So, we're into week 3 of my internship @ redWEB.
I spent most of monday writing notes about how we can "package" video for redWeb's clients. It was a tough task, one that comes down to estimating how long it takes to do all the little jobs involved with making a film. For example, research … How long do you schedule for good old fashioned research? Storyboarding … How long might it take to storyboard a 3 minute video?
I think the estimates I made are actually quite accurate. But, of course, there are always things that take longer than expected, so, it will be interesting to compare these estimates with how long it really takes to do all these film-producing jobs.
Tomorrow I start filming on the mini-documenatary about redWEB. This will be part of my final thesis, which, I have decided will be about how to use video for marketing and communication. Their will also be a healthy dose of Social Media thrown into the mix, too.
In preparation for filming tomorrow I spent today testing my set up. I have a small Panasonic HD video camera, a H2 Zoom digital audio recorder, and, a set of lights that I borrowed from Aarhus Film Workshop. In fact, I am very grateful to Aarhus Film Workshop. It has been quite some time since I have been there doing filmy stuff, so, it was very generous of them to agree to loan me a few lights for my degree project.
The lights I have are known as "Red Heads". They're small, not particularly lightweight, but, they are perfect for small video projects like this one.
The lights worked perfectly. The small HD camera isn't "professional" by any stretch of the imagination. But. It shoots 1080p at 30 frames a second, or, 1080i at 60 frames. I chose 1080p at 30 frames. The "p" stands for "progressive," which, also means that the video has more of a fluid look, as opposed to the dryer "video" look of 1080i -- the "i" stands for "interlaced," which, means you're shooting video with two fields, upper and lower, and that's what gives it the dry video look.
I used to shoot a lot of stuff on DV. I remember spending a lot rendering time de-interlacing DV to give it more of a fluid, or, film-like, look. Most HD cameras have the ability to shoot progressively, which looks great and saves all that rendering time later on trying to achieve the film look.
If video becomes a regular thing for me down at redWEB, then, long term it will be a good idea to invest in a Canon EOS 600D, or, something similar. Why? Well ….
http://vimeo.com/24000798With the small camera I have at the moment we can produce good quality video with a little fiddling in post production. But. With a Canon EOS, well, you get the quality of HD video combined with exceptional optics, and, the results look amazing.
The best thing is that the Canon cameras aren't hugely expensive. So. If video continues down at redWEB after my internship, meaning, if I continue making videos for them, then, I'll go out and buy one with my first pay cheque if necessary -- we'll really be able to start producing videos that look way beyond good quality. But. Good quality is where we'll start for now.
So, today's aim was to test the Panasonic HD camera and see what we can do.
I started out with setting up the camera and shooting a shot of myself sitting in a chair, as though being interviewed, no lights, no jiggery-pokery of any kind:
I am sitting in a room with white walls. You can see there is a window behind me. The light in the room is actually not bad. But, with a little jiggery-pokery, it's possible to enhance what the room already gives you. That to me is the secret of lighting a subject well. Be aware of the setting, aware of what the room is already giving you, and, add to it to highlight what you are filming.
The next thing I did was set up the lights. Nothing fancy. I used two lights. A key light at a 45 degree(ish) angle off to my right, and, a fill light off to my left. The fill light was low, about a meter off the floor, and, actually pointing away from me. The fill was pointed at a large white reflector, which diffuses and reflects the light back onto me:
So, with the lights on, I become the subject, rather than just part of the background, which (to me) is what the first photo looks like.
One of the nice effects of placing a large reflector board in front of a subject is that it gives a nice highlight to their eyes. The reflector is actually reflected in their eyes. You see this effect all the time, and, perhaps never really think about it. You can use light directly to give a subject highlights in their eyes. But. For this situation I felt using the reflector board to do it gives a much more subtle and natural looking effect.
I filmed myself for a minute or so. I also had my H2 Zoom audio recorder set up just off to my left. It's a really good quality recorder, and, I know it gives much better results than the onboard microphone on the little HD camera. On top of that, I can place the H2 Zoom much closer to the subject, where as the camera, in this situation, is actually about 4 meters away.
Having filmed myself, it was then time to transfer the results onto the computer and perform some more jiggery-pokery:
The first thing I did was load the footage into Adobe Premiere. I have been a Final Cut Pro guy for quite some time. But. I have tried several times to get to grips with Final Cut Pro X (the latest version) and have always come away utterly frustrated. Apple decided to totally reinvent the software -- changing its interface and how it operates completely -- which left long time FCP users like myself tearing their hair out because, basically, it was a whole new program … To stick with it, you'd have to learn it all over again, and, considering that programs like Adobe Premiere still work like non-linear editors have done for years, it's actually easier to switch.
I opened up Premiere and literally started editing right away. With FCPX I spent the first few hours trying to figure out where things were. And, even when I started editing, Apple (in their infinite-loopy wisdom) decided to change the workflow as well, so, choosing clips and getting them onto the timeline -- which actually isn't a timeline at all in the new FCP -- was just plain awkward.
So, I have switched to Premiere for the time being. I am also interested in checking out Avid Media Composer. But. Premiere works, so, Premiere it is for now.
Having loaded in my test footage the first thing todo was take the audio I'd record in the H2 Zoom and sync it up. This took me about five minutes. It's just a case of dropping the audio onto the timeline and moving it until the sound from the camera and the sound you have added are aligned. You can really hear it when it isn't aligned, even if it is off by a single frame, you can hear a slight echo. Once the external sound was synced, I switched off the sound from the camera footage, and, suddenly I had much fuller and clearer sound.
I should say at this point that I did spend some time in Bias Peak Studio cleaning up the sound from the H2 before dropping it in Premiere.
Now it was time to perform the extra jiggery-pokery. I wanted to add an effect that works really well for further highlighting a subject. I wanted to blur the background slightly. This involves layering two lots of video on top of each other. The bottom layer is your footage with a slight gaussian blur. The top layer is your footage with a garbage or shape matte applied to it. A garbage or shape matte cuts out your subject and removes most of the background. But. Because you have the background, etc, in your first layer of video, the effect makes your subject sharp, with the background slightly blurred. Feather the edges of the garbage or shape matte and the two layers blend together seamlessly.
It's an effect that works really well, and, I have to say that I like it so much that I will have to make sure I don't over use it from now on.
The next step was to do a little colour grading. I adjusted the contrast and brightness a little, and, the saturation. I had manually set the white balance on the camera when I filmed to adjust it to the lighting conditions, so, no real colour correction was needed.
The final result, I feel, is actually a much richer looking picture. The subject is highlighted in a very subtle and natural looking way. All in all, good results from a camera that, as I said earlier, is a long way from the more professional cameras available.
The test, however, had two down sides. The first, the room was very echoey. While I managed to tweak the sound from the H2 zoom, I couldn't get rid of what the echoey room did to my voice completely -- it sounded a little hollow. To combat this tomorrow -- I am interviewing my boss Christian in the same room -- I went out to a hardware store and bought a bunch of thick moving blankets. I will hang them in suitable places in the room which should help deaden the echo.
The second down side, which, isn't really a downside … rendering! Having performed all the jiggery-pokery I wanted to do on my test footage I went to render the results. The compute came up with an estimate of three hours! So, I left the rendering until it was time for me to go home. I then set it to render and will see the full results of today's test first thing tomorrow morning.
I did take the still images I posted here, so, from those I have a pretty good idea of what the final results will look like.
I can't wait to film the interview tomorrow. I have been talking about filming and doing videos down at redWEB since I arrived, and. tomorrow I will finally be able to start doing more than just talk.
