Fetus is correct about the cameras. Anything will work. Most people trying to do this buy a decent camera, and then degrade the footage in post, allowing for better low light performance etc. Most of your work in something like this is sound design, and you can get some impact if you have a good script and a good sound designer. Unfortunately, it doesn't really work even then.
Here is the part no one seems to understand about the Blair Witch project and it's success.
1. They spent over 25 million dollars on the film. $35,000 on the actual film, and $25,000,000 on viral marketing, a one time thing where a clever trick was used to get results similar to a $75,000,000 marketing campaign for a normal film. Adjusted for inflation, that would be 42 million dollars in today's money.
2. They were the first found footage film with real marketing, and actually tricked the audience into believing that it was real footage, something you can't accomplish now
3. They did this at a time when the web was much different (199
, and it was far easier to trick people with a small scale hoax
4. Basically everyone in the indie film world tried to repeat this formula, except no one bothered to do any research, and there have been about 3000 failed found footage films per year for 20 years now. It's the most misunderstood story in the history of film. To this day I get indie filmmakers coming up to me and saying "you can make a movie for 20 grand and make 250 million dollars" It has literally never happened even once in history.
Cannibal Holocaust was perhaps the grandfather of all these films, and made small splash in it's day, it's actually a much better horror film than Blair Witch, but did not have the financial support, and earned much of it's success as an underground cult film after being banned in many countries on suspicion of being a snuff film.
Paranormal activity, and it's much better cousin REC, were some of the only successful clones released after Blair Witch, both with heavy marketing funded by large companies. They broadcast the shooting cost, which is very low, and downplay the marketing cost, since this makes for an interesting and popular story, even though it is completely false. To my knowledge no person without the backing of a large corporation has successfully marketed a found footage movie since February 1980. If you go to film marketing expose events, you will find, to this very day, literally thousands of unsold Blair witch clones per year, a market driven directly by the misunderstanding of what actually happened.
Here is an article detailing the actual story. When the studios approached the filmmakers, it was sort of a one of a kind film. Today there are literally tens of thousands of found footage films out there, and few have been purchased by a major studio since paranormal activity. On the rare occasion a found footage film is released these days, studios generally produce then in house, to reduce unwanted legal complications, such as paying out even 1% of profits to an actual creative artist. Even in the case of the Blair Witch project, the single most successful indie film in history, the director of the film was paid a reported 1.1 million dollars out of a net 250 million, coming in at around 0.4 % to the creator, and 99.6% going to wealthy people who never created anything.
We take a look back at the way the web was used to make The Blair Witch Project a hit, and the era of viral film marketing it began...
www.denofgeek.com
If you want to make a found footage movie, go ahead and have fun doing it. Just don't expect great financial results. I just publish this as a warning, to help correct the record, and so people don't get their hopes up trying to replicate a story (about no budget films that made it big) that never actually happened.
As successful as Blair Witch was financially, it was even more successful at creating an urban legend, one about cheap indie movies raking in huge profits. It basically doesn't happen.
Anyone is free to correct me if you can evidence that any such thing happens on any type of repeatable (non fluke) basis.