Skip to content

Little Frog in High Def

Adventures in Editing
Little Frog In High Def

Archive

Archive for September, 2006

Using a PC laptop in the field for backing up P2 footage:

The first day of shooting happened days after we received the cameras. The DP and producer spent those days setting up the camera and testing the P2 workflow they intended to use in the field. They purchased a PC laptop with a PCMCIA slot and loaded the P2 software that came with it calledthe P2 READER (Windows based only). With this software they were not only able to mount and see the card, but also plays it in it’s raw MXF form before you copy it to your backup hard drive.

When they were in the field, they would have two 4GB cards in the camera, with three others on reserve. The 8GB cards were not available at the time we received the cameras, so they provided five 4GB cards instead. My producer work out a system with the cards and the rubber caps that cover the connectors. In white permanent marker he wrote FULL on one side and EMPTY on the other. When a card was full, it was ejected (when the camera was powered down) and the cap was put on with the FULL side face up. Also the copy protection tab was flipped to protect against accidental erasure. Then the cards were brought to the download station where a tech would slide the P2 card into the PCMCIA slot and copy the footage from it to one of several 80 GB Lacie Porsche drives. They were organized into individual folders to keep things separated, and the folders were labelled A, B, C and so on. The tech would copy over everything on the card…the CONTENTS folder and the LASTCLIP.txt file. That TXT file is very important. You cannot import the footage into FCP without it.

After the contents of the card was copied over, the card was then erased and the rubber cap put back on it with the EMPTY side up. By the end of the two day shoot, There were three full 80GB drives with all the footage on them. The footage was both 23.98 and 59.94…the latter intended for slow motion purposes.

My producer took the information from the smaller drives and copied them onto a 1TB G-Raid. It was the G-Raid that I received to import the footage. The Lacies were put on a shelf…cold storage.

When I imported the footage I went to FILE>IMPORT>Panasonic P2. I was then given a window called IMPORT PANASONIC P2, and the option to name the reel. I did not, so it defaulted to P2 Import. I clicked on the “+” next to the VOLUMES/PATH menu to navigate to the folders containing the CONTENTS folder and LASTCLIP.txt file. When I clicked OK I saw a thumbnail picture of the clip next to the name, and timecode information. The file names had some pattern to them, but still a bit of randomness: 0001E4, 0002RC, 0003GH. The first numbers…0001, 0002…where in order as they appeared on the card. Clip 1, clip two, and so on. Imported I viewed the clips, and added descriptions to the names, like 001FG.mov-CANNON FIRING RIGHT and 0013TU.mov-MEXICAN SOLDIERS MARCH IN FORMATION.

Now, when I began importing the footage, I did a bad thing. I forgot my basics of organization. For some reason, when I imported the P2 footage, I dropped all the clips into one bin. One bin for each day, so I had two bins. Two HUGE bins with over 1000 clips in each one. I don’t know what I was thinking. When I captured the Varicam tapes, I used one bin per tape…known as a Dailies Roll…and then organized into scene bins later. But the P2 footage I threw into one bin for each day. What a nightmare.

What you SHOULD do is drop each card into a separate bin. This way viewing and labeling them is easy and in proper order, as they would appear on a tape. I always create a folder for each tape. This way when I am looking for something and ask the producer for help, if they say “It was on tape 4” or “It was on the last tape we shot,” it give me another option for finding the footage. I do create SCENE bins that I OPTION-drag the clips into. Good organization is key for fast editing, and a successful project. I will stress this as often as I can.

NEXT: The P2 workflow with the P2 Store…and after that importing directly from the camera.

When we first started this project, we always had the HVX-200 in mind. Our original start time was in September, when the cameras were slated to become available. But just as our start date was pushed, so was the release date of the camera. All the while we were researching and testing…and anxously awaiting the HVX.

Then came the LAFCPUG meeting on Nov. 16, 2005. Panasonic reps were slated to explain the P2 workflow…and we were all hoping to get to see a working camera. But no…just the mockup. BUT…there was a huge silver lining. I learned a bit more about the workflow, and my producer befriended one of the Panasonic reps and explained our project. She was every interested in it, and our wanting to use two HVX-200 cameras to support our main camera, the Varicam. He explained how we were shooting in the first week of January, and would really like to use the cameras. The rep was very interested in our project, so they made arrangements for us to test the HVX-200, alongside the Varicam.

