December Miscellanea
Posted: December 8, 2008 at 10:13 pm, by Isaac
Since posting my last article on video-capable DSLRs and RED’s latest announcements, Jim Jannard has changed the specs of his upcoming product line yet again, mostly to clarify a few things and offer some advantages to existing RED One owners. I’ve also received a few emails asking me what, if the existing cameras aren’t so great, could be done to improve them.
So, I’m working on a list of what I would like to see in an affordable and flexible video camera capable of shooting everything from documentaries to feature films. Stay tuned. In the meantime, I want to ask you, my readers, a favor. I’ve tried to keep this blog on topic and focused entirely on filmmaking, but this is a noble cause, so bear with me while I shift gears for a moment.
The Art of Manliness website is about to name the Manliest Man of 2008, and a friend of mine is in the running. Matthew Chancey is one of the manliest men I know, and I can’t think of a better man to win the title. Ever since I’ve known him, Matt has been quick to stand up for his friends, represent the underdog, and help the weak, no matter what the cost and without ever shirking his duties as a father.
I could go on and on, but the contest ends on the 14th, so go vote for Matt, and while you’re there, poke around on the Art of Manliness site. There are some great articles on things like how to start a fire without matches or land an airplane, and why most of the modern ideas of pop counterculture and pop psychobabble are so stupid.
Canon, RED, and Digital Cinema
Posted: November 26, 2008 at 11:20 pm, by Isaac
A couple of things happened yesterday on the digital camera front. First, the Canon 5D Mark II reportedly hit stores, and second, Jim Jannard announced that more changes to the RED lineup would be announced on December 3rd. I doubt that the two news items are related, but the RED announcement is odd because there was already an official announcement last week that made the direction of the new DSMC line pretty clear.
 Since it’s been awhile since I’ve written anything on this, let me recap the situation. In April, RED announced that they would be adding the larger Epic and smaller, $3,000 2/3” fixed-lens Scarlet to their product set. In late August, Nikon announced that their D90 SLR would have 720p video capability and cost less than a grand. A few weeks later, Canon responded by adding similar functionality to their new 5D MkII, offering full frame 1080p for $2,700. A few days later, RED decided to redesign their camera systems in light of these new market changes. It took almost two months for them to work out this new modular plan.
And since this all unfolded on the internet, much forum furor ensued as various photographers and filmmakers organized themselves into vicious factions to fight over which option was actually “the future.”
As it turns out, the DLSR options from Nikon and Canon are fantastic still cameras, but not so great for video. The D90 has serious jello-cam issues and is limited to 720p at a minimal bitrate, while the 5D has a fixed framerate of 30p and some unpleasant latitude limits in the final footage. Neither camera offers much manual control when in video mode, and audio support is minimal.
 This is totally understandable, because SLRs aren’t really designed with video in mind. The video features seem like more of a marketing afterthought, but it’s an obvious progression; my 30D takes much bigger and much nicer pictures than my XL H1, and if only it could take them faster (say, 24 pictures per second), it would be a vastly superior video camera (for most things). How hard could it be to just up the processing power and bandwidth a little bit so I can shoot faster?
Canon and Nikon’s latest cameras don’t really do this, at least not yet. They only record HD video, not super megapixel frames, and they save highly-compressed h264 files, not raw images. The RED options are superior in every video-related way except two: They cost far more (A Scarlet brain competing with the 5D’s sensor is now $12,000, not including any recording, output, or viewfinder options), and they don’t exist yet.
Besides, it’s a little unfair to compare products that you can buy right now, today, with something that we’ve been sort-of promised in the future, especially since it took RED almost twenty months to ship the first Red One after its announcement. I think the various Scarlet options will be great when they show up, but I imagine that by then Canon and Nikon and others will have made a lot more headway in bringing better video features to DSLRs that already have great sensors, lenses, and raw data options.
The first-generation Scarlets are close to what I want, but the second or third-generation Canon DLSRs may be just as powerful and could beat RED to market. In any case, the competition is good for filmmakers. And as usual, Stu Maschwitz’s blog has been the best place to get info on the DLSR and RED developments.
Outside Hollywood 2.0
Posted: October 15, 2008 at 12:44 am, by Isaac
It’s been a while since I’ve posted, but expect more content soon. In the past few months I’ve been working on a number of different projects, the most of recent of which was an upgrade to this site. I upgrade websites just often enough to have to learn everything from scratch each time, so it wasn’t as quick a process as it could have been. Outside Hollywood looks the same, but there are some new features under the hood.
In addition to a new install of Wordpress, I’ve got a number of social networking and bookmarking buttons up on each post. Just click this post’s headline or the permalink button to see how they work. From there you can email an article or add it to Digg, Technorati, del.icio.us, Facebook, or your Google bookmarks. You can also open a printer-friendly version of any post and print it off with or without comments.
That’s right, I’ve activated public comments, so those of you with questions can post them right here in full view, instead of watching them disappear into the swirling vortex of my email inbox. For now, all comments will be moderated until I have an idea of what kind of responses I’ll be getting, but that may change once I figure out how to deal with the inevitable wave of spam that all blogs seem to be getting these days.
