Skip to content

Little Frog in High Def

Musings of an NLE ronin…
description of the photo

Archive

Archive for January, 2010

Are you going to MacWorld in San Francisco? Or are you going to be in the San Francisco area around the date of Feb 5, 2010? Or are you close enough that a drive there wouldn’t kill you?

And do you like hanging out with other Final Cut Pro editors and networking and swapping edit war stories? If so, have I got a place for you to go!

The 9th annual Supermeet in San Francisco is happening on Feb 5, 2010.  And not only are there opportunities to meet new people and network and share workflow ideas, but you can also learn some things from the guest speakers, and also, WIN STUFF!  Yup, these meetings are chock full of raffle prizes.  Not cheesy ones like T-shirts and keychains.  OH, well, OK, those are there too, but they are usually the EXTRA prizes you get for showing your enthusiasm for winnng the big stuff like capture cards, cameras and hard drives.  This year they are giving away a Canon 7D camera…and who doesn’t want that?

Want the details?  Click the link, but here they are, just so you can see at a glance:

Where? - Robertson Auditorium, Mission Bay Conference Center (Rutter Center)- UCSF

1675 Owens Street

San Francisco, CA 94158

Phone: 866.431.UCSF

  • When?Friday, February 5, 2010 (Doors open 3:30pm with SuperMeet Digital Showcase)
  • How Much? $15.00 online
    $20.00 at the door
    Cost includes 2 raffle tickets per person.
    **Tickets on SALE now**

The twenty-first episode of THE EDIT BAY is ready for your listening pleasure. Not everyone has the talent to be the next big director…so go where your talent lies.

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

To subscribe to this podcast in iTunes, CLICK HERE.

OK, if you follow me on twitter (twitter.com/comebackshane) then you might have seen me vent my frustration from time to time about getting stock footage masters without matching timecode.

Here’s what happens.  You are editing a show and need to use some stock footage.  So you contact various news agencies or stock footage houses and they provide you with a DVD with window burn timecode (visible timecode you can see on the picture), or a QT file with the same.  You convert this footage, import it, and if you are SMART, you change the reel name to be something you can track to a log of footage you should have (a vault database or spreadsheet the producer holds onto), and you change the timecode to match what you see on the screen.  So that when you media manage, or decompose, you can simply recapture this from the master tapes.

If you weren’t smart enough to do that, don’t worry, you still have the visible timecode to refer to.  You just use that code to find the footage on the new tape, capture that, and manually replace it.  No problem.

Unless the news agency or stock footage house sends you a tape that has brand new timecode, which is something that happens WAY TOO much lately.  What is SUPPOSED to happen is that you get a tape with the footage, with 5-10 second handles, and the timecode matches each clip.  So the timecode won’t be continuous on the tape.  One clip might be 1:03:44:00, then the next one is 3:56:18:42, then another at 10:34:15:00, then another at…you get the idea.  You can then capture all these separate clips, and then find them easily when you replace the footage.

But no, recently I’m getting tapes that contain 5 to 20 clips, and the timecode on the tape is continuous, and new.  It doesn’t refer to ANY clip timecode.  What does this mean?  This means that you now have to spend a long time searching for the footage and eye-matching it into the cut.  And if they happen to send something like a full dub of the footage (say of a court case in which you use 25 clips, but all within a 2 hour timespan) on two tapes, and the timecode on the tapes DOES NOT match the source that they gave you.  That means that you have to watch the footage to find at least one clip.  Then you can figure out the offset and search for the rest.

Say that the original clip had a timecode of 3:45:13:00, but you found it on the tape at 00:34:14:00.  OK, the next clip on your list is 3:54:20:00.  So about 9 min and 7 seconds from the last one.  So add that to the 34:14:00 and you can look for the clip at 43:21:00 and you might be close.  But still, how hard is it to just send a tape with matching code?  I worked in a dub room, I know how easy this is.  Put a little effort into this guys.  After all, we did just shell out $30/second for this footage.

Oh, and when you are sent footage on digibeta, from something shot like 2 years ago…it’d better look good.  I recently got a tape that was horrible and muddy.  It looked like a dub from a DVD, or something off air from my SD home satellite dish reception.  Turns out I was close.  It was a satellite feed alright.  From London to NY.  HORRIBLE.  How they thought we’d accept that as a master, especially when I need to upconvert it to 1080i HD, is beyond me.  We complained, they made a new dub in London, and overnighted it to us right away.  Something they should have done to begin with.

