[Robert Withers]

This is interesting. It seems you are struggling with a basic editorial structuring issue more than with the technical sound transitions. It sounds like the real problem is how the segments contrast and interact. No sound dissolve will take you from Holocaust to comedy and back. But an idea might. It’s hard to comment without seeing the movie. Good luck.

Robert Withers

Independent/personal/avant-garde cinema, New York City

[Robert Withers]

Thank you, @Rafael, I appreciate your patience and the theory. I still don’t understand why there are alpha indicators and choices in FCP 5.5 menus if there is no alpha in this Apple codec. Why are they there? This version predates ProRes, by the way, so they are not there to deal with ProRes codec. Maybe they are there for graphics codecs.

Also, one thing you said doesn’t match what I am seeing: “changing from NONE to Straight/Black/White in the “Clip Properties” window, that changes nothing.”

What I see is that it _drastically_ changes the appearance/quality of the footage as it appears on a television monitor. If there is no alpha in this Apple codec why is this happening? More important, _what_ is happening? If it makes the footage look better, don’t I want to do it?

Thanks again.

[Robert Withers]

Thanks @Rafael, this is so interesting. This is exactly what I’ve been doing, editing NTSC footage with the NTSC settings in FCP 5.5. But this alpha weirdness nonetheless has appeared.

If I understand what you are saying, FCP took the standard def NTSC signal from the miniDV tape in the Canon miniDV camera and somehow imported it as “Apple DV” codec. And from what I see in the clip properties info, FCP somehow created an alpha channel for this footage in this codec.

I am cutting plain footage and alpha has demanded a role in the process. I didn’t create or ask for or select an alpha channel–it’s just sitting there in the properties looking at me. FCP enables modifying the alpha channel in the properties to change it from None to Straight.

I discovered by accident that when I do this it removes the jaggies from the image that appears on the HD television fed through firewire and a miniDV camera.

It seems like removing the jaggies is a good thing, no?
But maybe not.

I’m inclined to to change the alpha and remove the jaggies for all 750 clips in the timeline and would rather do it as a batch than a clip at a time.

Do you understand what I am saying? I’m not making this stuff up 🙂 I didn’t ask for no alpha channel. But it’s just sitting there in the fat old Apple codec that I’m editing as native standard def NTSC. I just want a clean image without jaggies.

Maybe compressor can change the alpha from none to straight for the whole timeline–I’ll do an experiment. But I don’t know what this means for the footage when I’ll bring it into an HD timeline. Seems like doing this alpha modification from none to straight won’t in itself change the interlace, the frame rate, or the pixel architecture. But it does remove the jaggies.

I have no idea why FCP created an alpha channel for this standard footage or what it’s good for. Since it’s part of the FCP Apple codec I can’t ignore it. What it does do to my program is create a jaggie image if I don’t set it properly. Have you ever run into this?
The manuals don’t say. None of them. I just want to make the picture look good, though I’m not averse to learning something about this Apple codec FCP file in order to do it.
Cheers

[Robert Withers]

Thanks, Rafael. This is miniDV standard def NTSC interlaced footage shot in 2006 and imported into FCP 5.5 from tape (in maybe 2007). I don’t know anything about the codec than that. NTSC standard def is a codec, right? These are not graphics, it’s camera footage. No key has been applied. One or two of the shots have fade-ins from black, but most don’t.

I just discovered by accident that if I change the setting of the alpha channel in FCP for individual shots from none to straight the jaggies disappear on the HD television I’m using as an output check monitor. I don’t know why this works or what FCP is really doing. I’ve read that FCP 5 creates an alpha channel to modify things like transparency. And that this is used for fade-ins or for keying if you do that. I don’t know if that’s true or not. I just try to read manuals, but can’t find much about this.

There is an alpha channel setting that appears when I look at the clip properties. I don’t know if these clips really have an alpha channel or not. I haven’t done any keying or anything to this footage. All I know is that changing this to “straight” from “none” makes the footage look better.

Rafael, thanks again for your interest in thinking about this. I’d rather not even be editing in FCP and this version of FCP but I’ve fine-cut 750 shots in this timeline and can’t export it as XML because they’re based on subclips, it seems. So I want to finish this as well as possible and then export as a Quicktime. I think. Maybe there’s a better way, maybe not. I’m now working in Premiere CC, but can’t get this timeline into that program.

