Home Made HD Cinema Cameras - Technical Discussion

[Wayne Morellini]

Yes I know. I haven't got around to analysing the page.

With my idea, you basically take in a chunk, apply compression, when you come to edit it, you convert/uncompress it to the desired format, and edit and do finale version of film. In work flow terms it is simple. The compression is for storage fo the huge amount of data that maybe usuable in future, and to cut down on the amount of drives needed in recording (and the data rate). You could use it for finale version to, but with 5:1-30:1 shooting ratios that may not be needed.

So if you are taking advantage of file compression is there any chance you will ever go to lossless soon (less the unwanted niose).

Do you think it is a viable interim recording solution, or is processing power way in excess of 3Ghz?

Wayne.

[Laurence Maher]

Wow Guys,

Summix now claiming a cine-worthy camera now in March? This is what I was worried about guys. See, the fact remains that a bird in the hand is worth 2 in the bush. So far, we've heard about cine-worthy cameras "coming soon" from Summix and SI and what not. Every once in a while someone gives vague hints of something their planning to have out as a sellable product soon, at best there are like 3 or 4 guys on dvinfo who actually have a working semi-working prototype of a home-built camera or an HD mod or whatever . . . okay . . . .

The point . . .

I think we might prepare ourselves for the posibility that these things won't really get off the ground until it's far past too late (i.e., by the tiime real, cheap, efficient home-grown cameras are a pratical reality, Sony and Panasonic and them will have a far more reliable cameras on or near the market that when all was said and done, won't be much more expensive for what you get.) I mean, we all know it's not as good to have 1080i at 25Mbps at 4:2:0, as it is 1080p at 200 Mbps and 4:4:4. But when you start to figure it can't be more than a year before someone announces a 3 chip true 16:9 camcorder at 1080 lines / 4:2:0 / and 25Mbps . . . with progressive frames and cine-gama and very possibly pre-compression RGB out . . . well, we all know it's coming.

Meanwhile, 1 month before it's announced, we finally get our home-grown thing going for say 5 k minimum when talking storage and power and laptops and yada yada, and here comes this camera now for just under 10k that's 10 times as user friendly and portable . . . and with each passing month HD projection in movie theaters is more popular and more higher quality methods of video to film transfers are becoming available . . . .

Hate to be the pessimist, guys, but we might brace ourselves for the reality in March when Summix/SI says it's coming in June, then September again, and then maybe if we're lucky, we'll get what we want in December or so . . . if the companies will still put them out after they know how many people are going to get the camcorders to come.

Just say'n I had a (now validated) feeling last year at this time we'd still be looking . . . and I now I've got a feeling about next year.

Let's hope not.

Summix?

SI?

Bueler? . . . Bueler?

Replies?

[Laurence Maher]

Man, I can never remember . . .

Is it better to have "global" or "rolling" shutter for film-like cinematography?

[Richard Mellor]

I don't think the big guys are ever going to give us a raw capture to hardrive. a friend of mine at avid was told directly by sony that
there mission between camera and software was to preserve the tape . not for the sake of a better product . but for the profit in making and selling tapes. fox news waited years to go hi def because the big guys were just so greedy. when nikon makes a 35mm camera the optics and machined precision require great skill. all of us can build a computer equal or better then a dell
this is now happing to the big camera guys. we can buy the same cmos they buy but like the computer we don't have to take the packaged features that they hand us.if they want to try and sell a camera for $75,000 that has a$2,000 1280 x720 camera as its base they just open thereselves up for competion from sumix and the like. the hottest selling audio product was not made by the audio giants It was made by a computer company. the ipod

[Wayne Morellini]

Unlike most people here, I had a grasp on time frames, so thought that 3 months instead of one year was quick to get this camera, and it should have been possible but then everybody turned to Altasens. Now they have turned to FPGA, which is even more delays (to program it might take them six months, I know it can be done in one, but few people are that fast in programming them from scratch). The 3 chip is probably not too much delay. So things have changed and mounted up. The ussual. The camera could also be slotted in between other jobs.

At the moment there are working camera, just not cine grade software, and this is the release date for drake, who have been reworking their prototypes to make it work better for commercialisation. This life is not so good sometimes, I have seen it so many times in the last 20 years it is not funny, all prople can do in front up to help speed it along (extra money would help) or wait.

