export HDV from FCP 5.1.3 to Compressor for best output
 

Alright I have read and read and tried and I think I'm missing something here. All I'm trying to do is take the video that I shot on my canon A1 and turn this into an SD dvd at the best quality Below is my workfow and specs. Any idea on how I can improve upon this would be great. Im looking for tips and tricks that I might be missing like field dominance being set to something other than progressive etc... or changing the GOP length. Thanks in advance


My workflow

Shot 1080i24f with canon HXA1

dropped into FCP without problems using the Easy setup of HDV1080p24 edit native HDV

exporting using compressor and choosing the "Using DVD best quality 16:9" preset

Questions so far:

When using this preset is there anything I need to tweak to optimize the quality of the ouput?

Also once I output using this setting and drop this into DVDSP wont an additional compression get added, if so what format is the absolute best to get this to DVDSP?

Now when I do this I end up with my audio files and my m2v that has a res of 720x404..is this caused by the rectangle pixels scenario and when ouput it should be at 72x480. I feel as if I'm doing something wrong and would really appreciate any suggestions in regards to this. Thanks again and have a good one.

All of your steps are right while working with your 24F footage in Final Cut. Before I can give you my tips on how to tweak the settings in Compressor I need to know how long your video is. The duration has a lot to do with what settings can be changed.

Hugh

Hugh,

Thanks for the reply. My video is 4 minutes long. Thanks again.

Jourdan, you're doing everything pretty much right.

When you export an m2v out of Compressor, that's exactly what DVDSP is looking for and will NOT recompress it further.

There's only two things that you could do:

1-Bump up the bitrate for the video. Apple default is 6.1mbs, I believe. You can push to 7.1 average with 8.1 peak, if you're also encoding AC3 audio.

2-For HD originated material, I often add a a sharpen filter to the Compressor preset. It's a couple tabs over on the preset inspector...you'll see a list of possible filters to add in a text list. Down the list is "Sharpen". Go easy...between 2 and 3 is often enough. I won't go into the reasons why this can help your footage here, maybe do a search on my name and downconversions and you can catch my downconvert manifesto elswhere.

Here's Nate's original "On the subject of HD Downconversion" thread. It influenced my workflow approach greatly.

http://www.dvinfo.net/conf/showthread.php?t=63494

Thanks alot for taking the time to write guys. I really appreciate it.

Nate,

I went ahead and read your post on using 10 bit uncompressed when working in the timeline. This may sound like a dumb question but ...in that method you are using hardware rendering correct. Because I have sent my sequence from FCP to compressor and used compressor convert it to an HDV 10 bit uncompressed and it looks amazing from what I can tell, I have a 2Ghz iMac G5 with 2GB of memory and an external firewire drive which simply isnt cutting it anymore. But that aside when using the 10 bit uncompressed method stated in your post you are using hardware to render correct? Please correct me if Im wrong because that post was very informative and if I'm going to get a new machine that sounds like a good workflow in fact here is what I was thinking..


Proposed workflow:

Shoot 1080i24f

convert HDV to HD-SDi via HD CONNECT LE

import via blackmajic HD SDI card and work uncompressed using hardware

I might even have an extra step in there but thats just what I was thinking. I dont know that I even need to work in HD SDI but it was just a thought. Anyhow thanks again for the previous response.

I am a artist that works in video and I have found my needs are usually different than your typical film or video maker. First off, my videos are almost always less than 5 minutes long and sometimes as short as 17 seconds in duration. This being said my encoding settings are usually different from Compressor's presets, although the presets are often a good place to start.

After editing my video in Final Cut I always export directly to Compressor. I have read that people sometimes export their video as a "Quicktime Movie" first, but there seems to be a distinct advantage to exporting from Final Cut right into Compressor. Before every cut in your sequence being exported from Final Cut Pro Compressor will add a compression marker automatically (adding an "I" frame to the GOP structure of the encode). This is best since an "I" frame contains the most information in the GOP structure. The more "I" frames, often the better your image quality, but the larger your file size. With a short video file size is not a concern.

Once I have exported my HDV 1080p24 sequence from Final Cut to Compressor I often start with the "DVD best quality 16:9" preset. The Inspector will recognize the footage as "Progressive" and should display the correct frame rate as well. From the Inspector's "Video Format" menu you can choose "Automatic (SD)", but then you will not have a whole lot of control over the encode quality. I choose "NTSC" and make sure the frame rate and field dominance are set to match my Final Cut sequence (In your case: Video Format "NTSC", Rate 23.98, Aspect Ratio 16:9, Field Dominance "Progressive").

Then I click on the "Quality" tab in the Inspector and set the Mode menu to "Two Pass VBR Best". For the best possible quality you want the highest possible bit rate, but if you go too high DVD Studio Pro will flag the bit rate when you "Build and Format". I have found that if I keep the Max Bite Rate at 8.9 or slightly lower then I rarely have this problem (again, these settings are for short videos, approx. 0-10 minutes). As you move the sliders around you will see that there is a relationship between the Bit Rate slider and the Max Bit Rate slider. I usually set the Max Bit Rate as high as possible (approx. 8.9). Then I move the Bit Rate slider as high as it will go before pushing the Max Bit Rate slider further up. The Motion Estimation menu should read "Best".

Next, I click on the GOP tab of the Inspector. I have found that changing the GOP settings can have a dramatic effect on the quality of the encode. For a short video where file size is not a concern I set the GOP structure to "IP" (leave it "Closed"), and I change the GOP size from 15 to 6. The point of all this is to have Compressor create as many "I" frames as possible. Generally, the more "I" frames the better your video will look. The "IP" structure does away with the "B" frames and replaces them with "P" frames (better quality than "B" frames). Next, by changing the GOP size from 15 to 6 you are increasing the number of "I" frames by 2. Again, changing these settings does wonders to image quality, but you will pay a price in file size.

Having made these changes to the Inspector I save the settings as a Custom Preset for future use. Once I have chosen the "Destination" from the "Batch" window I am ready to "Submit" the batch for encoding.

Once the batch has been encoded to Mpeg2 and whatever audio format you have chosen you simply Import the files into DVD Studio Pro. In DVD Studio Pro make sure that your project is NTSC and has a 16:9 aspect ratio. After you have created your DVD menus, placed your video into a track, etc. you are ready to "Build and Format".

These are the settings I use and have had great result with so far. I am no expert, but have experimented and researched encoding techniques enough to have found a solid combination of settings for encoding short videos.

Hope this helps,
Hugh