3 channel 36 bit 1280 X 720 low $ camera - Viper?

[Wayne Morellini]

Sorry Laurence and Rob, I have been recovering from flu bouts, and just realised I have been getting you two mixed up lately, my appologies.

Laurence I agree about you girlfreind's father, he would be a great help.

Guys, good on the website idea. There are a number of camera links in the homemade thread that could be put in there.

[Laurence Maher]

Great idea for the wiki. Count me in.

Hey Obin, you said you'll be "getting the HD camera". You're still referring to your own or did you say you were going to also purchase one from silicon imaging. Oh, which reminds me . . .

Didn't get too far on that website, but are the silicon imaging cameras too expensive? Do they not use 24p? Seems like the subject kind of came and went.

By the way, I should put out there that I need to get a new editing system up and running some time soon, whether I like it or not. My Digisuite LE has crashed one too many times and I have stuff I need to get done. I was thinking on getting a mac based NLE, but all this various talk about Linux and a bunch of stuff I've never heard of like the free cine . . . whatever it was is kind of pulling me back. I don't want to get something uncompatable with whatever camera system we go with. Truthfully, as far as editing is concerned, I need something both user friendly and stable. I've never touched a linux in my life. Before I buy, maybe you guys should confirm . . . it's okay to get a mac, right? I mean, it'll be compatable here? (Looking at G5 with igniter X HD card, if I go crazy, maybe Cinewave or Kona 2 when it comes out, although this is probably overkill . . . I just want something that kicks ass and still works!!!!! WAAAAAAAAAAAHHHH!!)

Maybe you guys can give me system advice and/or more cost effective advice, but whatever I get, remember I'm no real tech head. It has to keep working on it's own, without the weekly "fix-it" session. Too many hours go down the drain.

Thanks!

P.S.

I'll talk to my girlfriend's dad then (electrical engineer) and keep you posted.

[Steve Ipp]

Hi guys!!!
Sumix isn't dead. Here's the update. Send your requests for the MI-MV40 pdf to my mail, I'll be happy to bounce it back as long as you don't come by hundreds :))

Vladimir will be back at the office next week and will contact you.
Please take a look at the attached pdf file for MI-MV40 sensor. We are
considering to use this sensor for two reasons. 1. The format is larger
than the current sensor we use. Wide angle 35 mm objective can be used
with this sensor by losing about 40% of field of view. 2. The 200 frame
rate will extend the application of this camera tremendously. The two
major problems are 1. It is expensive, $1500 small quantities and $1000
for quantities of 1000 or more. 2. We will have tremendous bit rate,
200*4megapixel per second, the interface and storage will be nightmare.

About adapter for 35mm objective to be used with with 2/3" sensors we
had several discussions with our optic engineers. The problem is that
about 80% of light will be lost if we use ground glass. If we do this
without ground glass, then we must make one adapter for each particular
35mm objective. So we do not know yet what to do about this.

Please let me know what you think

[Valeriu Campan]

Laurence,
Regarding your NLE dilema: I am using both platforms but have settled for serious and reliable work on Macs. I don't want to start a platform war, but I cannot recommend it highly enough.

Good luck,
Valeriu

[Wayne Morellini]

Laurence

I read that April Linux format (you should see if you can buy it). The problem is that you have to do command line stuff, and it isn't the most freindlist/complete feature list. The only one really that addresses HD is Cinelerra (that I know of). Now that sort of stuff is fine with some people but most it will not be simle enough.

When we find out what Sumix is doing, and decide what we are putting on it, we don't really know what to buy. Some of the guys here though will be able to address the compatabilities issues with you. I advocate we really need to use plugin codec for the capture software and plugins for the most popular editors. This will solve most of these problems and allow us to freely use Mac, Win, or Linux at will. So I personally think there might allready be softare out there to do this on the Mac, but I don't know what it is.

Steve, what is your email, I really would like to read the pdf?

