Cameras for an experimental set up
Posted 01 June 2007 - 02:37 PM
In case you aren't familiar with my posts in the 16mm thread, I've been experimenting with three strip cinematography, whereby the three primary color records are captured on individual strips of BW neg and recombined to produce vivid color (like Technicolor). Because of the expense and complexity of three strip photography with film, I want to change direction and build a three-camera set up using mini-DV cameras. Below is a crude schematic of my design. Basically, using beamsplitters, mirrors and colored filters, this system will enable each camera to record one color record. Then, in post, I will recombine the images for full color. IN theory, the color saturation, accuracy and overall vibrancy should be markedly above standard cameras, or even three chip systems. Getting to my question: I need three cameras for what I intend to construct. They need to be reasonably small, and not too expensive (so no XL2s, alas). For this first set up, it will likely be some sort of consumer level model (i.e. sub 1000) Obviously, I don't care about how each camera records color, since I will be shooting using the BW function through colored filters, and adding the color in post. However, I do want to get something with as fine and sharp a picture quality as possible for my budget. At this stage, I was planning on SD cameras. Of course, I don't have a ton of money to spend, so I was wondering what make/model of camera you all might recommend that would be the best quality for my purposes, and that would cost between 500 to 700 dollars per camera. I'd really appreciate any help! Thanks!
Posted 01 June 2007 - 02:54 PM
But why with 3CCD DV cameras when they already do this in the camera? Isn't it redundant?
Get a few B&W lipstick CMOS cameras and mount them in an array with filters and point them through your prism scheme. then record out to a deck.
Posted 01 June 2007 - 03:04 PM
But why with 3CCD DV cameras when they already do this in the camera? Isn't it redundant?
Yes, that's my thought too -- recombining b&w into color is cool, but basically this DV set-up is just an elaborate way of getting RGB from a 3-CCD (probably) DV camera without going through the 4:1:1 color subsampling. Using the Andromeda retrofit/hack for the DVX100 may be a better route if you really want 4:4:4 out of a DV camera.
Maybe you should do this same set-up with three 16mm or Super-8 cameras...
Posted 01 June 2007 - 03:42 PM
It'll be great to see another three strip film.
Posted 01 June 2007 - 06:56 PM
Thanks a lot for the feedback. To address your concerns, I am in absolute agreement and believe that film is still the way to go. It is the ultimate dream for me...a practical, three strip process for 16mm or (god willing) 35. But there are a number of issues involved with shooting film that I have encountered with my two strip experiments which have convinced me to go in a new direction, for the time being.
1. Alignment: Proper color depends on a precise alignment of all cameras concerned so that they may each capture the same identical image. This is a very difficult and time consuming process. I tried as much to align them properly by sight, using charts and reference points to ensure close duplication, but there is still a great deal of trial and error that takes time and money in film stock, lab costs and telecine. Furthermore, as another board member pointed out, it might not even be POSSIBLE for perfect alignment of the cameras because of inherent flaws in the individual lenses being used with my bolexes, if they were not from the same "batch."
By using digital, alignment is simpler because I can view my footage immediately, and make adjustments as needed. And, since I can buy three cameras at once, the odds are better of uniformity between the three.
2. Synchronicity: it is vital the the film cameras run at precisely the same speed, and in sync with each other. Otherwise, color bleeding of varying degrees of severity is inevitable as the three records slip in and out of sync. I was able to make do with spring wound bolexes for these first tests, but if I wanted to produce anything with the process, I would have resolve this issue, which can be done by only two ways: 1. common shafts/drive belts that ensure synchronicity or 2. crystal motors linked to a common frame controller so that they are in sync with each other. Either prospect demands much time and money, which, as a grad student, are in very short supply for me.
I want to produce at least one short subject using the process I have developed. Film would be great, and I would prefer it...you are absolutely right on that point, but ultimately, I want to demonstrate that it is possible, practical and preferable to create color through multi camera systems. I have roughly two years left of school, and considering my financial situation, it just would not be economically feasible to build a multi camera system using film within that time, and with my money.
3. Sound. This is a minor point, but I would like to do a short with sound, to "seal the deal" in terms of practicality. My current set up, using bolexes, is far too noisy. As with Technicolor, I would have to blimp it severely to quiet it down enough sufficient for recording levels.
With digital, this issue is moot.
To address you concerns over the quality of using digital, that remains to be seen. I will do some camera tests with a single 3 chip miniDV camera, and do side by side comparisons to see what will happen, and what the quality will be like. I'm not sure what will happen, to be frank, but I believe, based on the theory behind what I am doing, that the color I can produce will be superior to single AND three chip technology. Even three chip cameras ultimately have to record the three color records onto a single miniDV tape..and these chips have their limits in terms of color reproduction. I believe that by devoting a DV tape for each color record, and using three cameras, I can improve upon the color vibrancy, accuracy and saturation in a significant way. And even if the bump in quality is not that marked, it will still demonstrate that my process can work. It could help when the time comes to ask for grants or funding to pursue my greater goals. But at the very least, it is an interesting, new spin on image making. And consider the possibilty three cameras open up for high dynamic range cinematography...I've only begun to read and explore this method, and I am already excited about how this could be applied to what I'm doing.
Ultimately, I want to produce at least one short subject demonstrating this process, which is why I'm looking for some cameras that can record sharp, detailed video. I aim to make a presentation piece, which is why I'm hesitant about using lipstick cams...I'm just not sure the overall image quality will be up to snuff. But, I really do appreciate these comments, and I hope you understand where I'm coming from and what my goals are at this point. I have by no means abandoned film, but merely set it aside out of necessity.
So to reiterate my original question, would you all be able to recommend some nice, reasonably affordable miniDVs. Like something from the Panasonic GS line...something in the range of 500 to 700, I think.
Posted 01 June 2007 - 11:39 PM
I'm not try to discourage you or say that it won't make for acceptable results, but you might want to give Clairmont Camera a call to ask about the "Crazy Horse" rig and how much time and effort it took to pull it off properly. And that's with just two cameras.
Posted 02 June 2007 - 12:12 AM
You'd be better off just coming up with some sort of hack like the Andromeda/Reel-Stream idea of pulling the RGB signal off of the CCD's before it gets subsampled. See:
The 2-strip and 3-strip film camera set-ups produced striking results because it was a radically different way of creating color than conventional color negative and reversal, but in this case, all you are going to do is make a rather mediocre video image in the most awkward way possible.
You could just get a 3-CCD HDV camera and probably create a reasonable 720 x 480 RGB from that and save yourself a lot of bother.
It makes no sense at all to me when a 3-CCD camera already does exactly what you are planning. All you are trying to avoid the color subsampling.
Posted 02 June 2007 - 03:02 PM
I appreciate your candor, which is why I'm going to do some rudimentary camera tests first before going off and buying more equipment. I'll do as I did with the first bolex tests, setting up a miniDV camera on a tripod, and photographing a static scene three times. I will also shoot the scene normally for side by side comparison, and then I will share what I find. I figure the only way to know for sure is to try this out, and see what happens. Most likely, the color I produce won't be worth the added trouble, as you say. But who knows?