Little Frog in High Def

First, I’d like to point out two excellent articles about FCP-X and the future of post. First, a blog post by Oliver Peters, and then a Creative Cow Article by Walter Biscardi.  Both very good and in depth.

OK, now on to me.

I finally downloaded the FCP-X trial and explored the application for a full day. Prior to this, I used it briefly for two hours. But now, while spending all day trying to make something with it, I discovered that I disliked  just about everything about it.  Every minute I spent using it made it worse because it was backwards from the way I like to work.  But I guess that is how it is designed…to be unlike any other NLE, and to do things very differently.  But is the different way better?  Not for me. Am I too tied to TRACKS?  Maybe. To tied to two monitors when working?  Maybe. Dislike that I needed the Skimmer on to view footage in my EVENT, but that meant that the Skimmer would be on in the timeline too, and every time I moved the mouse, I’d be hit with a barrage of hyper fast audio? Definitely.

I had a list of all the issues I had with FCP X, and I was going to gripe about every one, but then my blog went down for five days giving me time to think about things and I’m not going to post another rant.  I am only going to say that I will not be using FCP X in the forseeable future.  Why?  Well, it doesn’t solve any post issues that I currently need solving, and the whole reason why I moved to Final Cut Pro in the first place was that it solved a big post issue I was facing

You see, I started this blog many many years ago, April 2005 to be exact, when I made my leap from Avid Media Composer to Final Cut Pro.  I had been using FCP for a couple years before that (starting with FCP 3) on smaller side projects, like actor demo reels, a handful of short films, a couple of corporate videos.  But I didn’t think it was quite right for me to use on broadcast work…even though FCP 3 was enabled to do this with the RT Mac and Cinewave hardware cards.  It wasn’t until FCP 4.5 came out with it’s native workflow with DVCPRO HD that it caught my attention.

See, I was working at the time on a National Geographic series that was shot with the Varicam to tape at 23.98…720p 23.98.  But the Avid Meridians that we were using couldn’t deal with that format…they were SD only…so we dubbed all the footage to DV tape and offlined that way.  And then when the time came to online, we were faced with a big issue…The Avid Adrenaline that we were onlining with didn’t do 720p…only 1080.  So we needed to upconvert everything, and deal with the 29.97 to 23.98 frame rate difference, and that was complicated, and costly.  We went over budget by just over a hundred grand for 9 episodes.  Not good.

Shortly thereafter I was asked to edit a History Channel series on the Mexican American War…and the producer wanted to shoot with the Varicam.  I was hesitant, given my recent experience.  And while Avid did release an update shortly AFTER our online to allow for 720p onlines…I had just been to a LAFCPUG meeting where I saw FCP 4.5 demo’d showing how it could capture DVCPRO HD from tape natively.  No offline/online…it was online from the start.  And it was 720p…23.98.  Final Cut Pro offered a solution to a post production issue I needed solving.  So I leapt on it.  Then we were going to also try to shoot with the new Panasonic P2 cameras as b-roll…and FCP was the only NLE to actually work with that format as well…so it was a no brainer.  (If you want to see my experiences with that, dig into the archives).

So…with the release of FCP-X and how Apple seems to have changed the way it feels editors should work…it doesn’t offer any solutions to any post workflow needs I have.  In fact, it actually lacks many features I need for the type of work I do. Other than being able to string pictures together to tell a story, and make the audio sound decent and picture look OK…it is missing everything I need to master for broadcast.  You know the list…no OMF for audio mixing, no output to a broadcast monitor for color grading, no ability to export to Color or Resolve for grading, no way to export multi-channel audio that I need (oh, wait, with the update I now have ROLES…), and on and on.

So, instead of trying to make it work…or wait for it to eventually work…I will be looking at the alternatives.  Going back to Avid Media Composer…and exploring Adobe Premiere Pro…both of which are making advances yet retaining the basic editing methodology that editors rely on to edit quickly, and concentrate on the creative and not the technical. They solve the post issues that I am currently faced with.

The thirty-seventh episode of THE EDIT BAY is now available for download.

This one is about how even the editors…and DPs and Production Designers…need to get approved by network and feature film heads in order to be hired on a show.

To play in your browser or download direct, click here.

To subscribe to this podcast in iTunes, CLICK HERE.

I made a quick tutorial in response to someone asking in a forum “Are there any free plugins for Avid for flash frame or blur dissolves? I don’t have Boris…”

That tutorial can be found here.

OK, while I am also transitioning some projects to Avid Media Composer (that I know fairly well, having used it for 10 years before I switched to FCP), I am also transitioning to Adobe Premiere Pro CS5.5 as well.  Because in playing with it, I find it VERY similar to Final Cut Pro in many ways.  And how I can manipulate media on the timeline is very FCP-like, how to composite graphics and layer footage, add titles and a whole list of other thing, is more natural to me. That’s a big bonus because that’s how my brain works.  I have always been more comfortable (and faster) with basic editing in FCP opposed to Avid Media Composer.

Anyway…while there are many many similarities, there are enough differences to make one get a little frustrated with PPro, and swear at it.  Those differences will just take a little getting used to…you always have to learn how the other NLE does things, as it isn’t EXACTLY like FCP does things.  So it will take a little time.

What are these differences?  Well, Walter Biscardi seems to be leading the charge (from FCP to Premiere) and has started a list of things that are slightly different, in his quest to change NLEs.  And he has been kind enough to provide not only pointers about these differences, but has done so in video form.

Here’s part one of his “gotchas” video series.

