DEAD BEEF's guide to video production
I spent the previous year picking up knowledge from various forums and web sites learning everything I could about filming and digital video production. If you are an amateur video maker starting at ground zero this page should be helpful to you. In order to be an good producer you are going to have to be part engineer and part artist. These two areas don't usually mix well but if you have an interest in video production hopefully you can pick up any areas you are lacking.
In this document I use the terms camera and camcorder interchangeably. I use both terms refer to a video camera.
- Get an external microphone. The ones on the camera suck,
even on high end prosumer cameras.
Stuff to take with you:
- small tripod
External mics may have not been turned on, cables can be damaged or mis-wired, people can be stupid. Always listen to your audio as you are recording, at the very least check it when you start so you know it was working at one point in the shoot.
Wide Angle Lens:
Lapel mics (Wireless Mics):
Digital input on foreign (Non-USA) digital cameras:
Frame rate - this is the speed of the video, or how many frames per second are being displayed. Standard NTSC is 29.97fps. If you have ever filmed video displays you will know that nasty flicker that comes up on camera. You can stop this if your camera has a variable frame rate and you sync the camera up to match the monitor's refresh rate. However only higher end cameras have variable frame rates.
Upper and lower fields, interlaced, progressive scan, ect.
From the first days of TV video signals needed to be comprised of two images per frame, an upper and a lower field. Below is some very good info I once found on a message board that explains several concepts in one fell swoop.
A video picture is made up of lines. Okay, I know you're thinking "duh... and just what does this have to do with my question?" Just hang in there. On a picture tube, these lines are formed by an electron beam scanning across the inside surface of the tube from left to right. NTSC video (the standard in the US) has a frame rate of 30 (actually 29.97) frames per second. Each frame is a complete picture, but it is composed of two interlaced "fields". Each field is composed of every other line of the complete picture. The fields are displayed at (surprise) two times the frame rate.
The reason why video has interlaced fields dates back to the early days of television. At that time, picture tube technology was such that if the electron beam scanned every line, instead of every other line, the top part of the picture would already be fading away before the electron beam reached the bottom of the screen. To compensate, the interlacing scheme was devised. With interlacing, the beam gets to the bottom of the screen in half the time, and then starts over again at the top, filling in the lines skipped on the previous pass.
The "lower" field consists of lines 2, 4, 6, etc. The "upper" field consists of lines 1, 3, 5, etc. NTSC is said to be lower (or even) field dominant, because the field containing the even lines is displayed first, followed by the field containing the odd lines. The PAL and SECAM video standards (used in most other parts of the world) are said to be upper field dominant, in that the odd lines come first, followed by the even lines.
When rendering digital video, the field dominance designates which field will be inserted into the output stream first. Why does this matter? The fields are ordered in time. Huh? The first field of a frame represents the image as it appeared a fraction of a second (1/60 sec for NTSC) before the second field in the same frame. If you inadvertantly reverse the field order, this result is usually manifested as "jagged edges" and similar sorts of visual distortion.
In general, if you're capturing video from an NTSC camera and producing video that will be displayed on an NTSC monitor, you want to set Premiere to capture (and render) the lower field first. The opposite holds true for PAL and SECAM. If you have an NTSC clip that you want to render for output on a PAL/SECAM monitor, you want to reverse the field order so that the resulting fields are in the correct temporal order. If the field order is switched between capturing and rendering steps (for example, upper field first on capture, lower field first on rendering) the result will be a jagged looking video. If you have clip that looks really jagged when played back on a (CRT) monitor, you might try rendering it with the field order reversed to fix it. Some cameras get it wrong, some capture programs get it wrong, and so forth. If you have jagged-looking on a CRT monitor, reversing the field order is a good thing to try.
This is why when you pause your VCR you can get video that flickers or looks weird rather than being a stationary image: the video is bouncing between the two fields that make up an image.
Interlaced - This is a term for having upper and lower fields. Since the field scan lines fall in between each other the picture is said to be 'interlaced'.
Progressive scan - Fancy word for non-interlaced. The scan lines come one after another, ie: 1, 2, 3, 4 instead of 1, 3, 5 for one field and then 100, 98, 96, ect. for the other field.
LANC - a universal standard that gives you a cabled remote control for your camera. There are numerous places that you can buy these or learn how to make your own. The most common use of the LANC remote is for a remote zoom control. From what I gathered in my video production class zoom controls are judged by how fine the control is. Tiny LANC remotes with membrane keys are worthless, you get the same effect with your IR remote. A big honkin LANC remote with a HUGE zoom rocker arm is what is useful. My advice is don't waste money on a LANC remote unless you know you need finer zoom control or are looking for remote deck control.
Also some older equipment uses LANC for deck control. Supposedly you can use LANC for remotely controlling your camera or deck in programs like Premier, but I've never done this. Firewire (1394) also lets you remotely control a camera so I've never messed with LANC.
