wearing a red T-shirt. The red goes back to the shirt instead of spreading all around. Magic. This toguether with a CC will make your movie look as would have been recorded with a much better camera. Your DVDs will look also far much better.
rafael