After working with HDV for just over a year, I am finally working with Panasonic P2 again.  I was finally able to convince my boss to shoot some recreations with the P2 Varicam.  We were able to do this because his normal shooters were busy, and he wanted to see what all the hullabaloo was all about.  Sure, we’d be mixing this footage with the bulk of a program that was shot on HDV, captured as ProRes,  but because of the style of this recreation, it should stand out fine.

We shot in the office, and I told my boss that I’d bring in my Powerbook G4 to offload the footage.  I have this machine ONLY for P2 offload.  It has ShotPut Pro and Proxymill and a nice bus powered drive to offload to.  And it hasn’t seen action in…a long time.

At one point when the DP was setting up, I came wandered out of my dark cave to ask my post super a question.  I saw the DP and a producer setting up the background, and the DP took one look at me and said “Hey Shane!”  This took me aback a bit, because I didn’t recognize the guy.  And I am really good at faces.  But then he continued.  ”Hi, I’m Helmut.  I’m a fan of your blog.”

Aha!  Helmut…that name I know.  Not only has he commented here (hey Helmut!), but I have also seen him contributing regularly to the Cow in the P2 forum.  And we both beta tested P2 Flow from MXF4MAc.  We are both big P2 guys.  Man, what a small world.  And he was the shooter for the day, shooting with his nice HPX-2700 and four, count them FOUR 64GB E-Series P2 cards.  Thank goodness that at the end of the shoot he only filled up one.

OK, moving past the “small world” part of the story…

Now, when the shoot was over, the owner of the company, the producer, and the production manager all piled into my office to look at the wonder that is P2.  They wanted to look at how immediate this tapeless format really was.

Well, it wasn’t quite as immediate as they thought.  Sure, theres no digitizing from a deck, and sure the import process is faster than real time, but I first had to backup the P2 card.  So they all sat there as I put the card into my computer (and interestingly enough the E-Series card showed up without any special drivers beyond the main ones for P2 cards), launched Shotput and began the backup process.

They sat there for 5 min before I said, “you know, this part will take a while.  Yes, I can immediately import the footage into FCP and begin editing, but I like to play things safe by first backing up the card.  So why don’t you all go about your business and I’ll call you when it is done.”  They were a bit put off by the lack of the immediate…immediacy of the process, but they were fine with this.  Because normally this offload process is done either in the field, or back at the hotel after the shoot, so that when you arrive at the production company you can immediately start the import process.

So after 45 min (offloading 52Gb with verification) I called people in and showed them how this worked.  They were impressed that when I would import the footage, after the first clip imported, I could start playing the footage and start logging or editing, all the while the rest are importing.

And yes, I did the typical Log and Transfer to QT because we are on a SAN and the footage needs to be seen by multiple machines.  So the native MXF software would need to be put on all machines if we wanted IMMEDIATE immediate access. Maybe if they finally decide to start shooting with P2.  We’ll see.  One thing I might REALLY suggest is that we either work with pre-loaded scene files, or get P2 Flow so that we can log the footage and add settings before we import.  Use that to export a batch import list that will include a lot of the metadata that the regular L&T ignores.

It’s so nice to get back to a GOOD HD format.  And hey, Helmut Kobler, it was a pleasure putting a face to the name.

Just catchy enough for people to remember.  Enjoy!

When editing a documentary you very often find yourself using stock footage or photos.  And in case you didn’t know it, stock footage and photos cost money to license for use in your projects.  Sometimes they are inexpensive(say $5/second) and sometimes they are exorbitant ($60/sec with a 10 second minimum).  Because you have a limited budget on your documentaries, it is up to the editor to keep track of how much stock footage they are using, and to use it responsibly.  Because it is very easy to get carried away and use a lot of the good (ie EXPENSIVE) stock footage, and wind up with a bill of $60,000…when you have a budget of $12,000.  As editors, we have a few tricks we can employ to keep the prices low.

COLOR CODE THE FOOTAGE.

One way to easily keep track of how much footage you are using, and see roughly how much it might cost, is to color code the footage.  Color all the free stock footage, or the really cheap footage, the color GREEN, for example. Green means GO!  Means that this is the stuff you might want to look at first.  The really expensive stuff, color that RED.  Red is the color of STOP, DANGER!  This is the EXPENSIVE stuff…use it sparingly.  Then you might assign the next expensive stuff ORANGE, the stuff that costs just a tad more than free BLUE.  Purple somewhere in the middle.  If you using FCP your colors are limited, but if you are using AVID, you have many MANY degrees of color to choose from.  This way you can tell, by a quick glance at the timeline, roughly how much stock footage you are using.