3:2 and 2:3 pull down - this is a method of converting the film standard of 24fps to the video standard of 29.97 frames per second and vice versa. Here is a full explanation if you are really curious.
XLR - A standard for microphones, it's a cable with 3 conductors plus a grounding shield. You can get XLR adapters that cost anywhere from $100-$250 usually that attach to the bottom of your camera and provide power to your XLR mic. Beachtek is one company that seems to be well known for XLR adapters. Higher end cameras usually have XLR inputs on them.
Phantom Power - when your camera/audio equipment supplies power for the mic. If you are in the market for a mic, be sure your camera supports phantom power if it's needed. Most consumer grade cameras don't support it natively.
There are a few different classes of camcorders available. For our purposes here I will only discuss digital cameras. Consumer cameras are mostly automatic in function which can be a hindrance. They have a single CCD for video capture. There are two main types of digital cameras, Sony's Digital 8mm and all the other cameras that use MiniDV. If you have an investment in analog or Hi8 tapes, a D8 camera is very attractive. It will read your Hi8 tapes and allow you to dump your video to your pc via firewire just as if they where D8 tapes. Also the media is a tad cheaper than MiniDV from what I have found.
The first big jump are the 3CCD cameras which have a CCD chip for each of the 3 main colors (Red, blue, and green). Once you get into the 3CCD arena you start to find 'prosumer' cameras that have more manual controls and better quality video. Some cameras are just consumer models with better optics and 3 CCD chips. The top of this category however has been making waves in the industry. Smaller independent film makers have used these types of cameras for films that are regularly making it to the big screen. Examples are the Cannon XL1/XL2 and the Sony DSR-VX1000/VX2000/VX2100, ect. For the price ($1000 on the used market to $3000 new) these provide excellent video quality
The next step up are the actual pro grade cameras. For the bottom end of the pro area, these are essentially prosumer cameras that record on HDCAM format tapes. Any camera that records on HDCAM tapes can also record on MiniDV tapes too, but you usually only get about 40 minutes on an hour long tape when doing this. Pro cameras normally come with XLR inputs or an XLR adapter and a higher quality external mounted mic.
"What's the difference between HDCAM and MiniDV tapes?" HDCAM is a more hardy tape and is a bit wider than MiniDV. This means the signals aren't placed so closely together and the chances of signal loss is reduced. But beware, HDCAM tapes cost more and may not give you anything more in return so weigh your options before blindly getting a pro camera.
There's one other thing to note about the differences between the 3 classes of camera's I've mentioned here. The physical size of the CCD chip makes a difference (Think pixels, the bigger the CCD the more pixels it can represent). Consumer cameras use smaller CCD's that are more susceptible to debris and reduced light levels. Prosumer cameras usually use a 1/3" (.333") CCD that captures more light and has a larger focal length. Pro cameras have even larger CCD's from 1/2" and up. The larger the CCD the lower the LUX rating (Basicly the lower the LUX number the better the camera is a picking up light in low light situations).
A piece of personal advice: don't buy a camera based on a frivolous feature like it's ability to take still images. Video cameras take poor quality still images until you get into the $2000 range. If you want stills, get a digital camera, not a camcorder that takes stills too.
The MiniDV format allows for four separate audio tracks however most cameras still only allow two tracks at this time. The only camera that I know of that has this feature is the Canon XL1/XL2 line of cameras. However I understand that you must reduce the audio quality back to 12bit audio to take advantage of this.
I think it's safe to assume most everyone these days does their video editing digitally. The addition of the Firewire standard (1394) has greatly simplified video editing and makes it virtually trouble free compared to the cabling nightmare of previous generations.
Once you reach the need for more advanced editing software you will probably look to Adobe Premier. It's a bit pricey but worth the cost if you are doing a lot of editing work. There are a couple different versions out there, look at what's available.
Additionally some good 1D graphics software will be handy for creating overlays and menu's. Adobe Premier is probably the prime choice, but there are many other pieces of software to do this.
There are other editing packages for sure but these are the most well known.
Other random stuff
Overscan and safe action/safe title areas:
Post - video editing
Rule of thumb: it will take 1 hour of editing for every 1 minute of video you produce. More if you have problems or are still learning your editing software.
Sony Battery Types
"Attenuation cable" to use micro-in as line-in
1 Sig >----R1---+-----> 3 Sig | R2 | 2 Gnd >---------+-----> 4 Gnd
VCD Help - great site with tons of info on hardware & software
General all around good forums:
Poynton's color and Gamma FAQ's as well as other info you may
need to understand at some point:
LANC hacking of Sony cameras:
Reviving and rebuilding Sony lithium ion batteries:
If you are a Windows user and want to break into video
editing, try Windows Movie Maker. It's free with XP, be sure to get the free
upgrade to WMM version 2 from MS's website.
General software and editing site with loads of info:
Store data on your DV cameras tapes: