Workflow for compression to MPEG2 and for rendering

[Michael Sacci]

[Walter Miale] “a) Disenable all the clips (with a Control click) “
Disable all clicks? why are you doing this, are you making sure everything is unrendered? You can bypass this step.

[Walter Miale] “b) Make sure the Sequence Settings are set to ProRes 422, and Video Processing to “Render 10-bit material in high-precision YUV “
This is fine for rendering but has no bearing on Export with Compressor. This setting should match your prores clips that you have in you timeline.

c)
what you have listed with give you the best quality encode. The down side it ties up FCP and Compressor (not so with FCS3). It is better to NOT have the timeline rendered. It is sending a uncompressed frame so all the fx never get rendered down and compressed before the m2v. As for CBR, if and only when you are using a high bitrate there is no reason to use VBR. VBR is to boost quality at low bitrates, below 5Mbps. If you are able to go above 6 you end up wasting time and it is not as good. The time saving is huge also, especially when going from the FCP timeline.

[Walter Miale] “2) If I want to render the seq without compressing it, is it the same workflow except that one would render the seq instead of sending to Compressor it? “

You normally set the render to match you sequence setting or else you will be rendering every clip. Rendering applies compression unless you set it to Uncompressed. Normally your sequence setting should match the clips and the rendering should match the sequence. If you are mixing clips, the sequence should be set to match the higher quality or majority of footage. Of course to every rule there are exceptions.

[Walter Miale] “3) If I have a sequence of dv (sd) clips with sequence settings of dv. “
Sorry there is no question here.

[Walter Miale]

Thanks so much for your informative post.

Here’s the issue for me: I captured sd dv footage and edited it mostly in timelines set for DV-NTSC, sometimes adding heavy fx. Can I somehow enjoy the advantage of retroactively (as it were) editing in a 422 space, as though I had captured directly into ProRes 422?

The points are 1) to edit and apply fx in a 422 space to dv footage already edited with fx applied in a timeline with dv settings; and 2) not to re-digitize unnecessarily.

That’s why I thought I should “unrender” the clips.

This applies mainly, but not only, when I want to eventually send to compressor for encoding in DVD.

Walter

[Michael Sacci]

Okay, I think I have a clearer picture.

First for m2v encoding from the timeline there is really no need to change the render settings since it is bypassing these.

For other uses it depends on what you are doing to. If you are going back to DV tape there is no gain. If there are a lot of heavy fx and text then rendering or exporting to a 10-bit codec and 422 space space has it’s benefits. You can change the render codec of a sequence at any time so there is no reediting. Duplicate the timeline and change the render to ProRes if that is what you want, export a sample clip with and without the new setting and see if it is going to be worth the extra time.

[Walter Miale]

Let me see if I understand. My working timelines are full of dv sd footage. I did quick and dirty renders as I was working. So before duplicating the timeline, is this it: dis-enable and re-enable all the clips (to “unrender” them) so that re-digitizing takes place in a 422 color space, and convert the timelines to ProRes before either exporting to QT or Sending to Compressor for m2v encoding, right? Is this equivalent to having captured to ProRes originally?

If the only re-digitizing (or rendering) takes place in Compressor’s encoding for m2v, (or in the export to QT) does it then not matter if the un-rendered sequence was edited in a dv timeline or a ProRes timeline?

I noticed in some tests that when I exported “unrendered” sequences to QT using ProRes, i.e. without giving them a final render, the image seems a bit worse than exporting after rendering. Is that what you would expect? Or was this anomalous?

Thanks again.

Walter

[Josh Aderhold]

Compressor is rendering everything in a new unit of time. So I think the point of not rendering is that it’s not bypassing areas that may already be rendered. I was under the impression that once a timeline is sent to compressor, all the rendered files are “discarded.” They aren’t actually discarded, but for the purposes of encoding in Compressor, they are ignored.

What I understood from Michael’s post was that if you dropped everything into a new timeline with a Pro Res codec set to that timeline, you may see the results you’re looking for. I would try exporting to a self-contained quicktime file and then go into compressor. Then as a comparison, go straight from your new timeline into compressor. Compare the two results. If you think you’re getting better results in your quicktime file by rendering, go ahead and do that for the quicktime version. But I would avoid doing that when you go straight into compressor.

I have found the best way to seek results you are looking for is to troubleshoot everything yourself. Do it in small portions to minimize your wait.

MacBook Pro
Intel Core 2 Duo
2.6GHz
L2 Cache: 6MB
4GB Memory
Bus Speed 800MHz