SLOW THE FOOTAGE DOWN.

One trick that helps is to take the shots you want to use and slow them down 50%, or more.  This doesn’t always work, it depends on the context of the footage and what is being said.  But it works a lot of the time.  HOWEVER, a few stock footage companies have gotten savvy to this practice, so they have modified their licenses to state total PROGRAM time, meaning how long they appear in the program, opposed to total clip length.

RECYCLE, REUSE

One thing you can do is use a photo or bit of stock footage more than once.  If the stock company doesn’t have that “total program time” license, this will help save a lot of money.  Instead of using 5 pictures of a subject, use 1.  Yes, it can get a bit repetitious, but if you get creative with your cutting, and vary how close you see it each time, it might work out fine.  I once cut some VH1 show about Leann Rimes, and we only had 3 pictures from a certain point of her youth.  But use used each still about 10 times.  But, this being VH1, each time you saw the still, there was some crazy move or series of quick cuts that made it appear different each time you saw it.  This works sometimes, but not all the time.

KEEP THE INTERVIEWEE ON CAMERA LONGER

You want to cover what the person is saying with something to show what they are talking about, but you just can’t afford to.  You need to trim the fat.  Normally we editors use footage and photos to cover what ar called “pull ups,” parts of the interview where we either cut out long pauses, or verbal stumbles, or where we trim up sections of peoples stories to make them more concise.  We cover the edits.  But there are times where you don’t trim, because the person is able to get their point across without is needing to trim.  So in these cases, you might just leave them on screen the entire time.  Yes, it can be boring to just see a talking head, but when you need to cut costs, you do what you have to do.

To tell you the truth I was only able to really pull this off once successfully.  I have done it plenty of times, but only once did I have an interview subject so engaging, so energetic, that I just wanted to see him talk.  I had him on screen for 42 seconds.  Which was a good thing, because everything that he talked about we didn’t have artwork or photos to show.  And the funny thing is, the producer at the network gave me a note asking if we could see more of this guy.

GO ABSTRACT

You might find yourself in a position where your show is cut, but the stock footage used in the cut adds up to $24,000.  And you only have a budget of $10,000.  So you need to cut out a LOT of footage.  Let’s assume that you already tried the slowing down trick, or the companies have that “program time” clause in the contract.  Now what are you to do?  You have a lot of fat to trim.  Well, to do this we can employ a trick that a producer of mine actually preferred to using stock footage…getting “abstract.”

Instead of always seeing what the narration or the interviewee is talking about, show something that alludes to it. For example, on a documentary special I worked on, BLOOD DIAMONDS, there were a lot of personal stories of the atrocities in Sierra Lione, Liberia and surrounding countries steeped in the illegal diamond trade.  We did use a lot of stock footage, and yes, it was expensive.  But when it came to the personal stories, we tried to be more abstract.  A prime example of what I am talking about is in a scene that I have posted on my website as part of my editor’s reel.  That can be found here. (WARNING, the atrocities discussed are of a pretty graphic nature.)

Are you done watching?  OK, let’s continue…

We didn’t have footage of soldiers or people being taken prisoner.  But we did have lots of b-roll.  The b-roll used here is of a rainstorm that hit a village.  There are kids playing in the footage, reaching their hands out to feel the rain.  It’s basically just the DP shooting and trying to keep dry.  But as you can see, it works really well with the story.  The rain sets a gloomy and bleak scene.

And the slow motion and strobe effect and color corrected work in a such a way to make it appear to be flashes of memory. The hands you see reaching out, and the forms of people you see in the doorway are of kids playing and giggling.  But it was edited in such a way to make it fit within the context of the story.  The resulting visuals with the subject’s story set a very powerful scene.

Another example of this type of coverage was done more recently on a show I am currently working on.  It is a crime documentary and a big part of it was the court case.  Initially a lot of stock footage of the court case was used when people described it.  But when we added up the totals, it was triple the budget we had.  So the editor cut out a good 80% of the footage, and replaced it with stock shots of an empty courtroom.  Something that was shot for another show over a year ago.  But the generic footage used in this abstract way worked very well.  And we only used footage of the court case when it was needed to show specific elements and soundbytes from the case.  This is something you see often on DATELINE and 20/20 as well.