There are a few other changes, like a collapsible archive menu, and I’ll probably be tinkering for the rest of the week, so if things disappear just check back a little later. In the meantime, check out the latest news on the San Antonio Independent Christian Film Festival and Academy (January 5-10). There’s a great line-up of speakers and sessions this year, and more information about specific films and topics should be coming soon.
Review: Deshaker Video Stabilizer
Posted: August 22, 2008 at 6:33 am, by Isaac
As I mentioned earlier, VirtualDub has become a big part of our production pipeline, and it’s largely due to the power of Gunnar Thalin’s excellent plugin Deshaker. Deshaker is the best image stabilization tool I’ve ever used. I’ve gotten slightly better results with the 3D tracker SynthEyes, and Avid’s built-in stabilizer is more workflow-integrated, but there’s nothing that can touch it in terms of the combination of power, speed, automatability… and of course, price.
It also has features that I haven’t seen anywhere else, like the ability to detect and repair the distortions that can come from cameras with a rolling shutter. It can handle interlaced video, which a lot of tools can’t, and it has the ability to reconstruct images to prevent edge flicker, which a lot of tools don’t. It even has scene detection built in so you can deshake footage that’s already been edited together (although this is obviously not ideal). Take a look at a very simple scene below. Even minimal stabilization adds a tremendous amount of production value.
As you can see, it’s best at taking out intermittent high-frequency shakes; the unavoidable fast jerks that you get when trying to handle a tiny consumer camera. If I cranked up the smoothness settings (which are adjustable on every axis, including zoom), it would be even more stable, plus I could isolate the area that I want tracked to just the sky, so the rippling ocean waves aren’t confusing the plugin. This is the only problem with Deshaker - it is so configurable that almost any shot can be properly stabilized, but it’s not automatic enough to get the perfect results every time when batch processing a whole drive full of clips.
However, I have managed to come up with base settings that give decent results for the kind of shooting we’ve been doing with the HV20, and I adjust them for the light conditions of each set. Each shot could be tweaked further for a better result, but for the most part I’m happy with what I get by applying the same presets to an entire directory of clips, and each time I adjust the presets to accommodate a new set, I learn better how to use Deshaker to its full potential. Everything is adjustable, from every aspect of the tracking process to the individual controls for correcting the image and compensating for frame edges, an area where it really shines.
This unstabilized shot is pretty smooth for handheld, but the pan could be a lot more consistent and it would be better to keep the rider in the center of the frame. Unfortunately, stabilizing the jerky pan would reveal the edges of the frame. Deshaker has an adaptive zoom feature which will enlarge the image so that the edges aren’t visible, but that will crop out a lot of detail and soften the image. The solution is an intelligent edge compensation, or image reconstruction. By comparing each frame to the frames that come before and after, Deshaker can actually rebuild the image and extend the frame out to where it should be, like this:
click to enlarge
This is a particularly good example of how great Deshaker’s image reconstruction algorithm is, but bear in mind that this is an accurate result since the spectators’ heads are relatively still and the background is static. Still, it’s not completely perfect. If you look closely you can see some smearing near the bottom of the image, and you’ll also notice that the men on the right are looking at where the horse was, and on the left they’re looking at where the horse will be.
But Deshaker can now pan around inside that image without ever revealing edges, and it shouldn’t have to pan far enough that we’re seeing vastly out-of-date frames. Also, there’s a feature than can extend the color from the edge pixels, which looks like this. The only downside is that image reconstruction is pretty time-intensive - to rebuild this panorama out of about 120 extra frames took almost 10 seconds per final frame, but most shots only need a few extra frames to fill the gaps, so our main edit box generally churns through HD video at about 5-6 fps (after a 9-10 fps analysis pass).
So, give it a try. It’s easy to learn, doesn’t cost anything, and shouldn’t add too much extra time to a production schedule. Everything else that we use VirtualDub for, like the deinterlacing and the denoising, could be done just as well in other programs, but Gunnar’s plugin is one-of-a-kind.
All Hail VirtualDub
Posted: August 20, 2008 at 10:28 am, by Isaac
Generally, our production pipeline involves editing all the footage down in Premiere to the final cut and then, depending on length and complexity, on-lining the project or project sections in After Effects for all color correcting, stabilizing, deinterlacing, frame rate adjustments, or whatever else is required. This the best way to maintain video quality, since you go straight from the source files to the corrected final in one shot, without multiple generations of adjustments and transcoded video files.
However, sometimes there isn’t time for all that, and sometimes video needs to be converted or processed prior to the editing phase (also, with a good codec, transcoding once or twice isn’t too bad). After Effects is great for this too, obviously, since it’s easy to set up batches of files with preset effects, but sometimes this is overkill, and that’s where VirtualDub comes in.
 VirtualDub is a free video utility developed by Lee Avery. It’s not an editor per se, but it can append multiple video files or trim portions out of them. Its main strengths are capturing, processing, and compressing digital video. It can do anything from stripping or adding audio streams to AVI files without recompressing them, or heavy-duty image remastering using third party filters.
