Wayne: sure, sponsorship would be great. But I'm not sure how
that would work or how to approach etc. Not too sure on who
would actually want to do that (can't really see a reason why).

Jason: he is using the Sony CineAlta, same one as George is
using for Star Wars.

I forgot to mention, if anybody has cameras they can post I will be discussing it on the Home Made HD Cinema Cameras - Technical Discussion thread.

<<<-- Originally posted by Rob Lohman : Wayne: sure, sponsorship would be great. But I'm not sure how
that would work or how to approach etc. Not too sure on who
would actually want to do that (can't really see a reason why). -->>>

I see your reasonning. They do this stuff quiet a bit for the public recognition of being the leading company making things happen (getting the job done), or for profit, or business political reasons (firing a shoot over the shoulder of the competition), or just benevoulance (like their giving moiney to charity). They like to be seen as the ones solving problems. Companies with lots of money are the most likely companies to fund thius sort of thing. The free publicity they get out of it canbe many times the outlay, and keeps them in the industry, or public's, mind as a leading coimpany to do business with. But you don't just storm the beaches with people onmass asking all at once, you have one, or two, select people that work togther to liase with the companies.

We have a bit of a problem though, because the companies in the market might not like to do it for business reasons. So we have to find other companies. As it is a computer, machine vision/surveilence related they are the most likely companies, with computer companies also have the most cash to spend on these sorts of things.

Just think of all the things IBM. Xerox, Apple, MS etc etc etc etc get upto that have no real bearing on their profits, but on their reputation. You can buy industry and customer loyalty through reputation and status.

But at the moment we are so divided between different camps (CL, GBE, FWb, USB2, HDSDI) that require different software for each, with x amount of people doing software, it is going to be hard. One software solution for all cameras would be best. If we did have one solution accross all thesed cameras and the ability to register for each camnera then there is a sure market including broadcast cameras cameras (HDSDI, Firewire). Still maybe we should stick with Cameralink machine vision, whereever it is GIGBE, USB, or Firewire implementation of it. Whatever the interface it is sent accross we end up with the same command, data, and variable structure.

Sumix (I'm sure SI allready has) needs to come up to you and say, Rob, and Rob ;), please support our cameras, what doi you need to know. They may even have their own capture software, but it would be good to clear the air. The Cameralink standards organisation should also support you (not saying they would though but it would be a great new market opportunity for their members) and co-develope a pro caputure tool with you, maybe some members will. But untimately it would be good to have support in windows as standard.


An interesting idea, have you thought of the idea of a translator program (emulator) that translate HDSDI commands to camerlaink, and cameralink to HDSDI, and camerlink data to HDSDI data, making cameralink appear to be an HD-SDI camera so we can use existing capture software? Then we can use Cinerella??

For anbody wanting a Apple Tablet/HF monitor, have a look at this article on a recent Apple patent:

http://www.theregister.co.uk/2004/08...ple_tablet_mac

so what monitor do I need to buy that will have a LVDS connector on it?

I'm not convinced that they'd touch a project like this. Sony or Panasonic could provide a product like this if they wanted, and there is a reason they don't -- they want to keep the price high for professionals. Apple (et. al.) want to stay on friendly terms with these companies, and a project like this could be seen as an attempt to undercut them. Politically I think it would be a hard sell.

If it was discussed privately and the approaches were done privately, instread of here, we would have room to manouver before somebody at those companies finds out. While Apple is cosy with it's editing line other companies aren't. Other comnpanies might be more inclined to support this, as they don't give too much of a rats ... what their competitors PS2 think, or have no connections to the video industry. Twianese, Indian, Chineese and Korean competitiors might not care what their broadcast competitiors think, and might do it just to spite those companies. Don't worry, there are plenty of people that knock even the idea of this project, but solutions canbe found.

OK I have the 3300RGB..Steve what needs to be changed to view this camera with our CineLink software? is it as simple as a larger ROI/area frame size? if it's more can you please send me the DOCS for this camera?

thx

OK, I take a 3 day weekend and get 5 pages in this thread to catch up on. Don't look for a video out from out cameras any time soon. Not until we put some intelligence in first so it will be awhile. For preview, I advocate using 4 pixel "quadlets" to create a 1/4 resolution color image with no processing.

