My Canon HV20 Short is making the festival rounds – some notes about the workflow
I shot a short film titled A Killer App as the final project for a directing class at UCLA Extension. We used my Canon HV20 exclusively, with the aperture locked to as wide an aperture as possible – a 2.8 f stop in order to get as much light hitting the sensor and to get a shallower depth of field. Footage was brought in via Firewire into Sony Vegas and then converted to Lagarith Lossless using the methods described by Eugenia to export into After Effects for color grading, cinematic cross fades, power windows, visual effects and titling and then exported back into Vegas using Lagarith again.
The final cut was exported as an uncompressed AVI (a huge file even for a ten minute short) which was then used as the new master video file so that various export formats could be created. Among these were the ProRes 4:2:2 using the Avid HDxDN codec 10-bit at 1080p / 23.976 which we dumped onto a drive and handed off to a friend at a post facility so that he could transfer this as an archival master to an HDCAM SR tape. Yes, you will likely tell me that this is overkill as the HV20 records in a 4:2:0 colorspace and is a highly compressed format to begin with, but some festivals (Sundance for example) still require HDCAM tape for exhibition, and so we wanted to have our bases covered.
While working in Vegas we assumed that we would export at 32-bit video levels but worked in 8-bit for smoother playback and due to the fact that we had at least 4 video tracks and multiple audio.
But I have learned a lot since then.
First of all – these cameras only have about 4 stops dynamic range. This means that the actual mise-en-scene is very important. Don’t shoot people wearing black against white walls. Instead work with a narrower palette – try to keep contrasts narrower – and that applies to wardrobe, set design, lighting styles. Also, learn to use a light meter, and use it, and shoot as flat and neutral as possible; although colored gels may seem fun at the time, remember that this isn’t high school theater – that red gel will squish your actors face into the same color space as the wall behind them unless you are very careful, and you will find getting a proper skin tone out of them in post next to impossible. (Not to mention the fact that the camera’s respective color channels respond differently to different colors – you may not seem to be clipping your highlights, but you could be clipping a color channel).
In post I would begin by capturing or immediately converting captured footage into Cineform using NeoScene (which converts to a 4:2:2 colorspace) and cut that directly on the Vegas timeline in 8-bit while piecing things together and again exporting at 32-bit video levels. This is not going to be a comprehensive post about all the settings or best practices, but just some notes about how we did it and what we got. The bottom line is, we tried to stick to first generation material as far as possible, but we ended up going into Lagarith twice, both times exporting at 32-bit, and then once more for the final output. Needless to say the tech guys at the post facility called to tell us that the HDCAM SR version they watched looked super sharp and high end. Unfortunately, I haven’t seen it off the tape. They also said the sound was really good:
While working with the audio tracks (which were recorded using the Rode VideoMic with a 10db pad on plugged straight into the HV20) in Vegas, each character’s dialogue was checkerboarded – cut onto its own track and heads and tails were trimmed right to the edge. We made sure to have recorded and included real room tone from the day, and spent another two days looping any dialogue that was bad due to extraneous noise or for being too far off mic.
On the set of A Killer App, shot with the Canon HV20. Photo by Jessica Marshall-Gardiner
A lot of fun was also had building the environments, background ambience, sound effects -these little details can help a production enormously. The well known golden rule is that sound is paramount. Your video can be just OK, but if the sound is no good, then you are sunk.
We found some excellent soundtrack music by simply asking some appropriate sounding bands on MySpace, and then sending them a music release form which they signed, scanned and emailed back.
Region 0 NTSC DVDs were made and shipped out to a couple dozen festivals and we won an Award of Merit at Accolade, and were accepted in the Glastonbury Horror Festival in the UK. We have yet to hear from about half our submissions who don’t report until later in the year, but so far so good.
Incidentally, if you are not yet aware of Withoutabox.com and you are an indie filmmaker – then go learn it now. It is a one-stop shop for thousands of film festivals and you can submit directly through the site. Note that while use of the site is free, festivals are not – each submission will set you back an average of 25 to 50 dollars US.
A side benefit of using Withoutabox.com, however, is the eventual IMDB page you are given. This is a great way to get your film indexed and noticed by millions of daily visitors to the portal. Also, Amazon owns it so there are opportunities to put your short up for sale through that company.
Regardless of where you are in your development as an independent filmmaker, I hope some of the experience I have shared above will be of use to you.