We went to a staging area that had a pre-lit stage with their several cameras aimed at it. So we shot with some footage with the Varicam (that was expertly set up in advance by a professional DP…for 20 min) and some footage with the HVX-200 (that was just taken out of a locked cabinet and handed to my producer…no setup whatsoever)…I got to the the lovely model. We were a bit perplexed by the camera settings, but luckily Gary Adcock, a DP from Chicago and beta tester of the camera, was on hand and helped us set it up. Then we shot the tests.

At the time, the drivers for the P2 were only PC, so they had to transfer the footage to our hard drives for us. When I got them to my system, I realized that I couldn’t convert them…as I needed a version of Quicktime that had yet to be released. So we went back to Panasonic and they converted the footage for us. We intercut the Varicam footage with the P2 footage, and noted that the footage was a little soft. But I suspected that was because of our inexperience with the camera. We added a Sharpen Filter at 15%, and did a bit of color correction with the 3-way, and were ready to export.

We took our G-Raids to Lightning Dubs, which recently acquired the Post Group…and their editing facilities. We arranged an edit session in a Final Cut Pro bay equipped with a Kona 2. We needed to output to D5 as it was our required delivery format. The Kona 2 has the wonderful ability to upconvert on the fly…as you output. We needed to deliver a 10-bit uncompressed HD master on D5 at 59.94fps, yet we were editing at DVCPRO HD 720p24 at 23.98fps…and all our footage resided on a G-Raid 500GB firewire 800 drive. The Kona 2 took our footage and upconverted it to 10-bit uncompressed at 59.94 and laid it to tape. No copying the footage into a 10-bit sequence and rendering…no fibrechannel drives. Just the original source footage and a firewire drive. The output looked great, and we sent it off to the network for QC, and waited anxiously.

A few days later we got the call. The engineer asked about our workflow…and we told him. He was quite impressed with the test, and we passed QC. The camera and format by itself didn’t pass the QC, but our entire workflow was taken into account.

So we were greenlit to use the camera. We were at the top of the list to get the cameras, and we received two of them on Jan 9, along with five 4Gb P2 cards (the 8GB ones weren’t available). Literally 3 days later we had our first shoot, and would test our field workflow for the first time.

COMING UP: The HVX-200/PC Laptop workflow.

San Diego Newspaper

While you are waiting my talking about the P2 workflow, here is a little story from SIGN ON SAN DIEGO:

With clouds of smoke and flashes of gunfire, Hollywood invaded Los Peñasquitos Canyon Preserve yesterday during a mock battle from the Mexican-American War.
For more you are going to have to click on the link.

http://www.signonsandiego.com/news/northcounty/20060122-9999-2m22mexwar.html

Sorry guys, but I have been VERY busy logging, capturing and importing footage. Three weekend shoots worth…Varicam AND the new Panasonic HVX-200 camera. We got the first two on the West Coast and immediately put them to work. Now I have LOTS of footage to sort thru. And I had been unable to mention any of this previously due to concerns at the network. But now those concerns have been abated and I am free to talk about using footage from this camera. I will write about it soon…promise.

Oh…and Panasonic will be interviewing my producer about this workflow this week as well, I will keep you posted when it comes out.

-shane

A week of logging, capturing and organizing dailies.

I am glad that my producer said, as he dumped a load of footage at my feet, “Take this week to review this, and let me know if there is anything you feel you are missing. We have another shoot this weekend,” because that is how long it took me to log, capture, view, label and organize the stuff. It was a three camera shoot, so there was a lot of footage. More cannon footage that I can shake a stick at. All three cameras at different angles of a cannon firing so that we can then make it three cannons. And they did a LOT of cannon firing. Good…I’ll need it. There was a lot of cannon fire in the Mexican American War.

My goodness this was a lot of footage. And I foolishly requested that they shoot time of day code…in case I wanted to sync up the shots. When would I ever do that? That made capturing this footage very difficult. Because of the cannon fire, there wasn’t a lot of pre-roll, so logging and capturing was out of the question. I ended up pre-labelling the clip and hitting CAPTURE NOW, and catching it as the timecode break hit. This did allow me to view the footage, and label it as I went, but I like to log and capture and then let it go as I watch it, or answer the occasional e-mail. Never again will I ask for time of day code. Dumb dumb dumb. They are shooting more again this weekend, and asked if I wanted time of day…this time I requested REC RUN code. Never again man.

Beautiful footage. They shot a lot of 60fps stuff for me to slow-mo with the frame rate converter, and a bit of green screening for compositing. I can’t wait to start digging in…but after another weekend of shooting, I feel I have another week of dailies to sort thru.

whew.

Using both to review footage on my Dell 2405 monitor.

The recreation shoot is over…now comes the time to capture all that beautiful footage. I rented an AJ-1200HD deck from a buddy of mine for a day, and was able to borrow a couple different HD SDI converters from another friend (he works at a company that rents FCP systems)…the Decklink HD Link and the AJA HDP…so that I could monitor the output from the 1200HD deck on one of my Dell 2405 LCD displays. I first tried the HD Link.

Decklink HD Link: This required that I download software from the Decklink site, which I did. It also required a USB connection between it and the computer, which I provided. Then I connected it to the deck via HD SDI, then to the monitor thru it’s DVI cable, and into the mixer via RCA for audio monitoring. I got a picture, but it was small…about half size…and had a large black border around it. I consulted the one-page manual and saw that it recommended that I set my monitor resolution to 1220×800. OK…well…how do I do that? I connected it back up to my computer and used the DISPLAY system preferences to make the adjustment. 1280×800 was my only option close to it, so I chose it…then shut down, connected it to the HD Link and…the same thing. Half size. I dig into the monitor settings in the monitor menu and at the bottom it states the resolution to be 1920×1280. Must be the default setting. I dig thru all the menus and find no option to adjust the resolution. So I am stuck with the half size image. Not what I wanted. Plus there was a lot of “noise” on the monitor…vertical waves is the best description. I made sure all the plugs were tight…I even tried the other monitor…same thing.

So I tried the AJA HDP.

AJA HDP: I swapped the boxes out and crawled out from under my desk to see the footage filling the screen. Not quite filling, there was about .5″ of black on the top and bottom, but the full image was present, and filling my screen as much as it could. And the video noise was gone.

OH….MY…GOD!

What an AMAZING image. I saw on my screen live fire from replica cannons. Big booms…smoke filling the screen. Sharp sharp image. I am now seriously considering this adaptor…all I need now is an HD capture card.

UPDATE: I used the HD Link Utility and switched the setting from LCD Monitor to HDTV and it filled the screen. However, the image was the raw squeezed image from the tape…and I had no way of making it look “normal.”

In using the DVCPRO HD Frame Rate Converter, I know have a piece of media that does not originate from tape. I plan on doing this quite a bit, and will end up with a bit of footage that did not originate from tape. This will be problematic if I ever need to revisit a cut at a later date, which has been known to happen. Or if a drive fails and I need to re-capture.

I also need to get this footage to my producer, who is keeping a duplicate set of all the media on his own set of G-Raids, so that when I send him a cut (project file, XML) then all he has to do is re-link. So here’s what I’m gonna do:

I’m gonna use the converter to convert the footage as I laid out in the previous post, and I am going to store them in a folder on the scratch drive called CONVERTED FOOTAGE. I will then back up this footage to either DVD as data files, or to another portable hard drive, and send them to my producer so that he can put the footage on his system. I need to get him other files as well, such as the temp music files I imported from soundtracks and from the composer.

Eventually that footage will reside on a back up drive with all the other graphics, music, stills and any other element we use in the show that doesn’t come on tape.

NOT AN ISSUE

Bearfeats.com reports (http://www.barefeats.com/quick.html#refresh%22%20target=%22NewWindow) that several people have complained of dropped frames when using the Dell 2405 24″ widescreen monitor. Apple reports (http://docs.info.apple.com/article.html?artnum=25117) that monitors with a refresh rate of 60 Hz cause dropped frames, but don’t mention what models. Since the Dell when connected to a Mac only has a refresh rate of 60 Hz…this could be an issue.

I have yet to suffer any problems with this, and the other two editors I know with this monitor also have no problems. So it must be intermitted.

UPDATE 1/18/05

I e-mailed the webmaster at barefeats and he stated that he had only one well documented occurance of this issue, and that the person in question was able to solve the issue by lowering the resolution and switching to 75hz. But I suspect it is the monitor in conjunction with something else, because I have heard from 20+ owners of the Dell 2405 who also edit with FCP and not one of them suffer the same issue…me included, and I have two.

Not to brag or anything.

A gotcha to look out for.

Panasonic wrote a handy little application for their DVCPRO HD codec….the DVCPRO HD Frame Rate Converter. This allows you to take footage shot at 60p (with the intention of it being used as slow motion) and convert it to 23.98 and give you smooth slow motion. It allows you to convert DVCPRO HD into many other frame rates, but this usage is the most common, and very useful. On a show I edited on an Avid, we had to take the footage shot at 60p and have a dub facility run it thru a special converter box and dub it onto another tape. This converter allows me to bypass that workflow.

But there is a “gotcha.” If you attempt to convert the footage in your main project, you can’t. It will give you an error message: “No video file found in selected clip.” I followed the instructions to a “T.” I selected the file in the Browser, went to the TOOLS menu and selected DVCPRO HD Frame Rate Converter…but still got that message. Like many processes in FCP, there is a workaround.

Create a new project with the same settings and import the clips you need to convert. Go thru the steps outlined in the instructions and BOOM, it works. What this does, if you set this up in the conversion settings (which I suggest you do) is create a new media file. Import that clip into the new project, then copy it into the old project and you are all set.

A slow motion clip that is as smooth as silk.

One issue with using any internal SATA, or external firewire drive (400 and 800) is that after a while of non-use, they will spin down. This can cause big problems when you are trying to edit. If you are using media from one drive in one part of your timeline, it will be up and running. But say you have enough footage to require two or more drives…if you don’t access the media from the other drives for a while, they will spin down. So when you are playing your timeline and then hit that piece of media, your timeline will stop dead, and wait for the other drive to spin up…the spinning beachball will appear and then the message that one or more frames were dropped.

This can be a HUGE pain in the ass.

But there is a simple fix. A small application that is really just an Applescript that tells your drives to not spin down while they are connected to the computer and it is on. It is called Spindownfix and can be downloaded for free at versiontracker.com.

Speaking of free downloads, it would be advisable to also get FCP RESCUE. If your preferences ever get corrupt (which is opt to happen with FCP) you use this to trash all the appropriate files. You can also use this to back up your preferences and custom keyboard layout so that when you need to trash your preferences you can then restore the backup set and have all your settings..

FCP Rescue

The Radio Cut…

The holidays are over…my previous jobs are done…so now I FINALLY begin the editing process.

I have been prepping for a while. Looking at the footage, organizing the bins, importing temp music. Speaking of temp music, we found a composer that we all like (two time Emmy winner) and he put almost all him previous music online for me to download. It is good stuff…I am really glad we are going with him. My producer also provided THE ALAMO soundtrack, and I found other sources online from the other composer we were considering.

I also went thru my sound effects library and pulled out some preliminary sound effects I think I will need.

And my producer also provided me with Temp VO…he has a great voice, and a talent for dramatic speech. Saves me having to do it. I’ve done enough temp VO in my time to know that no matter how often I do it…I stink. Well, not all the time, but most of the time.

While the re-creations aren’t scheduled to start shooting until mid-January, I begin the edit process now, with what is called the “radio cut.”

The radio cut is simply laying out the narration, host standups and interview bytes in order so that you can get a rough estimate of the show timing. Music hasn’t been cut in, nor the re-creations, so it will be off…generally shorter than when music and recreations are cut in, as I add space for dramatic effect. Here is an example of what the timeline looks like. As you can see, the only video present is interview bytes, or text placeholders. But soon I will be adding music, adding B-roll and re-creations and still images and begin to “flesh it out.”

It is very interesting working from the home…feels…odd.

Outputting to D5.

Today we performed a test of our workflow. We took various Varicam footage, interiors and host exteriors, and other sources such as one of our graphics machines outputs, and dropped it onto a G-Raid and headed to the post house that had a Final Cut Pro bay and a D5 deck.