That might work nicely with the VGA touchscreens - 720p is 1280x720 so a 640x360 preview image leaves room for some menu controls.

Just a design mockup for the distant future...
 

Lately I've been doing some modelling and designing camera-looks. So this is just a mockup for the very distant future, where we could really make the camera case any shape we want to. The physical size of this one should be quite small - maybe slightly bigger than an XL2 (with lens).

http://pupuedit.sourceforge.net/camera/pihlaja_HDO4.png

This is not that serious really. I'm just having fun modelling.

after taking test shots with my Canon 10D and the new SI3300RGB I have found that the range on the 3300 is about like the 10D in RAW mode..this is good news to my ears...anyone else?

Also it seems that to show the same amount of color I must boost the 3300 about 45% on the saturation level..not sure why but this chip seems to be less color heavy then others..not as bad as IBIS but not as good as Canon 10D...over all good stuff!

LCD preview, Case etc.
 

Joonas

I agree on the camera case, nice. I read you wanted to do a case, but I couldn't find where you said that.

I was thinking of doing a few cases, myself, but haven't got around to it and finding manufacturers. I have most of the specs and components in my head (very good virtual reality modeller ;). Actually i was talking to somebody, but they seemed to suddenlly dissapear and I haven't heard back, whatever happened to Rai??


Steve N

I agree about the preveiw, we are dealing with GBGR any way, that is close to RGB pixel. If you want a bit more resolution for focusing I suggest that with an RGB LCD with RGB cells in the pixels you can address the subpixels (so a 640*480 display becomes 1920*480 focusable display).

Another suggestion for the preview display, is to add a instant magnifyer feature for focusing, where you touch a button and the centre comes up double size while you hold it.

<<<-- Originally posted by Steve Nordhauser :That might work nicely with the VGA touchscreens - 720p is 1280x720 so a 640x360 preview image leaves room for some menu controls. -->>>

Obin

Is the 3300 presenting more natural colours, and the Canon boosting them?

I think the 3300 is exciting, not because it is the best thing since sliced bread, but because it is a camera we can use (without problems). I look forward to seeing it, the footage from the previouse model looked exciting, this should be more.

Rob I would, and we can only try. It's not going to be simple or easy, but a rejection should come fast, so it is worth a go, otherwise we won't know. Just look at how Bill sold IBM his non existent DOS, I know some of the history that lead upto that. They turned down buying Atari, doing CPM, doing 68000 (probably linux), they even had a crack at buying Apple, but Bill pulled it off. He out did those companies, despite not having their power, money, infrastructure, or reputation, he even had to go and buy the OS from somebody else. I hate to say it, but truely a victory for the little guy (at that stage). If he can do it, we should be able to get some help in return for a million dollers worth of marketable publicity. Something like this should be able to hit the news lines of every major IT news site and magazine, not to mention the video sites and magazines. What we are saying going to them, is that we are here, and we want to make something of it, permanmently.

The world out there is a jungle of competing politics and business interests, even in companies, that get mixed up and stirred around. With certain people, at a certain place, in certain circumstances and times, things are possible, but we won't know until we put the spoon in and pull it out.

<<<-- Originally posted by Rob Lohman : Wayne: sure, sponsorship would be great. But I'm not sure how
that would work or how to approach etc. Not too sure on who would actually want to do that (can't really see a reason why). -->>>

Wayne,

I think there's something you're overlooking in this whole situation when you're thinking of pitching this idea to a big company.

Money spent on marketing

These are all companies that are owned by shareholders. They're not going to be spending a ton of money on projects that they then have to turn around and market for millions of dollars to get sales. How much money do you think Sony spends on marketing? How about Panasonic?

With big company $$$'s comes big company beaurocracy, big company ideas, much of them the antithesis of what we're trying to do here.

Why do you think Linux is doing so well as an underground movement, finally catching the attention of the big boys (IBM)? This stuff that we're doing is grassroots stuff, not big company consumerism.

