Posts Tagged ‘HVX200’

 

Timecode Mismatch – 20. June, 2009

I’ve run into a post-production issue and I need to call upon the hive mind for help.

My film was shot on the HVX200 @ 1080/24pA (23.98) and the audio was recorded into a Sound Devices 702T as Wave Files (WAV). We jam synced both devices and also used a timecode (TC) slate. We then fed the audio out of the 702T and back into the HVX200 via the camera’s XLR inputs.

As the footage was transferred from the P2 cards and ingested into Final Cut Pro (FCP), we ended up with clips that had one video track and four audio tracks (a doubling of our stereo audio feed from the 702T, I believe).

clip-timeline

Now, I was always told that I’d need to take my FCP ingested clips, rip out the four tracks of audio, import the original stereo 702T WAV files, sync audio to picture, lock the new clip and repeat until done. Why do this? So I can use the “better” audio.

For anyone that has ever had to sync hours of footage, you know this sucks big time. But that’s why we jam sync the TC; it’s supposed to make this whole procedure less painful. Plus Sam from the Confidence Bay showed me an awesome way to use QuicKeys to cut tens (if not hundreds) of hours out of this process.

Perfect. I’m ready. I’m excited. I’m dying to sync all this footage so I can hand it over to my new kick ass editor.

Just one problem: the audio and image use two different TC counts.

The WAVs use a 24 (23.98) TC count (check out the TC in the top right window).

audio-fr23-hl

audio-fr24-hl

The HVX200 footage, well, that’s a complicated story. If I recall properly, 1080 24pA DVCproHD footage is recorded to the P2 cards as 29.97. Then, in FCP, you ingest using the advanced pulldown setting and TADA, you have 23.98 clips. Unfortunately, what I found is that the 23.98 footage still uses a 29.97 count. I kid you not. A 23.98 clip counts up to frame 29. The TC doesn’t convert to a 24 count.

image-fr29-hl

image-fr30-hl

And here’s the proof that 1) the clip is 23.98 and 2) that the sequence is set to cut 23.98 footage.

image-item-properties-hl

sequence-settings-2398-hl

Do you see my problem? I have one chunk of media that counts from 0-23 and another that counts from 0-29 and I’m supposed to use their respective TC to sync them together. “00:00:00:27” in the footage is “00:00:01:03” in the audio. Without resolving this discrepancy in counting, I can’t see a way to have FCP automatically sync the audio and image via TC.

How do I easily resolve this so I don’t spend the next month syncing my footage? Is there a way to resolve this discrepancy without 1) a massive re-ingesting of all the footage or 2) paying for some expensive hardware transcoding? Should I even bother with this now? Is the the audio routed from the 702T to the HVX200 via XLR that much worse than the original WAV files? Would it be simpler to just clone a drive for my editor now and deal with this problem after I’ve locked the cut? This violates the “5 minutes now saves you 5 hours (or days) later” rule I learned from my buddy Ken but maybe this is one where I just have to suck it up and sync the WAVs to the edited picture (that might take a month as well).

Thoughts? Questions? Solutions?

Field Dominance & Stills – 14. June, 2009

I just wrapped a gig and had the strangest issue pop up. I’m gonna share in the hopes that someone out there might be able to shed some light.

The gig: I was hired to edit a Spanish language medical video. The project was shot on the HVX200,1080i, 29.97 fps. I was using the latest version of Final Cut Pro (FCP). As the edit progressed, I was also given some animations (HD QT), layered Photoshop images, TIFF drop-ins to replace corrupted frames and 4K JPG stills from a DSLR. The final delivery was a high-quality SD DVD for client approval and then a native (1080 DVCproHD) QT to be sent to the DVD replication house.

Everything worked wonderfully until I started doing some minor color correction with Colorista. At this point I learned that my new favorite plugin (which I like much more than FCP’s 3-Way Color Corrector) is GPU based and as I’m editing on a 2.33 GHz MacBook Pro with 3GB RAM, the plugin can’t handle 4K JPG stills.

Okay, a minor setback but I had a solution. I only had two sequences of stills, one to open the industrial and one to close it.  Solution: why don’t I export each sequence of stills as a native DVCproHD QT. I’ll then re-import these two QT clips into my project, apply Colorista, export and I’m golden.

