[Tom Brooks]

[Sean ONeil] “I just played around with it. Do NOT put a DV clip set to “lower field” on a sequence that is set to “none”. Viewing it on my external monitor (via Kona 3) the image is horrible. It appears to be line-doubling one field only while throwing away the other field – which of course equals garbage.

Change the clip to “none” as well as the sequence and it looks great. So I think this is mandatory – not just a good idea.”

That’s not what I’m seeing with DV. Clip captured with normal DV NTSC settings (FD shows Lower), placed on a sequence set to None–looks OK. Clip property set to None, placed on a sequence set to None–looks the same. I’m observing this on a Kona LHe set to 525i (the only choice for SD) going to a Sony PVM interlaced monitor. Is there another possible variable in your sequence settings that is causing yours to look bad?

[Tom Brooks]

[Sean ONeil] “My understanding is that there is no actual difference between 60i and 30p within the confines of the DV NTSC codec. It is always stored as 60i, 25mbps CBR. All that matters is how QT/FCP interprets it during export or playback. Certain “flags” tell it to playback at either 60i, 30p, or 24pA. These flags can be changed – one way to change it to 30p is by changing the field dominance to “None” like I described.

It’s just like the anamorphic flag. It makes no difference if you capture it with or without “Anamorphic” checked in the capture settings. You can change it back and forth after capturing as you like. It doesn’t re-render anything.”

Sean, I’m sure you’ve seen that FCP doesn’t always “store” clips as 60i. 24PA material is captured with pulldown removal in real time and the clips are stored as 23.98P.

It’s true that Final Cut will OUTPUT that 23.98 timeline as 60i by inserting pulldown either in realtime over FireWire or on export. I think you’re right that the key thing to understand is that the output to DV is always 60i because DV itself is a 60i format. You’d have to ask the heads in any DV VTR and they’d tell you, “Yup, it’s 60i.” All the 24P and PA stuff is just trickery to superimpose progressive frames over a 60i medium.

In the case of 30P, though, there will be no difference in the output of captured video whether you change the sequence field dom to None or leave it at Lower. Final Cut knows what Dave LaRonde has said: “It doesn’t NEED to do anything different” to output 30P because the timebase of 30P frames is compatible with 60i. FCP just treats 30P like it’s 60i, no matter what–on output that is.

The sequence field dom set to NONE is to make sure moving titles are made with no fields and to let Compressor know the sequence is meant to be progressive for MPEG-2 that’s going to DVD. We’re starting to sound like a broken CD here.
-Tom

[Tom Brooks]

I don’t use HDV, but that rings familiar. Maybe HDV forces break on start/stop? Should be some info in the HDV forum.

[Tom Brooks]

I agree with you. The only problem with 30P is when it is displayed on interlaced systems. 30P looks great in a Quicktime movie on a computer. I believe it can look good on a DVD if the DVD player can recognize it as progressive and display it as such on its component output.

But 30P on BetaSP? I don’t see how that works–for the reason you state. It’s that point where a field from frame 1 and a field from frame 2 are displayed interlaced. Those fields are from different points in time and will make the comb artifact. (If “Friends” is shot 30P, I don’t know what they do to make it look good on my NTSC TV.)

For BetaSP final output, I’d be shooting 60i. I do hope you get some good answers on this topic.

[Tom Brooks]

Hey vv,
That older 30P thread got a little scattered in my opinion, by the the time it was all said and done. To the point where I’m not sure it is a reliable source of information. Let me clarify the workflow that appears to result in 29.97 progressive frames per second on a DVD.

1. Shoot 30P or produce 30P animation. (It’s actually 29.97 in NTSC land.)
2. Use the normal DV NTSC easy setups for Final Cut Pro.
3. Create a sequence, go into the sequence settings, and set the Field Dominance to None.
4. Edit as usual.
5. Export using Compressor. Choose the normal preset for 90 Minute High Quality encoding. Leave all settings on default.
6. Bring the resulting MPEG-2 into DVD Studio Pro and make your DVD.

When I follow this recipe I can verify that the MPEG-2 file is progressive frames by playing it in Quicktime Player. It has no comb artifacts when viewed on the computer monitor. Likewise, the DVD (created in DVD Studio Pro) has progressive frames. When played on a normal set top DVD player to an interlaced TV, it has no comb artifacts and the pause is a perfect still frame.

It was not clear to me that your camera-original footage was shot 30P. If your original is interlaced, this method will not work. The only way to get progressive from it would be to use some type of deinterlacer on the clips. Simply changing sequence field dom to none won’t do it.

I hope this helps get you where you want to go.
-Tom

[Tom Brooks]

Export Using Compressor will re-render everything according to the render settings in your sequence (high-precision YUV or whatever). This can take some major time when it hits text effects, motion blurs, glows and the like. I’m not sure how to explain the faster render on second attempt. That suggests FCP has cached the render and re-used it on the second attempt. You should get faster results by exporting a QT movie with current settings and then bringing that into Compressor.

[Tom Brooks]

I totally agree with that. The bulging of the center is another trick to make your brain think ‘subwoofer.’

[Tom Brooks]

I’d just do this in Motion (After Effects, better yet) and animate the scale of the shot. It’s probably best done with good old keyframes. Experiment with the amount of scale increase, the number of times it goes up and down, the speed of the up/down, and the shape of the bezier curves. You have to create the look yourself rather than relying on an actual kickdrum sound to do it, although that could be a starting point. I’m guessing that the scale motion curve will look like a sine wave, but the shape of that wave and how it decays will determine whether it looks to your eye the way the drum sounds to your ear.

[Tom Brooks]

Not sure if there is a P2 driver issue in your workflow–but drivers for 16GB are new. Upgrade software avail from Panasonic.

[Tom Brooks]

I see a couple of problems with your workflow. You should stay away from Quicktime Conversion. Try this. Make a copy of your sequence, change the compressor to uncompressed 10-bit, and then export Quicktime Movie, using Current Settings. Bring that movie into Compressor and choose the 16:9 setting of your choice.

Quicktime Conversion is recompressing your footage. Export Quicktime Movie does not require this and it should not be necessary. Not only that, if your render settings are set for 8-bit YUV, I think your QT conversion is rendering only at that quality instead of the full 10 bits. This will cause the banding you are seeing.

Try a short sample with this method of changing your compression settings for the sequence and then exporting QT Movie (not Conversion).

I believe Mr. Nattress has described how to change the HDV sequence settings in previous posts in one of the forums. Example:
https://forums.creativecow.net/readpost/162/864975