The system we output from had a Kona 2 card, and was routed to a machine room. The editor hooked up the G-Raid, opened a new project, and imported the clip into that project. THE MEDIA REMAINED ON THE G-RAID. He then added bars and tone and asked for the network specifications. He took them into the machine room to set up the D5 tape properly. Apparently there are dozens a things you need to take into account when setting up your tape. I am really glad that we had not only the online editor, but machine room operator present to make sure it was done right. After they got everything set up, he did the output with a total running time of 3 min.

It looked gorgeous. We were viewing this on a Sony CRT HD monitor (a discontinued model due to the health hazards involved in making it and disposing of it when it dies) and the image was SHARP. Interiors had a little noise, but it looked like film grain…the exteriors looked flawless.

We output 10-bit uncompressed to D5, upconverted on the fly from DVCPRO HD by the Kona 2, running off a G-Raid firewire 800 drive.

It appears that out workflow works. But we are awaiting QC from the network. When we get that, then we’ll know for sure.

Sweet!

BIG NEWS!

Wow! First trade show I have been to in…in…well, first video trade show I have EVER been to. Lots of stuff happening there.

Big Thing #1 – HVX-200…a working model. Actually, TWO working models. They had this camera, and 5 other Panasonic models including the Varicam, SDX-900 and DVX-100B all pointing at a well lit set…a kitchen. And they had nice Panasonic LCD location monitors on each camera, and I tell you, the HVX-200 looked DARN close to the Varicam. I only noted slight softness in the image, but that was due to the fact that the Varicam had this MONSTER Canon lens. It was very nice looking. While we are considering this camera for our B and C cameras for re-creation shooting, I have no doubt that it will hold it’s own as an A camera. The CCDs are 1/3 the size, which might also account for the slight softness…softness doesn’t describe it well…smoothing of the picture. VERY slight however. And the formats it supports are MANY. DVCAM, at various frame rates, on tape. And on the P2 card, DVCPRO 50 and DVCPRO HD, also with various frame rates…24p to 60i. It even has it’s own version of 24PA called 24PN…this is 24 FULL FRAMES of video, not 23.98…and not 24 recorded at 29.97 or 59.94…but straight up 24fps…for direct film out or export to DVD. Unfortunately, I can easily picture many people choosing this setting and still hoping to output to tape…just like people who shoot 24PA who expect to do the same.

Big Think #2 – The Kona LH. This card is THE card for so many reasons. First, the reason that I need it for. And this, my friends, is huge.
A) It will take my 720p24 DVCPRO HD timeline running at 23.98 and output that as 10-bit uncompressed 59.94…in realtime…through the card. Realtime hardware upconvert. No need for dropping into a 10-bit timeline and rendering…it just does it. I heard this, and had to ask the AJA rep twice…I HAD to confirm it. He confirmed it. This is something that the Kona 2 cannot do! That’s right. What I (and others) assumed was that it could do a hardware upconvert. Actually, I was told by an Apple rep with a Kona 2 that it could. Well, I think we had our wires crossed, for what the Kona 2 can upconvert is SD to HD…it cannot upconvert DVCPRO HD to 10-bit uncompressed HD. That is the power of the Kona LH. This credit card is burning a hole in my wallet.
B) For all those who shoot with HDV, especially the JVC JY-HD100U, you can capture 720p 24 HDV footage with this card into an I-Frame format…as DVCPRO HD…thus making editing MUCH easier. So if you shot 720p24, you can now get that footage into the computer in a very editable format. You will have to capture it at DVCPRO HD 59.94, but you will still have the 24p look that owners of this camera want. Unfortunately you MUST use the JVC HDV deck (BR-HD50U) to do this, for it has component HD outs…and the LH card has component HD ins. Yes, you will be losing quality, but very slight picture quality…nothing noticeable. This workflow was successfully tested by a producer who shot 70 hours of footage for a 13-episode PBS show with this camera. He shot all this footage without first checking for editing compatibility. We was sold an Avid, which didn’t work. So then AJA was approached and offered the LH and the deck…which worked beautifully. Oh, and the motion breakup that occurs with quick movement? That is all smoothed out with this card and codec. The deck retails for $3400. (Lumiere HD and HDVxDV converts the footage into an I frame mode, but doesn’t transfer the timecode. This card does)

So…the Kona LH is the perfect solution for my workflow, and is a very viable solution for 720p HDV. The HXV-200 is looking good as B and C cameras for our recreations…OH…and I got to meet a few Creative Cow leaders in person.

