Since mid-December I've been working on a video studio for my office. It started out with some rather modest thoughts about being able to record podcasts, screencasts, etc., from there it has blossomed into a full fledged studio. I mean that in just about every sense of the word from a high definition camera to a chroma key background (green screen), the studio is really pretty decked out at this point. While I have a few issues I still need to contend with (I'll explain later), however, for the most part I feel like I've managed to address the issues that most folks might run into. I wanted to enumerate the components of the studio -- including what I think of the tools thus far. I thought it might save someone else some heart ache when trying to put together a studio.
I have to admit that I didn't think about doing this until I've had a few people walk through and tell me how well thought-out the studio seemed. I realized that even though it seemed like chaos to me that there are a ton of little details that I "just knew" from my interest in video over the years -- things that people were saying that they would have never thought of.
So let's start with backdrops and lighting... To start with you'll need to know what chroma keying is. If you've ever seen the weather in the evening news, you've seen chroma keying in action. The weather man is standing in front of a green screen. Some hardware is dropping the green out and layering in the computer generated map behind him. It's cool and it allows you to be anywhere that you want to be. I picked up one at a local camera store, but this one looks very similar. Just because I may occasionally use the studio for photography I picked up two textured muslins (backdrops). One is charcoal similar to what is seen here. The other is sort of this sandy color.
That leaves us with lighting. We can start with the fact that I use florescent lights in my office and I specifically selected "daylight" bulbs which have a color temperature of 5500 Kelvin. I had to add some floor lighting to help evenly illuminate my backdrops so I put some standard 4 bulb florescent fixtures on stands with pivots along their long axis. I used two four foot fixtures for a bit of flexibility. (My shooting area is about 10 feet wide.) As it turns out, I am only using them with 2 bulbs each in them -- but I can ramp up the power if I need to.)
That's great for throwing light at a background but it will make some hard shadows on a face. To fix that you use a softbox. Basically it's a diffused light source. I settled on an Interfit Photographic Super Cool Lighting kit. The deal with these are they're light, don't produce much heat (because they use compact fluorescent lights), and they do a good job of putting a lot of light on the situation. They output the equivalent of 500 watts of light each -- at the same 5500 Kelvin color temperature as my other lighting. They are double diffused so they produce a nice even light and you can turn on or off lamps in the fixture one at a time. (There are five in each fixture.) I have two of them. They go on either side of the room behind the camera (or at least out of frame.)
With the lighting down, it's time to talk turkey. I ended up deciding on a Sony SDR-HD11. Why? Well, I was trying to balance a ton of things -- including cost. I wanted high definition because at this point it would be silly to not do high definition camera. I wanted to avoid MiniDV tapes because they're an hour long and some of the projects I'm working on may have recordings that are longer than an hour. I really wanted IEEE1394 -- but that meant tape and I ultimately relented for a hard drive based camera.
The other major consideration was cost/performance. For a fixed installation I couldn't see spending at least $1,000 more on a "professional" camera like the Sony HDR-FX7. Sure, the camera I have isn't a 3 chip camera, however, that's really most important in low-light situations and I've already figured out how to throw a ton of light at the situation. The other thing that I knew I was giving up -- which did come back to bite me -- is the audio (VU) meters. I didn't think that it would be that big an issue because I was going to be shooting in a fairly controlled studio environment all the time.
You've got to have something to set a camera on so a tripod was in order. I opted for a Manfrotto 756XB Tripod with a 701HDV fluid head. A fluid head is important when panning a camera to prevent jerking. I've spent a lot of time around video -- and you can tell the difference in a good tripod and head. The setup I selected is designed for lightweight cameras.
With video down it was time to work on audio. This is where the world gets funny. I have served as the technical director for the church I attend for several years. As a result I've learned more about audio than any sane person should have to. When it comes to microphone systems you've got Shure and you've got Sennheiser and a bunch of other people. I ended up with Sennheiser Evolution G2 microphones -- two of them. Yes, if you're doing math, you'll realize that I ended up spending more for audio than video. (and we're not done yet.) Why is that? Well, cameras it's easy to get good quality in the hands of consumers because they see it. It's much harder to get consumers to demand good audio -- most of the time you're at some sort of a child's sports game with wind and yelling anyway. The result is that I went with "professional" gear to be able to get the kind of rich tone I wanted to my voice.
In case you're wondering, I needed two lapel mics because my plan was to be able to do mini-interviews in the studio. Mostly just the conversations that I have with friends that I think should be recorded. I really get a ton of value from some of my conversations and want to be able to share. Unfortunately, I did a trial run with a good friend and totally overdrove the audio in the camera. In other words, the camera was getting so much signal everything was distorted. I thought I had calibrated everything but I made a mistake -- and 40 minutes of really good conversation ended up being wasted.
Upon further review, there is no way to determine how much audio the camcorder is getting -- and frankly Sony was no help figuring out what levels I should be sending to the microphone jack on the camcorder. So I decided to buy a separate audio recorder. I looked at the Zoom H4 Handycorder and the Roland Edirol R-09HR. Both are 24-bit recorders. This means more dynamic range in recording. In the end I decided upon the R-09HR. The H4 is more feature rich, however, in this application what I needed was ease of use -- and visibility. The R-09HR has an LED indication of peaking. Peaking is bad because it means that the audio is being distorted. I wanted to be able to see at a glance if I had a problem.
The audio off of the Sennheiser microphones attached to the R-09HR is beautiful. I haven't had problems aligning the audio and video, I line up the on-camera readings (which is synchronized to the video) with what the R-09HR picks up -- it maybe takes 60 seconds. What's better because everything in the whole setup is digital I don't have to have time code to line things up. We used to have to have something like time code to keep things in sync because tape would stretch and give a little causing sync issues. Pure digital means that once it's in sync it stays in sync.
That's enough gear to record, but I picked up a few other quick things to make assembling the video easier. First, I picked up a Line 6 Tone Port UX2. It's a 24-bit USB attached audio device. It's definitely designed for guitar work, but it's a solid microphone recording device. It offers phantom power if I ever decide to use any condenser microphones. It's also got VU meters so I can see input or output audio levels. I attached to that a Shure SM58. There's probably no better standard microphone than this one. It's the workhorse microphone we use for soloists and performers at the church -- and we're not alone.
Before I get to editing software, I have to share another surprise. I didn't think about the fact that I can't record HD video to a DVD... sure I knew that at some level but on another level I just wasn't thinking about it. So I picked up a LG USB BluRay burner. Why USB? Because I wanted to be able to move it from one computer to another easily. Of course, I hadn't yet purchased a bluray player -- so I ended up buying one so I could even see the discs I was burning.
All that's left to talk about is production and editing. This is an area that I was a bit lucky in that I already had the software. Because it's been an interest for so long -- and because I'm occasionally called on to produce some videos for the church, I already had the software. I have been using Sony Vegas for several years now -- a buddy of mine turned me on to it about 6 years ago and I've not touched another editing software since.
It's not that it's perfect (as I'll explain in a moment) it's that it's relatively stable. What I mean is that as long as I use video clips that aren't corrupted in some way I rarely see it crash. I'm running on Vista 64-bit so that helps, but even when working with it in 32-bit mode it was really stable compared to other tools. I should note that I'm using the "Pro" tool, not the consumer stuff they've put out in the last few years. Some of the compositing stuff that I want to do (think layering in a photo editing software) is only available in the Pro version.
I mentioned that Vegas isn't perfect, well, the biggest issues I've run into with it are: compositing/masking, and importing slides from PowerPoint (which isn't really Vegas' issue but more on that in a minute).
The issue that I have with compositing/masking (besides the fact that that documentation on the topic is woefully inadequate) is that I've been unable to figure out how to put one video underneath the positive part of the mask and a different video underneath the negative part. I've got an idea about how to do this with nested projects (because Vegas allows you to use a Vegas project as an input video file which is really cool). Basically I'm thinking that I'll try to get a Vegas project that outputs one side and then either drop out the rest, mask it, chroma key, or something to get the other side. Anyway, it's overly complicated to try to make that work.