[Rob Lohman]

Laurence: global shutter is better for cinematic work.

[Wayne Morellini]

Hello, people have talked about these things before here is the Windows XP PDA, toshiba also announced 80 GB 1.8inch drives:

http://www.tomshardware.com/mobile/20041214/index.html
http://www.tomshardware.com/hardnews...14_120939.html

9 new Notebook drives listed, finally 720p performance:

http://www.tomshardware.com/mobile/2...ook_hd-14.html

[Laurence Maher]

Here's what I don't get . . .

If it takes 1 month to program the software for the 3 chip cmos, why isn't this thing done? We could all be using that until the altesens is out. There's always going to be a new chip coming out, guys. If you're always waiting for the bigger better technology, you'll never make a camera. Make one NOW, use it, and when the next comes out, make that then. 1920 X 1080 3 chips uncompressed is as good as one will need to make real feature films, trust me.

P.S.

I'd program the thing myself if I knew how to . . . but I don't.

[Rob Lohman]

Laurance: the problem is obviously time. This takes a huge amount
of research, trial and error, optimizing, programming and testing.

Obin's programmer seems to be the one still moving forward at
this moment (Rai & Markus' camera is finished already!) and
Rob Scott and myself are currently just without time.

That is the whole problem. It is not as easy as slapping a few
parts together, especially not if you want 10 or 12 bit recording.

I for one do this in my sparetime and over the last 3 months this
has just been almost completely absent. Sorry.

[Wayne Morellini]

I missed the bit about one month to do the three chip software, where is it?

Laurence, we have been using two micro cameras and a IBIS camera for a while but it is the cinema quality capture software we are waiting for.

I am very frustrated too, after waiting for JVC HD10 replacement, new HDV cameras, and now for these cameras, but I have no real option at the moment. If i want HD raw, I have to wait.

[Juan M. M. Fiebelkorn]

That seems to be the problem.

Users are users and developers, although being users too, are a different breed...
Normal end users don't have a clear idea about how much time and frustrations are required to be able to get to a product a non technical user can operate...
I, myself as a stupid developer in ANSI C, have made lots of video software which is good for the things I made it, but which cannot be used by anyother guy because of the lack of user friendly features.When it is not working for me I just open it, tweak a couple of things, compile and keep going.A normal user cannot do that and tht's why it takes months to be able to give people what the want.

Just my 2 cents..

[Richard Mellor]

this is a link from a member .
It has some hd clips from that camera.

the street price is around $3,700


http://www.subgunvideos.com/videos/R...rike%20Videos/

[Steve Nordhauser]

Laurence,
I understand your frustration. There are two levels here - one is the basic camera and the other is a fully functional system solution for movie making. We've been selling IBIS-5 cameras for two years, 720p Microns for over a year and 1080 24fps Microns for over half a year. Obin has (two now) SI-3300 cameras and he is getting results on the system solution. If you want an Altasens, place an order now - the next production run is due in early January.

Yes, we are at the mercy of the sensor vendors but there is a development cycle to deal with. Frequently we have our design done, PCBs layed out and fabricated, just waiting for sensors. Sometimes the sensors are cancelled, sometimes delayed due to problems requiring another 8 week fab cycle and sometimes the specs change requiring board relayout before we can start testing.

The bigger problem is the length of time from when a camera is available to when a system solution can be done. This may change when people like Obin have their first product out and just integrate in a new camera front end but the first time through takes a long time as all the decisions are new and require research before implementing. That is why I see working with people doing integration as an investment.

If you need something now, buy an SI-3300-RGB-GR and Streampix and you can be recording 1920x1080, 10 bits, 24fps raw data.....tomorrow. Not elegant but it works.

[Wayne Morellini]

What a day, after findout about the Sony's HD output, I haved just found a Link to programming the JVC HD/PD series from a link over at camcorder info.

http://translate.google.com/translat...lr%3D%26sa%3DG

Very unclear (in multiple langauages) but it talks about programming it in the Toas language, the hidden programming port near the firewire, programming and flash, 24fps, and HD.

Starting a seperate thread about it.

http://www.dvinfo.net/conf/showthrea...threadid=36574

[Laurence Maher]

Wayne,

Do you know for sure that the component out is uncompressed?