Now to buiseness :-)

Obin, how 'bout posting some pics from the new 3300 :-)

BTW, were these new pics from your new capture app?

gone for a few days on vacation..hold down the fort for me!

I will be Mountain Biking in the North Carolina Mountains, maybe even get some white water rafting in ;)

I expect to have a working version of our CineLInk capture software when I get back!

Will post lots of stuff when I get back on Sunday

have fun !
keep it real!

I'm jealous :-P

C-mount-to-EOS lens
 

Does anyone know where to get an adapter to let me use a Canon EF-mount (EOS) lens on my C-mount camera? I can find tons of other adapters, but I haven't run across one of these yet.

Not sure if this will work Rob, but at B&H photo you can get a c-mount to t-mount adapter, and then a t-mount adapter for EOS lenses. So again, theoretically that should work, although I'm not sure if it will screw-up the back-focus because of so many adapters.

What's the difference between the 3300 and the 3170 Silicon Imaging Cameras?

Is it price?

Is there a reason why the SI-3170-U is not appropriate for this sort of application. It would seem to be ideal. What is the price of that unit??

Also, for Ben, do you have a raw bayer grab from the Sumix camera available?

Raavin

Jason: please check the silicon imaging site yourself to see the
difference between the camera's. Steve has enough work to
keep up with all of this as it is, and it's easy to look up the specs
yourself. Thank you.

Having said that check out his post on the 3170:

Yes, I probably should have been a bit more clear. I suppose what I meant was, are there hidden negatives, not described in the specs, which make the 3170's negatives greater than the 3300. From reading the specs I just couldn't understand why it wasn't the most obvious choice for this kind of work.

All I could think of was price. Also there seemed to be a bit more flexibility as the 3170 comes in USB2 and Camera Link versions where the 3300 doesn't (as far as I can tell from the site).

I will try to be a bit more clear. I'm not lazy at doing research it's just that not all the information is there (plus I'm new to the deep technical aspects so it's hard to translate some of the specs to actual use ie rolling vs global shutters etc.).

I was looking more for an opinion really. Something that can't normally be reflected in specs alone. Oh and of course the price.

Cheers anyway,

Raavin ;)

There's more than price.

The 3170 is a couple years old, and has a lot of thermal and fixed pattern noise. To try and suppress that you're going to waste a lot of those 12-bits in an attempt to maximize the signal-to-noise ratio and reduce noise. The new 3300 will get you only 10-bits, but those are going to be much cleaner 10-bits that you can actually utilize.

I'm impressed the thread hasn't grown that much this time, with the new camera floating around and all ;).

Jason, you are missing the point we are discussing, which is good press coverage for helping us a little (unless we go back to talking about product support in Apple products). Super Bowl, local baseball, charities etc etc etc. Sometimes they get zilch out of it, some giving to charities even done in secret (called good will), and charities and trust funds for arts, science etc, give for community/sectors benefit. So the principle is out there, and businesses (especially those plush with money) do get involved. But, what I am talking about is much more practically beneficial to business. They would spend very little helping us and in return the press releases and official websites, say they are the sponsors making it happen, which is a big return for their reputation. Big companies have the resources to get on the news wires and get good news coverage, and a news release is as good as an advertisement of the same size (that is marketing themselves). Forget marketing, forget setting up production lines, we are using unique software to use already existing products together for a new purpose. This would benefit SI and all companies (Irony, SBS was just advertising that their coverage of the Olympic Games was brought to me by Microsoft (they even paid for that one ;). We are not just doing this for the benefit of our own personal cameras but it has a greater more symbolic significance that is a benefit to the video/indie community.