The other issue was getting a clean look on slides imported from PowerPoint. I just finally got this one licked (I think). The issue is that PowerPoint files aren't directly supported as input files in Vegas, and PowerPoint doesn't have an option to create a video file output. (Without third party tools). The output that you can create from PowerPoint includes PNG files (that end up with a resolution of 960x720 and no anti-aliasing on the text) and PDF files.
The PNG files need anti-aliasing to really look good and since you can't turn that on... If we had enough resolution the down sampling in Vegas would get us close to the effect, but 960x720 just isn't enough resolution. I also tried outputting to PDF and then using Adobe Acrobat Professional output them to TIFF files. TIFF because that's what Adobe would output. This got me a higher resolution but not much in the way of anti-aliasing. Ultimately I just couldn't push this technique to the quality level I wanted.
Ultimately the answer turned out to be exporting the slides from PowerPoint as Enhanced Windows Metafiles (EMF) and converting them with TechSmith's SnagIt product. It's an excellent screen capture tool that I used for The SharePoint Shepherd's Guide for End Users. It also has the ability to convert file formats so I converted from EMF to PNG. The difference between the PNG files created by PowerPoint directly and those from SnagIt? well, SnagIt applied anti-aliasing -- and the resolution was MUCH higher (11999x7999). Sure that's more resolution than I needed (by a wide margin), however, it makes a really nice looking result.
To be fair, I have to say two things. First, Vegas has a titler which does alpha channels and anti-aliasing. It's very good, however, I didn't want to regenerate all of my slides. Second, at the suggestion of someone in the Sony Vegas forums I tried TechSmith's Camtasia to record a screen cast of the PowerPoint being presented. The video compression artifacts caused more problems than trying to do it from the still image files. This made me realize that part of the real issue was the fact I was (and am) using the Chroma keyer to generate my mask/alpha channel. It's not really designed to do what I'm doing with it and as a result it's particular about the source. The Camtasia file had the normal video artifacts and that caused the Chroma keyer even more struggles. Ultimately the best answer was to let Snag It do the conversion.
The bottom layer that I have in my work is a moving background. I am trying to add some pizzazz by spicing the backgrounds up a bit. I ended up buying some from Digital Hotcakes/TriLab Productions. They're clean -- although they're a bit fast for the subtle backgrounds I want. I was able to slow them down by putting a velocity envelope on them in Vegas -- so I got to where I wanted to go.
That pretty much handles video, but it doesn't quite handle the audio. Yea, Vegas has some tools for managing audio, however, that's not what it's primarily designed to do. Enter Sony Sound Forge. It's role is to manage audio. I'll be the first to admit that I'm not done learning this tool. However, I do know that I can render the video and open the rendered video up in Sound Forge directly. So I can do some minor cleanup on the audio after I get the video work done -- if I need to. As a point of fact, I've been using Sound Forge longer than Vegas -- it was the original product of Sonic Foundry -- the company that Sony bought.
With all that software, a really good question would be what am I running it on. Here's where I made a cost conscious decision. I realized that I can walk away from my processing if I need to. I ultimately selected a HP Pavillion computer (similar to this one). Why? Well, the cost/performance curve was right. It's a really inexpensive machine for the power. I know, that I could buy about 65% more processing power in the highest end machine -- but at a cost of several thousand dollars. If I need to upgrade to process video -- I can do that later.
I took the money I saved and bought a PAIR of Westinghouse 26" displays. I'm really glad. The displays are brilliant. They make an impact on everyone that sees them. With both of them running at 1920x1200 and setting side by side my biggest issue is having to pick up the mouse to scroll from left to right. Seriously, they're great. They've made editing very nice thus far.
One final note, I've purchased but not received a teleprompter kit and a monitor for it. I did that because even with the monitor (they are big remember) next to the camera it was obvious when I was moving my focus to the computer to read my notes. I didn't like the appearance this created. If things work out well with the teleprompter I'll pass on what I did -- total cost for that part should be less than $300.
So a few disclaimers... I don't have this all worked out yet. I'm still playing with the setup and trying to see what it will -- and won't do. However that being said, thus far I've been impressed. I'm not RECOMMENDING any of this to anyone ... I'm just telling you what I used. Your mileage may vary and you should do your own research. (In short, I don't want to have any whiney emails if you try it and it doesn't work for you.)