Please feel free to add any notes in the comment section below about working with your Canon HV20 and Sony Vegas, Lagarith, Cineform, After Effects, post production workflow and deliverables or how you have been marketing and distributing your final product.
A Guide to Getting Footage From Your Canon HV20 Into Vegas and Burning a DVD
I want to assert that this site, the Canon Vixia Experiments, which I started as a love for my Canon HV20 is by no means meant to be a de facto authority on the subject of High Definition video nor of the vast amount of information that one must learn to handle everything from production techniques to capture and digitization to post. It is, in fact just a lab for learning – a place where I collect my thoughts and discoveries, ask questions (and get answers) about optimizing and perfecting the workflow for this series of cameras to maximize the value of my investment in the excellent camera.
I have owned my camera for a year and a half, I was certainly an early adopter. I purchased my HV20 for US$900 and even then considered it a steal at the time. Now the camera is available readily for a third that price.
Canon’s newer versions of the camera over that time have not changed things much and so anything that I write about here still applies to the HV30 and 40, respectively, except for the fact that the newer models can also shoot 30-frame progressive.
Since I have had my camera, I have learned a lot, but I have also found that there are questions that remain unanswered. In dealing with Canon’s way of storing data to tape (for example 24p inside of an interlaced 60i container) I have learned some rather Baroque techniques for converting the footage to true, non-interlaced (that is to say “progressive/full frame”) material, or that there is software (Cineform) and hardware-based capture cards (BlackMagic’s Intensity Pro, Matrox’s MX02 MINI) that can do this conversion on the fly.
I read an article yesterday at a blog by a very prolific and knowledgeable lover of the HV20 named Eugenia (her exhaustive and almost overwhelming blog is a must read for getting into the real nitty gritty) about making a 24p DVD with Vegas.
Some argue that it looks so much better to make a Progressive DVD, others say its overkill and something that you can’t really notice for all the trouble. New 1080p monitors may betray the non-progressive, interlaced NTSC version, or poorer HDTV monitors that don’t handle the de-interlacing as well as they should. But for the most part I want to share with you how to output your materials from Vegas for making a conventional DVD for HDTVs properly when you DON’T want to create a true, progressive 24p DVD. (Read Eugenia’s article “From Vegas Platinum to 24p DVDs” about how to do it, if that is what interests you.)
First things first – assuming you going to be digitizing your footage by capturing from a Canon HV20 camera into Sony’s Vegas non-linear editing (“NLE”) software via a firewire cable, you want to first create a session as follows:
Capturing and Editing Your Footage – Setting Things Up Correctly
If this is your first time capturing from your Canon camera this way, from the File menu, select “New” then from the Template dropdown menu select – HDV 1080-60i (1440×1080, 29.970 fps). This is the native format coming our of the camera so you want the session to reflect that.
If this is not your first time, but you have gotten your settings mixed up in the past, then you can click the “Match Media Settings” icon at the top right of the New Project window – the icon looks like a folder. Once you have clicked on it, it will prompt you to select a file – choose one of your previously captured .mt2 files and it should snap to those same settings.
We’re not done yet – make sure to set “Full-resolution rendering quality” to “Best” and “De-interlace method” to “Interpolate fields.” I also like to set “Pixel format” to “32-bit floating point” as opposed to 8-bit. But make sure you do this all of this before you start editing on the timeline.
Now you can drag your clips onto the timeline. Although it is the default, right click on the footage in the timeline, go to “Properties” (down at the bottom) and make sure it is set to “Smart resample”.
Now you can edit away.
Preparing the Video Stream For DVD
When you are done and you want to prepare your movie for DVD (this is assuming you have Sony’s DVD Architect, but should work with any proper DVD composition software) follow these steps:
Go to the “File” menu->”Render As“->”Save as type” and select “MainConcept MPEG-2 (*.mpg;*.m2v;*.m2t;*.mpa)”
(We are thus outputting in MPEG2 from Vegas for more control over our options).
(If you are dealing with content that is over 90 minutes or you intend to include a lot of bonus material or features on your final DVD, read the ++ note at the end of this article).
From the Render As window, click the “Custom” button. A second window comes up. From the new Template dropdown menu at the top you have four choices depending on what you want/need:
In North America and Japan:
DVD Architect NTSC video stream
DVD Architect NTSC Widescreen video stream
DVD Architect 24p NTSC video stream
DVD Architect 24p NTSC Widescreen video stream
additionally, if you are in a country that uses PAL, you can select
DVD Architect PAL Widescreen video stream
DVD Architect PAL Widescreen video stream
Hit “Okay” at the bottom and now name your file something relevant. Make sure you remember, or copy the exact filename, because you are going to use it for a separate and identically named audio stream.