Not quite. I first noticed a problem after I exported my final edit (the full industrial with the two 4K JPG stills sequences transcoded to DVCproHD) via Compressor for DVD Studio Pro (DVDSP). The “closing” QT stills clip wound up looking like this:

bad-pp

Here I should note that 1) the rest of the industrial looked perfect and 2) both 4K JPG still sequences were exported with the exact same settings:

sequence-settings.png

These settings matched the original DVCproHD sequence settings perfectly, but I only had issues with one of the QT stills clips, not both. Also, the problem only cropped up when exporting from FCP. The entire industrial looked great in FCP.

Time to trouble shoot.  First I thought it must be Colorista. I removed the plugin, exported the corruption-prone section of my final edit to QT but the resulting clip still had the same problem.  I went back to the original 4K JPG stills sequence and tried exporting another DVCproHD QT with the hope that it was a one-off issue.  Nope.  I tried different export settings but still had the exact same stumbling block.

Again, I went back to my original 4K JPG stills. As the problem looked like a field dominance issue, I tried exporting another DVCproHD QT movie with one exception–I set the field dominance to “none”:

sequence-settings-fdn.png

I then imported this closing QT stills clip with the field dominance set to “none” into my final edit project and something weird happened. The QT stills clip suddenly had its field dominance set to “upper” (which is what the HD video was set to):

item-properties-whl.png

Okay, I started to freak out but my inner empiricist convinced me to follow through with my experiment. I applied Colorista to this new QT stills clip and exported the entire final edit to a native DVCproHD QT. Bingo, it worked:

good-pp.png

I then sent the sequence to DVD SP via Compressor and that worked too.

Why? I have no idea. Do you?

Posted in Post-Production

Importing HVX Footage into Final Cut Pro – 28. January, 2009

As I’ve had a few requests, I’ve decided to post instructions for importing footage shot w/ the Panasonic HVX200 into Final Cut Pro.  I created this document for my DIT.

For reference:

Lastly, you should know that I culled this from Creative Cow and Shane Ross, an editor I once worked with.

Best of luck.

Turn Off That iPhone – 2. January, 2009

While editing a PSA for a national women’s health clinic I’ve run into an audio issue, one that manifests itself exactly as on my film. In both cases there’s a noticeable buzz on the recorded production track and I’m suspecting it’s related to these guys:

iphone.jpg

In my film, it takes place during a scene shot in a bathroom. On that set, there were at least 3 iPhones I was aware of. On this PSA, the set was a small office and I counted at least 6 iPhones.

Now I know cell phones can cause speakers to buzz but in all my years of post-production I’ve never had this buzz show up on a digitally recorded track.

Have any of you noticed issues when recording audio to an HVX200? What about to a stand-alone recorder? Have any of you noticed this related to iPhones within close proximity of the recording device?

Regardless, the next time I’m on set I’m asking everyone to put their cells (iPhone or otherwise) in a room far from set and far from the production mixer.

Posted in Post-Production

Back Up At Bat – 3. November, 2008

Once again I’m supposed to finish shooting my film this weekend. Once again I’m keeping my fingers crossed that FilmLA doesn’t boot me form my location again.

Also, I need a gaffer and an HVX200 for the shoot. Anyone want to make a donation?

Lastly, I’m trying to keep the psychological wear and tear to a minimum but I feel like I’m losing that fight. Say a prayer for me, send me the good vibes or sacrifice a virgin in my honor, you know, whatever is your thing.

🙂

Posted in Last Night

Small Format Monsters – 31. January, 2008

If you’ve seen CLOVERFIELD you’ve witnessed the prosumer high-definition revolution co-opted by a huge studio event picture… sort of. Most of that “film” was shot with the incredibly popular Panasonic HVX200 (the same camera used to shoot QUIET CITY). Still, in order to maintain maximum image integrity for all the CGI, the production employed the much heftier (and costlier) Thomson Viper and Sony F-23 high-definition cameras. You can read the Videography magazine article here.

Posted in Cinematography

Quiet City – 29. January, 2008

mum•ble•core (noun): an independent film movement featuring low-budget production, non-professional actors and largely improvised dialog.

