[Matt Lyon]

Well I\’ll take a stab at it but I don\’t know if you\’ll like my answer 🙂
I can\’t think of any way to do this in fcp. I would hire a programmer to write a python script to parse an EDL or XML. It shouldn\’t actually be that hard to code, but it would need to have some smarts built in re: timebase. Do you already have someone on board the project who is doing any coding to parse the gps data? Maybe this is a job for them…

Matt Lyon
Editor
Toronto

[Matt Lyon]

Hi Chandler, I’ve never worked with 59.94 timelines, but my gut says that trying to mix 23.976 and 30P footage (or 29.97 for that matter) on it would cause more problems then it solves. If you are concerned about choppiness, I would do a test! Convert a one minute clip with some fast motion and camera panning and see how it holds up. Alternatively, if you have access to a broadcast monitor and a video output card (like a Kona) that will do downconversion to SD video w/ proper 3:2 pulldown, you can get a good preview of what the motion will look like on an SD DVD. (I presume you are going to burn a 23.976 fps DVD, right?)

I guess the alternative approach would be to convert all your A camera footage to 29.97, and use the “conform” tool in Cinema Tools to change all the 5D footage to 29.97 as well. But that seems pretty labour intensive, no? Unless you lock picture first, isolate the XDCAM footage, and convert. Then assemble it all together on a 29.97 timeline.

As for noise reduction, I don’t have much experience with what’s out there. Have you tried searching the boards? I’d definitely start a new thread though if you don’t turn up anything! But the good news is because you are downconverting to SD for your DVD output, you can probably do some aggressive processing and still maintain a decent resolution. Start with a simple “unsharp mask” in the FCP filters and see how that works.

Hope this helps,

Matt Lyon
Editor
Toronto

[Matt Lyon]

I’ve transfered a fair amount of super8 over the years and my preference is to have it done at a proper facility, ideally to digi-beta (or directly to uncompressed quicktime files). The crazy amount of grain and wow and flutter of super8 is pretty hard on lossy compression formats. DV25 to me is right on the border of acceptable; it’s totally workable, but you get a bit of “digital edge” on everything, at least to my eyes.

If you are filming on a prosumer camera, I think your best bet is to hand off the footage in whatever native format the camera shoots. Let the editor worry about converting it to match their other material. The footage isn’t going to get any better then what winds up on the camera tape (or SD card), so my approach is to minimize the amount of conversions the footage undergoes.

Matt Lyon
Editor
Toronto

[Matt Lyon]

Thanks man! Glad my post was useful. Judging by the amount of threads on the subject, these cameras are being adopted incredibly fast. I’m hoping they will improve by leaps and bounds in the next few years.

Matt Lyon
Editor
Toronto

[Matt Lyon]

Hi Bob,

I’ve also never gotten this to work. My somewhat kludgy workaround is this:

select the clip with the keyframes in question. Make sure the video track above is clear and the audio track below is clear.

Option click and drag to make a copy of the clip, and use the shift key to constrain the clip vertically.

Slip the new clip so the audio keyframes are where you want them to be.

With the new clip still selected, hit “command-c” to copy the attributes.

Select the old clip and press “option-v” to paste attributes (or right click and select the paste attribute command).

In the option box, de-select “scale attribute times.” Now select the attributes you want to transfer (“Audio Attributes:Levels” in this case).

Hit “okay.” Now your keyframes should be moved, but the timing of the clip is preserved.

Hope this helps,

Matt Lyon
Editor
Toronto

[Matt Lyon]

Hi Chandler, I just finished cutting a little sizzle reel that had the exact same shooting formats. My workflow was a little different then the one described in your link. It worked for me, but, as always, test for yourself to see what you like best!

We also used 5D footage for inserts/cutaways, so sync sound wasn’t a concern. I received the 30P footage on an HD and converted using compressor. Since the bulk of my material was XDCAM EX 1080p24 (35 MB/s VBR), I wanted to match the 5D footage to that spec.

Initially, I tried converting the frame rate using the “Best (high quality motion compensated)” setting in compressor, but I found this created very objectionable motion artifacts. I ended up using “fastest (nearest frame)” and was much happier with the result. All this does is toss 6 frames out of every 30 to give you 24 fps. Also of importance is that I converted to all the clips initially to 24.00 fps, then did an additional step of opening the files in Cinema Tools and using the “conform” option to change them all to 23.976 fps. Could I have converted directly to 23.976 with the same results in Compressor? Maybe — but I didn’t go so far as to test that. I was happy with this workflow and wanted to get to cutting!

One other point: I converted all the footage to 50 Mb/s CBR … this is a higher data rate then the EX1 footage, but I wanted to minimize generation loss when converting the 5D footage, so I figured the more bits the better.

The resulting footage definitely has some strobe to it. But even the original 30P footage is pretty strobe-y (it looks like it was filmed with a very fast shutter speed — so there is much less motion blur then with footage coming from a film or “proper” video camera). I don’t think my conversion method introduced much more strobing but YMMV. It intercuts reasonably well with the EX1 footage, although I did have to do some color correcting to match.

Hope this helps,

Matt Lyon
Editor
Toronto

[Matt Lyon]

Hi Matt,

You could try downloading the program “metadata hootenany,” which lets you inspect the hidden metadata within quicktime files. (Although a quick glance at an anamorphic file I have here doesn’t turn up any kind of “pixel aspect” flag)

https://www.applesolutions.com/bantha/MH.html

Not sure if it’ll turn up anything useful, but it’s a good tool to have around anyway. I recently had a problem with FCP incorrectly thinking a file was 720×486, even though every other program seemed to think it was 720×480. I edited the “frame” attribute in MH and it fixed the issue.

As for AE, I would chalk it down to quirkiness. I think it’s a good habit, regardless of how well things seem to working, to check that any given program is interpreting your footage correctly (be it Compressor, AE, FCP, etc). Especially before running long render batches!

Matt Lyon
Editor
Toronto

[Matt Lyon]

Josh,

You’re description of your problem is a little confusing (to me at least) which is maybe why you aren’t getting any responses. But I’d check to make sure the sequence settings match for your two timelines. If they mismatch, that could be causing the strobing.

Matt Lyon
Editor
Toronto

[Matt Lyon]

Try making a new, empty project. Copy just your current timeline into it. Then try to “save as..” Might work. Failing that, try the same thing, but copy the contents of your timeline into a new, empty timeline in the new project, then try saving.

Matt Lyon
Editor
Toronto

[Matt Lyon]

Just to add a thought into the mix: have you verified that you\’ve properly set the render scratch directory to your esata drive? If you are rendering to a slower internal drive or the system drive, that could be the culprit. Also, does your system drop frames randomly, or always at the same spot in the timeline? That could point to a corrupt render file if the second case is true.

Matt Lyon
Editor
Toronto