Hit the “Save” button, and Vegas will render out your video stream.
Preparing the Audio Stream
Now we are going to do all of this again for the audio stream.
Go to the “File” menu->”Render As“->”Save as type” and select “Dolby Digital AC-3 Studio (*.ac3)”
Hit “Save” and Vegas will render out the audio stream – this should go much more quickly.
You can now save your session, close Vegas and open up DVD Architect.
Assembling and Burning Your DVD with DVD Architect
To create a simple DVD for instant playback (no navigational menus) in DVD Architect, from the File menu, select “New” and then the “Single Movie” icon.
A window will open labelled Open Media. Navigate to the video and audio streams you rendered from within Vegas and select the file for the video. Then hit Open.
If you did things correctly, and saved both the video and audio streams with the identical file name, Vegas will automatically associate the audio file with the video and place them in perfect sync on the timeline.
To check it out, hit the “Preview” button at the top and you can watch a preview to make sure everything looks right.
Let’s get this finished!
Click the “Make DVD” button from the top toolbar, and then click the “Prepare” button.
From the Make DVD window, click the browse button and select or create a folder where you want to store your DVD prep files. I recommend using a dedicated media drive for this; as a rule, it’s not ever a good idea to handle or store media files on your boot drive.
Click “Next“.
On the next page you can choose to click the “Optimize” button, particularly if your movie is long. Otherwise, click ”Next“ again.
Click “Finish” on the next page and DVD Architect will prepare your DVD for burning.
Once that is completed, click the “Make DVD” button again, only this time, choose the “Burn” option.
Select “Current Project” and then the ”Next” button twice.
Enter a name in the “Volume” field. This short title will help identify the DVD when placed in a reader.
Your burning device should be identified automtically. If not, select the corret drive. (Make sure you have a blank DVD in there!)
Under the “Speed” option, you can either choose the fastest speed if you are in a rush, or if you want to play it slightly safer, choose a slower burning speed to make sure the laser really gets in there and burns those pits in the dye. Some might argue that this is irrelevant, but in my experience, burning at slower speeds assures the DVD will play on any consumer DVD player – especially these days when you can buy super budget cheapo versions new for less than $2o.
Click “Finish” and watch the magic ensue.
++ If you have more than 1.5 hours of video to put on a single DVD, you must do some math. Find the total time of your movie in minutes, and use the following formula: (34400 / (Time*60)) * 900,000. This result is the average bit rate you should use to fit the video on a single DVD and leave 400 Mbytes free for the audio and menu. Click CUSTOM: VIDEO. and type this number under Variable bit rate: Average (bps). (this tip courtesy Williams College)
If you have any thoughts, comments, critiques or suggestions on how to imporve this guide or apply it to other software platforms, I invite you to post your comments.
Tips on Capturing HV20 Footage to your Vista PC
In windows Vista, the best way to still get the best footage out of your Canon HV20/30 is the method explained at Eugenia’s Rant. Also, I noticed that in Vista at least Premiere Pro CS3 will not properly capture footage from the HV20. Unless yon are using Vegas as outlined below, use HDVsplit. Eugenia explains:
HDVSplit is a popular HDV capture freeware utility. While it generally works OK, unfortunately it occasionally crashes Vegas, because HDVSplit’s captured .m2t scenes erroneously have some black frames at the end of each scene. So, if you will use HDVSplit with anything but Vegas, that’s OK. Otherwise, use Vegas’ own HDV capture facility (which is one big continuous .m2t file for Vegas version 7 or earlier, and with scene detection support, in smaller .m2t files in version 8). Obviously, version 8 is recommended over other Vegas versions for that reason alone. The Pro versions of Vegas 7/8 support scene detection too after you install its official updates from Sony’s site.
Please read her site for excellent articles on the subject of capturing from the HV20/30 into Vegas and other software. Sadly, she will no longer be updating her blog.
I can usually capture quite easily in Vegas although I still think their capture utility leaves much to be desired. I WISH I could capture with Premiere since I like its logging in/out feature for capture – which is the way any decent capture program should work) but no dice with the HV20. At one point I did find a patch for Premiere that was essentially an import template specifically for the Canon HV20/30. A Google search may turn up the most recent location for this small downloadable and very useful file.
This seems to have been corrected and improved in Adobe Premiere CS4 but the verdict is still out.