This is now something I try to do on a regular basis.  I’m incorporating it into my “style” of editing, where possible.  And by “where possible,” I mean that not every doc calls for this, or you end up with camera people who don’t shoot enough footage to allow you to do this.  So this can be a note to all shooters too.  Shoot a LOT of b-roll.  Get plenty of angles, and plenty of interesting things.  It is good to know going in what the doc is about, and what the possible stories are so that you can keep this in mind when shooting.  I know many talented DPs who do just this.  And they are even kind enough to comment, on camera while shooting (which I hear when I capture), “Hey, editor…this might be really good for that section on blankety blank.”

It is also useful if you the editor have access to a camera (hopefully one that is similar to what the rest of the show was shot on) so that if you wanted, you could go shoot some of this abstract stuff yourself.  I have been able to do that on occasion as well, and it really makes the producers happy.

In my current job I am working on two TV series for Investigation Discovery.  One is a half-hour series that is currently running called CALL 911.  The other is a 10-episode, one-hour series called TRUE CRIME with APHRODITE JONES that will start airing in March.  Both have very different workflows, as CALL 911, due to the fast turnaround needed, is kept fairly simple to finish, while TRUE CRIME is more typical documentary, with LOTS and LOTS of footage, and stock footage, so it is a tad more complex.

CALL 911 is shot on HDV and captured as ProRes.  It only consists of interviews and some b-roll, and recreations.  Not a lot of footage.  So onlining this is a snap and takes only 2.5 days.  Media manage the timeline to the local drive, with handles.  Send to COLOR, color correct, send back to FCP.  Put back in the text (spell check that text), add credits, a slate…the mixed audio…output.  Done.

But TRUE CRIME…that takes a tad longer.

It too is shot on HDV, but this is captured as DV anamorphic to save space.  Because we have to also include not only the interviews and b-roll (of which there is far more than Call 911), but also lots and lots of stock footage and stills.  THIS online takes a tad longer.  More like 5-7 days, depending on how many stills there are.

You see, that is the main bottleneck in the system…those stills.  Because all the moves on those stills, many of which are low resolution, are done at 720×480.  When I online the show, I need to bring everything into a 1920×1080 sequence, and replace the low res stills with high res ones.  So that means recreating each and every still move, manually.  That can take a while.  BUT, my post supervisor, also a graphic artist skilled with AE, stumbled on a plan to use Automatic Duck to send the still sequence (I separate the stills from the rest of the footage into their own sequence, but keep their relative positions) to After Effects, where adjusting the keyframes, mainly the scale, is far simpler.  Replacing the low res with high res and keeping the same relative scale is apparently relatively easy as well.  So we are hoping that this will shave a couple days off the online process.

Because I will be busy with other things, like recapturing all the HDV tapes, capturing the new stock footage masters, going through the sequence shot by shot to make sure that the footage hasn’t slipped sync (this happens occasionally), redoing some of the text…building a credit list, replacing transition effects that go beyond broadcast legal (like VAPOR ACROSS!)… and then building the TEXTLESS ELEMENTS to appear at the end of the show.  And note the start and end times of acts, locations where the network can put in those big lower third promos for other shows (they need 40 second gaps with no on screen text)…Oh yeah, and send the footage to COLOR to color correct and render and send back to FCP.  Making sure all the text is broadcast legal (they can’t be at 100% white, that puts them at 105 IRE). Then dropping in the audio stems (separate tracks of audio) from the audio post house, assigning them to the proper channels.  Making sure that the SPELLING is right in all the text we have (all the text for the 911 calls, in the CALL 911 show…wow…lots of text), because editors tend to be horrid spellers.  Blur license plates…keyframe the movement… blur the addresses in legal or personal correspondence, blur faces of people in pictures that aren’t involved in the show.  Redo those blurs because when we go from SD to HD…locations shift.

So many small details.  But that is the job of the online editor…all of the small technical details.  Making darn sure that every detail is taken care of.  If not, you will fail QC (Quality Check) and the show will be kicked back to you for fixes.  And we don’t want that.

This offline process will be better once the office upgrades to FCP 7 and we can start using ProRes Proxy.  That is a FULL SIZE HD offline codec, with a data rate close to DV.  That would make this process a LOT quicker.  Because I tell you, I much prefer the CALL 911 onlines.

The twentieth episode of THE EDIT BAY is now available for download.  This one is about being a freelance editor.

I post this mainly for the people who don’t download the show through iTunes, so to play in your browser or download direct, click here.  To subscribe to this podcast in iTunes,CLICK HERE.

Oobie, well, Grampoo, (from a real kids show on Disney Channel) shows us how easy it is to make a video. Dang, and I have been slaving under the delusion that this was hard.