The footage is coming in now. So far only the interviews and b-roll, but a good chunk. My producer played the role of Assistant Editor for this part. He rented a deck and borrowed a high end color correcting LCD monitor mainly to look at the quality of the footage. In doing so, he thought that he’d kill two birds with one stone and capture some of the footage. He was so taken by the sharp, crisp, beautiful image that he ended up capturing all the footage. God bless the tech loving producer. And he knew what he was doing. He used the 720p24 preset and all the footage was brought in at 23.98 fps.
And being such a “hands on” producer with over 20 years experience (maybe more, but I won’t age him any further) he knows the basics of organization.
The script we have is locked. That is, it was written and re-written until the producer and network were happy with it. Typically with this type of script we only capture the interview bytes used in the script. But just in case we might need another sound byte, and because we are renting a deck for the captures, we captured the full interview tapes as well. We have the space, so why not?

Anyway, here is my basic organization:

First thing I did was break up the show into two separate projects, as this is a two hour (90 min in real time) show. I do this as to make the projects into more manageable sizes. The smaller the project, the less memory it takes and the faster things work.

As you can see I have bins for all the current cuts, and bins for all the previous cuts. Whenever I make changes to a cut, from the Radio edit (assembling the narration and interview bytes in order) to a rough cut, I duplicate the sequence and work from the duplicate. I will then take the previous cut and put it in the Previous Cuts bin. So when I am working on a cut, and want to recut a section, I make a duplicate and stow it away…because many many times I have said to myself “hey, the way I cut it two weeks ago works better than the way I currently have it.” Or if we cut something out and now want it back, we go to a previous cut. That has saved time more times than I can count on both hands.

Besides by CUTS bins, I organize things by B-Roll, Interview subject (then by sound byte), Host standups, stills, imported graphics, music, sound effects, whathaveyou. Organizing things like this makes looking for things very easy and very fast…VERY important for an efficient workflow.

This is how I organize my b-roll and interview selects:

When Music and SFX are imported, they will have their own bins and so on.

I also will create a project to store ONLY the final sequence for output…less things to bog down output.

I will be tackling the Radio edit in mid-December and get it all laid out so that we can plan for the re-creation shots. Recreations will be shot in January, and then I will start digging into the cut.

OH…and we are really hoping to get our hands on two HVX-200 DVCPRO HD cameras as soon as they come out, to be our B and C cameras for the battle recreations. We will shoot some test footage, do a quick test edit and upconvert and output to D5 for the engineers at the network to test. I’ll keep you updated as to the results.

Sorry that it has been a while since I posted. I have been working many many hours on my current DVCPRO HD series for National Geographic (this series edited on Avid). Constant script changes keep us editors busy. That and last week I worked another job at the same time for a commercial post facility. Getting my foot in another door, and working with a friend that I met at the Apple FCP discussions. That and I am STILL doing testing and research and playing with footage to see what I can make it look like. Finally, today, I have a night off.

I had a good friend check out my system last week. Besides being a friend, he is a professional colorist and HD expert, working for a well known Final Cut Pro dealership setting up High Definition Final Cut Pro suites. He wanted to see what I built as my offline system.

He seemed impressed. He really liked my monitors, the Dell 2405 (which barefeats has beating the Apple 23″ pretty soundly – http://www.barefeats.com/lcd). So far I have been using Digital Cinema Desktop to view my DVCPRO HD footage, and must say that it looked darn good. I showed it to my friend, and he seemed pretty impressed too. He asked to look at the HD bars, and again, was impressed. The blacks were black, the whites white, and the colors VERY well represented. He said that he’d be confident in the color representation, and would perform color correction with the current setup.

Color correction for LOOK, not deliverable…just to be clear. For final tweaking and checking for broadcast safe colors, he still only go for the high end CRT monitors and external scopes. But hey, I’d only trust those too. But is is nice to know that the color representation is so good, that it is fine for preliminary color correction. No HD card and HD link that converts the HD SDI signal to DVI…just straight DVI via Digital Cinema Desktop.

Oh, and the “?” in the subject line refers to the fact that DVCPRO HD is an offline, and often an online quality codec.