For our Sumix people: GG glass is not the best, but recently the Static adapter thread has been talking of better alternatives: Chemical ground glass, Minolta Microolense based ground glass camera focusing screens, the Reflexite Intenscreen Beattie camera focussing screen (I think an array of microlenses), holographic projection screen (rejected, but I don't know if all avenues where researched). The focusing screens are available cheaply in a plain format. The microlense screens will add upto 1 1/2 stops (more for medium format) of light over a GG screen. They have been condensing the projected light to also increase brightness. The problem with many screens is visible grain, which is worse in HD but the company that does the Beattie screen will do a custom made order, so I imagine the array canbe made smaller. The Reflexite group probably has the broadest optical experience (including LCD backlighting) and should be considered. I have not read all the Static, 35mm and HD10 adaptor threads, but they are a good source for information on increasing light intensity.

http://www.intenscreen.com/

That sensor (but haven't read the PDF) will probably be great in the sub $3K camera, though if it could get within 90% of the performance of the big boys for $800 cheaper that would be even better. About the data rate etc, we could simply use lower resolutions until we can afford the extra storage / newer technology etc. To solve the 35mm adaptor problem, we could have a set of lense convertors for each of the popular or desired formats to connect them to the adaptor.

Nice Wiki.

[Steve Ipp]

Wayne, look in my profile, you'll find the e-mail there.
I'll also send you a document detailing exact PC configuration for a particular NLE to edit HD @ speeds 120 MB +.
Regards

[Wayne Morellini]

Whoops, thanks Steve. I use so many forums I forgot about those little email buttons on some of the posts.

[Laurence Maher]

Well Wayne, IMHO, $800 to be knocked down 10% image quality is a bad trade, I mean yes, none of us want to pay any more than we have to, but fudging $1000 or so I think is something we need to be able to stomach. Maybe not 5k, but 1k? There's an old saying that filmmaking is 90% work and 10% talent. Don't kill the last 10% that might get your film on the big screen, I say. When it comes down to it, one advantage we'll have with these systems is that we will be indistinguishable from big boys image wise. I don't know as much technically as you guys, but I've spent a lot of time trying to get my feature sold. The only time I got looks was when they saw my relatively decent image and production value vs. someone else's shady job. Beyond that, it's all star power. So I say fight for that 10%, man. It may be all we have, and unless you have a name, it will also be your film's first impression.

Again, just IMHO

[Laurence Maher]

By the way,

So ya, I've been looking at the Kona and stuff. Wayne, I think your advice is solid if I was . . . smarter (lolol) . . . but I'm simply not able to do this code writing stuff. (No background in it at all, I can barely get into dos management) As long as you guys think it will be mac compatable, that's what I'll go for.

Any warnings against it?

Thanks!

[Laurence Maher]

By the way,

Got a lead on another possible camera. But so far they say it would be 1280 X 1024p. 24p and a little faster. I guess we could just crop the image for a 16 X 9 720p right?

[Wayne Morellini]

Thats why we need these plugins, for people like yourself to keep familar simple environments, I wasn't trying to suggest that you write them, but it would be good if the manufacturer, or somebody out there (as an open source project), incorporates the ability in the capture software to use plugins, and maybe write some plugins (if special ones needed).

Most people will not notice the 10% drop in quality, if we were at a movies we would have to look for it. Our perception would probably tell us that it is more like 1-5% drop, as most of these performance issues concern production lighting and only certain scenes will show it, and then we can control the lighting a bit more to get rid of it, and make it nearly identical to the big boys). One of the generalisations in life is that the last 20% costs the last 80%, so if we just go up and nudge the borders we can get great results without a, say $4K, chip. I really want to put it to the HDV market, by offering the next best thing to the $50K ENG class at the same price ;) For me that performance will look credible on screen, though maybe not George Lucas credible.

[Laurence Maher]

Okay, I see what you're saying about the plugins. Ya, good show.