(Almost the New Oxford American Dictionary‘s 2007 word of the year; it was narrowly beat out by locavore.)

Apparently this tidal wave of tiny films crashed upon the scene a few years ago at SXSW but I wasn’t aware. Yes, I’d heard of Andrew Bujalski, I’d seen (and loved) THE PUFFY CHAIR and I was envious of Joe Swanberg‘s Nerve web series but I didn’t know they were all part of a movement and I sure wasn’t tracking any of their releases.

That was until I caught this trailer on Miro:

I was impressed with the emotions evoked. These beautiful images wove together into an appetizer that whet my appetite for the feature. Thankfully, I finally caught the film at a Spirit Awards screening (it’s been nominated for the John Cassavetes Award for best feature made for less than $500k).

QUIET CITY is far from perfect but it is a delicate and affecting gem. You should absolutely throw this in your Netflix queue.

After the screening, the cinematographer was on hand to relate some startling facts. 1) The film cost $2,500. 2) It was shot HD on the Panasonic HVX200, probably 1080i 24p; I would have sworn it was shot SD. 3) One scene was captured in two takes… each one 40 minutes long! 3) The entire film was shot over 7 days.

I was absolutely floored by these factoids. In fact, they’ve made me a bit tipsy. Should you bump into me on the street I’ll mumble something about how “I could do that.”

On a final note, director Aaron Katz owes much of the film’s effectiveness to its female lead. She has a warm, soft beauty that shines through her eyes and captures the camera. I’m positive a thousand men have proposed to her since seeing her in this movie.

Posted in Reviews

P2 Field Report – 1. September, 2007

First, a warning to save often. I had written a massive post on this subject but wasn’t saving regularly. When I finally did, my wi-fi was down and I got the “try again” message. I hit “reload” and everything was gone. 🙁

Frustrated and angry, I’ve decided to blog using short bullet points.

Recently I was hired to AC (1st & 2nd) a shoot using the HVX200 and P2 cards. Here’s what I learned:

  • Panasonic recently released a 16 GB P2 card. That gets you 16 minutes at 1080i (this project’s format).
  • You need a driver for your Mac and Final Cut Pro to see the new 16 GB P2 card. Otherwise they won’t show up.
  • The P2 records MXF (Material Exchange Format) files but ingesting the data into FCP will place them in a Quicktime wrapper. Note: you must bring in all the metadata, not just the “video” or “audio” content in the MXF, for FCP to see your footage.
  • “Incomplete” clips aren’t really incomplete. They’re clips that span over two P2 cards. Ingest the data from both cards at the same time and FCP will make them into a contiguous/”spanned” clip.
  • You may find “spanned” clip on a single P2 card. Don’t worry. Because the cards are formatted FAT32, no single file can be over 4 GB. Therefore, continuous shots over 4 minutes long in 1080i are stored as two files. When you ingest this footage into FCP, the clip will appear as one contiguous shot.
  • We were shooting 1080-24pA (that’s 1080 vertical lines of hi-definition resolution at 24 frames per second shooting progressive with an advanced pull-down of 3:2:2:3) but we’re actually recording 1080-60i (60 interlaced fields per second), as if we were laying down to tape. The HVX200 can record 24p-N (24 frames per second shooting progressive in the native frame rate – no duplicated frames and no pull-down required) but only when shooting 720p.
  • The filmmakers rented a P2 Store and it stayed in the box the whole weekend. Why? First, we had a G4 Powerbook (with PCMCIA slot for the P2 cards) and two 320 GB G-Raid drives to store and back up the data. Second, because of a hardware glitch, although the P2 Store is USB 2.0 capable with Macs it can only transfer at USB 1.1 rates. That means a 3:1 ratio and 48 minutes to copy 16 minutes of footage is just plain stupid.
  • The filmmakers should have rented a Firestore FS-100. It’s a 100 GB portable hard drive that connects directly to the camera via firewire. I’ve only heard two complaints about this device. First, when it heats up the fan can be obnoxiously loud. Second, all reports indicate that you can’t record in the “native” format (although the manufacturer says they’ve repaired this with a firmware update). Still, this point would have been moot on our set as we were shooting 1080i and the “native” frame rate is only an option when recording 720p.
  • Although some people suggested we use HD Log or P2 Genie to transfer the data from P2 card to our hard drives, I transfered material using the old fashioned method of “drag and drop”. It worked fine.