[Robert Withers]

OK, I’ve discovered that all the shots in this timeline have the wrong alpha type (none instead of straight). And that it’s causing the jaggies. So I need to modify all the clips in the timeline (which have been mostly edited from subclips). I can’t seem to do it by selecting all. Is there a way to modify them all instead of each of 750 clips one by one?
I need to finish this in FCP 5.5 because I can’t XML import it into Premiere CC, which I’m working in now, because the subclips don’t export.
Oh I hope someone has a solution.
Cheers,
Robert

[Robert Withers]

Thank you, Rafael.

My Visual Quickpro Guide also tells me that “FCP video clips automatically have alpha channels assigned to them.” So I like your answer that Straight Alpha is OK and that it’s a standard export format for FC. It seems like I can set all the clips in the timeline to Straight Alpha, a little rendering goes on, and they all look good on the HD television.

I still don’t know what’s really happening–it almost likes like the image is being deinterlaced, but I think you’re saying if it looks good it is good.

The text images are a slightly different story. When I first create them, they show a green line in the timeline and look fuzzy on the TV. But when rendered in the background they look clear on the HD television (Even though a little fuzzy on the Canvas.)

Still wondering what straight alpha really means. I hate mysteries when trying to use the technology. Of course mysteries are good in the final program.

[Robert Withers]

Clarification: It’s not on a video monitor that I see the jaggies, it’s on an HD television that I feed the signal to through firewire and a miniDV camera, into an S-video input. (4:3 aspect ratio) My theory is that I’d like to see what the movie will look like in the form that people will see it–played back from a DVD on a television.
No jaggies appear on the computer display in FCP–only on the television.
Does this make any sense?
R

[Robert Withers]

Ontological FAIL softened by Adobe rep chat:

Robert: ok, what’s happening?
Prerna: the sequences are nested ?
Prerna: i see that there are subclips
Robert: When I imported from my camera, I used the option to automatically mark change points between shots. Then I used FCP to automatically convert these marked segments to subclips. I edited with subclips, which behave like shots. There are no nested sequences, as best as I can tell.
Prerna: ok
Prerna: please download cs 6 trial and check in that once
Robert: OK, that may take a little time. According to my readings on the Adobe site, the CS 6 trial is no longer available for download. Shall I start hunting for this?
Robert: Is there a problem with CC7?
Robert: Are you there?
Prerna: there is a known issue with multicam footage and nested sequences
Robert: OK, I’m not using multicam footage or nested sequence. So what do we think my issue is?
Robert: Hello?
Prerna: the change points which are now converted to subclips are not getting detected properly with premiere pro
Robert: OK. And that is how my fcp show was edited. Is it not possible to import this into PP?
Robert: If I convert my timeline into a single QT movie, will that import properly?
Prerna: yes you can try that, mov file will be imported
Robert: Will I be able to access the segments and subclips from this show in PP?
Prerna: yes
Robert: OK, so this could be an interesting workaround. Or I could go back and finish editing this project in FCP.
Prerna: ok, and you can then export timeline to QT and import that in premiere pro
Prerna: you can contact us anytime
Robert: OK, thanks. I want to copy this chat.
Prerna: You’re welcome.
Prerna: Thank you for your patience.
Robert: I will sign off now. Thank you

So . . . my works-in-progress in fcp will not translate because I used the option to automatically create subclips using DV start/stop detection. FCP treats these subclips just like any clips. But can’t export the metadata or the markers so that PP can read them.

Nothing is as simple as it seems. Looks like I’m back to FCP, until I am ready to imported QT files with the final edit versions into PP.

Cheers,
Robert

[Robert Withers]

Third attempt: FAIL Better?
This time followed somebody’s tip on the internet and did the XML export from the BINS in the FCP project. Grouped them in fcp into one for original footage and one for sequences.
This time lots more items exported and came into Media Browser to PP.
Hundreds of clips, different sequences, etc.
More of the main sequence appears in the timeline, no barred out shots, BUT it’s still not the timeline I edited. Still weird repetition of shots overandoverandover.
Will go back and try to export again from the fcp project in all their funky disarray, not changing anything.
I wish I could start editing, waah 🙂 🙁
Am I close enough for someone to offer some help?
If I’m still just talking to myself is this interesting to anybody?
Best wishes and thanks,
Robert

[Robert Withers]

Of course you’re right Michael. I was wrong about use of slug. And that would work, if clumsy and time-consuming.
Selecting tracks forward and advancing them the length of the deleted clip portion is relatively easy. But is there another way than reading the timecode length of the deleted clip portion and re-entering it after selecting the tracks?