About camera:

Perhaps I'm just selfish, but if I can compete (with origin of signal, not effects of course) with George Lucas, for 5 or even 6 k, I'll think I got the right camera. I think most indie filmmakers would agree with me . . . well, I think. Surely they'd rather pay the extra $800 to not have to work that much harder to get the image they want if they're doing say Noir or Horror or extravegantly shot sci fi. On the set, and in post, all of the money you saved will be spent again in hours, sweat, and worst of all, frustration. And that's just with the first feature. After you make your second, third, that $800 you didn't spend at first is now several thousand in payback. Of course, I'm a big perfectionist and I'm also not too savvy on just what constitutes that "george lucas" level as opposed to a film that "isn't lucas". It seems to me when I see one of his films that the image has a (despite the stories recently sucking, lolol) certain "glow" to it that most images don't have. Maybe I'm talking about something I have not enough knowledge of? I couldn't find the chip. What's the specs?

P.S.

Please don't think I'm dizzin on you or anything, just interested in what's best for all of us. You DO have some valid points.

[Obin Olson]

HD camera should ship thisweek...mean while I am working out how to use a beamsplitter to keep the viewfinder on the K3 16mm camera...too bad I have to take it all apart..this is a very tough wellmade camera that is in good working order ;)

[Wayne Morellini]

I understand Laurence, actually Geoge's camera is a bit inferior compared to regular film ;). I'm just looking for good enough, that $800 sensor will maybe add $2000 to the retail price and another $3000 for 3chip ($1000 each extra and cheap $1000 prism) configuration. I'm interested in good enough for me and us, and cheap enough to drive sales, because that is what will make a product, and bring in bulk discount, so as many people out there can benefit. I think the $200 sensor (only a guess) might be 50% better than what many would accept. We will probably get much greater problem from bayer than from a 10% drop in latitude s/n niose ratio). Realistically the best sensor might be say $10K and only have 10-20% better range than something a tenth or a 50th of the pricee, again a big general market educated guess). But this is where all the good stuff comes in, people wanting a cheap camera could go for the $200 (bulk OEM discount) sensor, people who want the next 10% could use the next level up for double the finale system price, and people who want another 10% can go for a unit that adds 50-100% on the finale price again (or a Kinetta etc). But for all these different after market cameras we could use a common cameralink capture direct to disk/PC editor board (as I have suggested here or else where), and the same plugins. So you get a standard to allow you to put together whatever format, editor, plugin, camera sensor you want with the capture board. I think we can all get what we want, there are a number of different priced camera projects out there.

[Wayne Morellini]

OK,

That Micron chip brochure leaves me scratching my head, I like it, but it says a lot about nothing using some terminology that I am unfamiliar with, could we get some real explanatios on it.

How does True bit work, how does True color work, and what negative effects do they have on picture quality? Some of these sorts of schemes have proved less than desirable in the past.

The 4MP of the unit is spread accross 4:3, so we will be getting around 20% more pixels than 1080 in 16:9 format, but great 720p footage. 200fps at full res is great, but I don't want to pay much more.

Power consumption, good.

Digital rsponsitivity, what on Earth are those figures? What is the real lux range, min & max ratings, at what gain?

Internal Intra scene Dynamic Range ???, maybe I do or don't underatnd that: 59db. OK, I forget how to read this at the moment, but as far as I know brightness doubles every 6dB, bits double every extra bit, so 10-bits = approx 60db, so you would be able to get accurate pixels?? But what about with gain completely up or compleely in negative (gains not even mentioned) as this can effect the range.

The signal to niose ratio is very improtant, you need 60db +6db (for accuracy below the lowest bit) + the desired gain level (that gives grain free footage at the lux desired 5, 3 or 1 lux) (i.e 30db). So for example 96db, gives a stacking clear good picture i low light.

The respone curves of these, and luminace and color would also be good. Pluss all the other regular artifacts, smearing, interpixel leakage causing blooming etc.

Maybe the specs of sensors are a bit much for us, or others, to accurately understand here without explanation from an expert. Maybe we should specify what we are looking for, even if it is "something cheap within reach of the quality of an CineAltar" something to that nature. We could realistically specify no artifacting, dynamic range, s/n, min grain free lux, min gain, desired perforamnce curves etc, but could we ever agree?

Another piont about upscaling, is that single chip footage is effectively allrady upscaled because of the bayer filter, and trying to upscale that might prove a bit more difficult than normal.