(If you want to see other videos he has done, such as how to configure PPro to work with an AJA capture card, go here)

OK, I can finally come up for air.  August was a VERY busy month for me.  On top of my regular day job, I took on two side jobs (after hours work).  One was really easy…online a 23 min reality show.  That was straightforward and I was able to do it in two four-hour nights.

But the other one…well, that was a doozie. It’s the one I blogged about last time…the one that required the DNxHD Quicktime with 12 channels of embedded audio.  That wasn’t the only tricky part. The show, that was edited at 23.98, needed to be delivered at 29.97.  This was easy, actually.  Because of MIX AND MATCH (available since Avid MC 4) I could easily convert the timeline and have it look right.  We were given the uprezzed project (they edited low res, DNxHD36 from XDCAM EX, uprezzed to DNxHD175) as 23.98.  Then I would open that project, and remove all the matte graphics…all the lower thirds and other keyable graphics they had in the project.  Because while I can convert the media to 29.97, mattes won’t.

I put the prepped sequence in a new bin.  Then created a 1080i59.94 (29.97) project, and dragged the bin into it.  I then opened the bin, and double clicked on the sequence.  I was prompted with a message stating “This sequence is of a different frame rate than the project.  Would you like to convert it to 29.97?” Why yes, I would!  So it did.  When we tried this with the mattes in the sequence, it said “whoa whoa whoa…I can’t do that.  You have matte graphics in here!”  (I’m paraphrasing)

Now, when I did this, the timecode was way off.  I mean, the original sequence was 48:00:00, but the converted sequence was over an hour long.  In looking at the sequence, stepping through frame by frame, I noted that several timecode numbers were missing.  At first every 5th number, so I was missing 5, 10, 15, 20, 25 and 00.  But then later, I was missing 1, 6, 12, 18, 24.  Really odd.  To correct this I loaded the sequence into the Source monitor, made a new sequence and just cut in the old sequence into the new one.  That fixed things.  We were back to picture ending at 1;48;00;00.

But what caused that? Well, it appears to be because they cut with a 23.98 Drop Frame timeline. Wait…what? 23.98 DROP FRAME timeline?  But 23.98 is a non-drop frame only format…right?  Well, yes.  On tape, and with QT files, 23.98 is non-drop only.  But apparently Avid MC 5 (not sure about earlier versions) allow you to have 23.98 drop frame sequences.  I’m guessing they do this to allow you to cut to a proper drop frame time for delivery.  Clever.  But, it does have that small hiccup of an issue.  Figured that workaround though…

OK, the frame rate conversion was done. And it was delivered in high resolution, all I need to do now is color correct it.  Yes, I could do it in Avid MC, but I don’t that much time, and I am a tad rusty with color correcting in Avid MC, and I really like the control I have with Color…so…I thought I’d go with Color. But, I cannot SEND TO Color like I can with FCP.  So, what did I do?

Simple.  I exported a Quicktime file from Avid MC 5.5…encoded as ProRes 422 (Because Color doesn’t work with DNxHD).  At first I tried exporting as DNxHD and then converting to ProRes with Compressor, but when I did that, I got the famous gamma shift.  But I found if I exported directly to ProRes (something that requires FCP be installed on the system) I didn’t get that gamma shift.  So I exported the QT file, and then I exported an EDL.  What made that easy is that all the video was on one layer.  Well, after I prepped the cut and moved things to one layer.  And then I launched Color and imported the EDL into Color. When you choose the option to use it as a “cut list,” Color then knows that there’s a media file that this references. So it asks for the path to the QT export.  So I selected the EDL, the path to the QT file, chose settings for 1920×1080 29.97, and clicked OK.

And Color imported the media, all chopped up…perfectly.  And yes, where there was a dissolve, Color added dissolves.

I color corrected and then…hmmm, now what?  Rendering.  The options I have are to render as QT…ProRes, AJA 10-bit codec, or Uncompressed, or as an Image Sequence.  I could do 10-bit, but that requires a LOT of space.  And I did still need to do a playback with the client, and have them give changes, and I wanted to do the changes in real time…so I opted to render as ProRes HQ, and playback in FCP.

Yes, this is adding adding a layer or two of compression.  DNxHD175 to ProRes 422 HQ, rendered again as ProRes HQ.  And then exported out as a self contained file (when all the color notes are done), and then using AMA to bring that into Avid MC, transcode to DNxHD220 (the delivery requirement, and because we would be adding titles in MC), again, being able to avoid the gamma shift (the AMA clip and transcoded material matched exactly).  That’s three conversions (DNxHD to ProRes HQ, render to ProRes HQ, transcode to DNxHD)… but that is fine. DNxHD and ProRes are very good compressed formats and hold up well after many conversions.  And, this is not any less than I would be doing if we, say, output to tape, color corrected on a DaVinci, then output to another tape, captured that tape in Avid again for titling.  It might be one more than I’d get with Resolve (as it reads the Avid media, and renders back out Avid media)…but it did hold up VERY well.

Slightly tricky?  Yes…but it worked.  FCP was used in this case only as a means to get the footage from Color to Avid (export self contained QT file)…and as a means of playback.  Well, that’s not true.  I did do the blurring required in FCP, with Andy’s Region Blur.  Because it is far better than the blurring the client was able to do (more subtle).  But other than that, just an in-between option.  So it looks like I can get a bit more mileage out of COLOR while being able to move to Avid Media Composer.  And I was able to convert 23.98 to 29.97 inside the Avid with very good results. Something I couldn’t do inside FCP…and if I used Compressor, would end up taking quite a while rendering.