In the end, I found this workflow to be incredibly addicting. There’s nothing more satisfying than watching dailies 30 minutes after they were shot. What’s more, someone could have edited the first day’s footage by the middle of the second day. Amazing. After this experience, I’m 93% positive I want to use this camera to shoot my thesis; that remaining 7% of uncertainty is in regards to the camera having a 1/3″ chip instead of a 2/3″ and my desire to use prime lenses.

Pulling Up Stakes – 27. July, 2007

It’s official. I’m moving my thesis shoot from September to early December, probably the weekends of December 8 and December 15. I’ll be using the additional 12 weeks of prep to get all my ducks in a row. Also, hopefully, it’ll let me save some money.

For example, I won’t be enrolling at UCLA for the fall quarter so I won’t be using UCLA equipment or sound stages. I won’t be building a set; I’ll be looking for a “practical” location. That means I won’t have to spend tons of money to build and dress a set. It also looks like I’ll be shooting with the Panasonic HVX200 instead of their VariCam (that’ll save me a ton of money). I’d prefer to shoot with the VariCam and DigiPrimes but the name of the game is creativity from limitations.

Truly, I hope this additional time will let me assemble a rock solid crew, secure a great cast and end up with the best possible short film I hope to make.

Posted in Last Night

Back Against the Wall – 23. July, 2007

Have you ever seen/heard of this:

I first heard it from from a low-budget filmmaker… or maybe it was Joe Roth, I can’t remember. This person said, “pick two because you can’t have all three.” I’ve found that this holds true in the world of short and student films. Yes, there are exceptions (Mean Gene, I see you raising your hand) but it usually works like this. If you want it…

Good & Cheap, it can’t be done yesterday.

Fast & Good, you’re gonna pay through the nose.

Cheap & Fast, kiss quality good-bye.

What does this have to do with my UCLA thesis? Due to a series of misfortunes, my film stands at a crossroads. At the moment, all I have is a script, an actor with a difficult schedule and me. I am approximately six weeks away from my planned start date and I’m facing two options.

Option A – although I only have a script and an actor, I stick with my original plan. I will enroll for the UCLA summer session and use UCLA locations and the UCLA sound stages. I will get my equipment from UCLA. Since enrolling for summer costs way less than enrolling for an official academic quarter, I will use my resources towards building a set, renting a VariCam with DigiPrimes and maybe a good jib or a small crane. My crew will be large, close to 20 I’m guessing.

The advantage of this plan is that I get my film in the can soon, I don’t lose the momentum I’ve been building, I maximize my use of UCLA facilities. With this plan I’m targeting “Good” and “Fast.” One of the biggest problems is that I’m committing to solving all problems with money. I can kiss “Cheap” good-bye.

Option B – I move my shoot to early December. I won’t enroll in UCLA for the fall quarter because the cost is prohibitive (the difference between summer session and a “regular” academic quarter = add one zero to the dollar amount). I shoot with the HVX200 at 720p on P2 cards. I will use practical locations. I will take a page from the book of Dogme 95 and be scrappy (no Fisher Dolly, no Technocrane… no, I was never going to rent a Technocrane). Maybe I borrow my ex-boss’ MicroDolly. I’d work with a small crew.

With this plan I’m giving myself 12 more weeks. I can really crawl around town begging for donations. I might be able to save a few thousand dollars (instead of building and dressing a hotel room set I just rent a room I like, instead of renting a camera I try to borrow one from a colleague). And if I’m in a real location, I probably have to minimize my production size. That means less equipment and fewer people. I’m now working with “Good” and “Cheap”. I commit to solving problems with creativity instead of a credit card. But that momentum I had, see ya later.

This week is the “make or break” week. I think I need to lock at least two of the following three: producer, cinematographer and production designer. If I’m still on my own by the end of the week, I think I need to pull up stakes and move my shoot to December.

Christ, I’m feeling ill.

Posted in Last Night