Because VirtualDub was originally built to capture and remaster television programming, most of the filters are for removing things like VHS color banding, ghosting from broadcast interference, and network logos, but there are also plenty of great film-related tools for de-flickering, de-jittering, and de-graining damaged telecine transfers. And that’s not all; since VirtualDub is a completely open source project, plenty of developers are contributing new tools all the time.
There are softeners, sharpeners, deblockers, super-resolution resizers, pan-and-scan filters, various color correction tools (limited by the fact that they can’t be animated, but still handy), deinterlacers, subsampling interpolators, vignetting fixers, etc. Perhaps the most powerful is Gunnar Thalin’s Deshaker, which is one of the best video stabilizers I’ve ever used, and merits a blog post all of its own.
So, for our latest project, which didn’t really require a serious on-lining process, we used Virtualdub to process our video before editing. All of our B camera footage (shot with the HV20) was deinterlaced, then denoised, then stabilized, and then had the audio tracks removed in a single generation, resulting in footage that was a much closer match to our A camera (the XL H1).
Once you set up a filter chain in Virtualdub, it’s very easy to apply all those changes to an entire directory of files at a time. More complicated, multi-pass settings may require other tools for the really easy batch processing of clips, but it is still be faster than After Effects for automated processing.
Filmmakers shouldn’t let VirtualDub’s humble origins dissuade them for making it a majpr part of their production pipeline. Whether it will be used for simple codec conversions or cleaning stock footage or reconfiguring entire libraries of video, it has some of the most powerful filters available, the fastest batch interface, and the lowest system requirements of just about any processing tool, and it’s free.
The Canon Hacker’s Development Kit
Posted: July 10, 2008 at 1:53 pm, by Isaac
I own two Canon video cameras, an XL H1, and an HV20. The first retails for about $7,000 these days, and the second can be picked up from just about  anywhere for around $700 (or you can get a slightly newer version, the HV30 for just a little bit more). If you’ve read my reviews, you know that the divergent prices reflect some serious differences in professional capability.
The interesting thing is that under the hood, the cameras are pretty similar. Yes, the XL has 3 CCD chips behind a big interchangable lens, and the HV only has one miniscule CMOS chip behind a little fixed lens — not to mention some serious differences in manual controls and connection ports for things like timecode, genlock, and HD-SDI — but the actual brain of both cameras, the chip which processes the images, runs the viewfinder, mixes the audio, and encodes the HDV, is the exact same Digic DV II chip.
The same is true of Canon’s still cameras. There are a lot of differences between their big black SLRs and their little silver point-and-shoot cameras, but most are run by  identical Digic II or Digic III digital signal processors. It’s cheaper for Canon to develop a single powerful chip that can handle consumer and professional processing than to design and manufacture multiple chips, so we get the same processor whether we buy a $1200 SLR or a $120 pocket cam.
So, why don’t we get access to the same features for a tenth the price? First, those who want a point-and-shoot camera generally want to point and shoot, not mess with a bunch of complicated settings. Also, the tiny form factor of the consumer cameras means there is a limited amount of space for all the controls that would be needed to adjust a bunch of settings quickly. And finally, most manufacturers want to maintain a big difference in features available in the consumer and professional cameras to protect the market for the more expensive products.
But, with a little aftermarket tweaking, you can access more professional features on the point-and-shoot Canons. A bunch of clever hackers have cracked the firmware that the little cameras run, and added new features supported by the processing chip. All you need to do is download the new firmware, called CHDK, and run it from a memory card in the camera. It doesn’t break the existing firmware; all it does is add a ton of new menu options to a new “alt” mode.
This will allow you to control the aperture, ISO, and shutter speed independently, and expand that shutter speed to extremes far beyond the original limits; anywhere from 64 seconds to 1/60,000 of a second. You can save RAW files instead of being limited to JPEGs, add features like battery meters, zebra bars, and histograms to the viewfinder, use the USB port to control the camera, and run multi-shot scripts for things like motion sensing, HDRI exposure, and time-lapse photography.
Time-lapse photography is one of many great reasons for a videographer to have a still camera. The video above was shot using a Canon 30D on my Manfrotto tripod and computer control (although you can also use a remote timer). I think I could have got similar results by duct-taping a cheap camera running CHDK to a tree. Maybe better results, since the computer dropped a few frames.
The downside to CHDK is that you need to make some fairly complicated settings manipulations using only the few navigation buttons that the menu offers, so it can take a lot of time to get the settings you need. And of course, you still have a tiny fixed lens and a cheap imaging sensor to contend with. CHDK doesn’t make your point-and-shoot equal to a SLR, but it does make it much more valuable than a normal consumer camera, and so I wouldn’t recommend buying anything but a Canon.
The next question is, do we have a hacked firmware for the HV20? No, or at least not yet. The HV20’s little brother, the HF10, has been cracked, and may get some extra features soon. There are people working on the HV20 now, however, and if we can get the ability to individually control aperture and shutter speed, or at least be able to turn the gain off (or adjust the image settings in more detail, or do numeric white balance, or export RAW data over HDMI, etc, etc), it will be a much more powerful camera, and more useful for filmmakers. I can’t wait to see what might be possible.
|
|