Please, have a think of how much money big business pumps into Linux (not to mention how much the distributors of Linux versions pump efforts to standardise it, that come from the profit from those non free distro's). The underground works on lots of small bit's (some of them funded under the same model I am proposing), but over all the complexity is great, and in need of authorities, or businesses, to jell it all together. I have said this for years, and it is increasingly dawning on more and more people now. I design my own OS, and I know that a horse designed by a committee is a camel, not a thoroughbred, some big companies know this, and that is why they are trying to genetically engineer the camel into a horse (or OX ;). I know that it is a very tall order. But a committee is not even a committee without structure, and any complex endeavour needs leadership (at least we have the Linux founder). But for us, we have a simple/straight forward package that can be broken up into simple sections that can be done by one, or a few people each working on a subsection. We would ask for help to finance the effort (very cheap, I doubt Rob wants 100K a year programmers wages to do it) maybe some simple driver or codec support in their products. Past this, I am not saying publicly. Sure they could offer to program it for us, but even then it would have to be open source and in line with our mission. I'm sure it's been done before, in Linux, and we don't have to bend to big company corruptions.

I am all for radical new ways to do things better, like having the whole Linux community working out each bit to one coherent and comprehensive model and working towards making it the best in the world. But human nature is that there are so many stubbornness, ignorance, lack of understanding, for sight and intelligence, politics, and competing interests (which even effect highly structured big businesses in top standards organisations (name deleted for legal reasons)) that this is near impossible in Linux (but now I think of it, it should be possible to make a program to organise things and gets past that). One leader starts with one person (Rob in the case of capture), and as I assume he is skilled, he should get much better results than ten unskilled people arguing. If you could see my OS design, and understand the paradigm shifts in it, Linux would look laughable. Actually this paragraph is a little off topic, I think I'll delete it latter.

Re: C-mount-to-EOS lens
 

<<<-- Originally posted by Rob Scott : Does anyone know where to get an adapter to let me use a Canon EF-mount (EOS) lens on my C-mount camera? I can find tons of other adapters, but I haven't run across one of these yet. -->>>

Rob, this was posted earlier (that guy had a few interesting, well thought out things, in his post):

http://www.birger.com/html/ef232_home.htm

It also allows you to controll all the electronic lense funtions from your computer.

Actually I just had a skim (I have been so busy it has been sitting my browser for ages):

But look at the price, yikes, are they for real. When you look at this, it should be able to be done with a few pic's, and actuators, for a tenth the price. I wonder who their target market was for a price like that?

Just looked around their site, interesting stuff, some camera electronics boards, FPGA based boards, and 68Billion colour Led display.

400GB Hard drive, with command queing.
 

Toms hardware has just posted a new article on a 400GB, five platter, command qued harddrive.

http://www.tomshardware.com/storage/20040820/index.html

I've skimmed it, and they left out a bench mark, but still a curiosity.

http://www.seagate.com/content/docs/...x_sata_ncq.pdf

Command line queing is a spec of SATA II, only this drive and a SG Barracuda 7200.7 have it (and only a few Intel chipsets support it).

It follows our discussion of setting up the read/write commands for optimal performance, to make sure the head doesn't have to wait for the desired sector to spin around again. Something that could even hinder a 720p bayer save. I'm not sure it would be better than writing the software to organise itself anway, but it is another consideration for the high end.

I must confess ingorance of how Manufacturers set up their multiple platters, wether all the heads opperate at the same time, or not, independently or in parralell. If they operate all at once independently, that would be interesting, as you could theorectically dedicate platters to capture, and one to the OS and program. Or if in parralell then all the slower inner sectors could be given to the OS etc, and old footage could be migrated there during pause, while new footage uses the faster outer sectors during shooting

Thanks

Wayne.

Wayne,
The heads inside of a hard drive are all connected to the same arm which is moved by a voice coil type mechanism.
A five platter hard drive would have ten heads... so you would need ten individual mechanisms for movement.
I don't know if you've seen the inside of a hard drive, but I doubt you could squeeze that many components into todays current hard drive casings.
Not that it couldn't be done... but we'll probably all be using solid state hard drives long before they would try something like that.

Back!

the trip was great!

Jason just for you I will post a pic or 2 of the riding around Brevard NC ;) ;)

i am waiting to hear from my programmer on the progress of CineLink while I was gone..I hope it was good! will keep everyone posted!

<<<-- Originally posted by Matthew Miller : Wayne,
The heads inside of a hard drive are all connected to the same arm which is moved by a voice coil type mechanism.
A five platter hard drive would have ten heads... so you would need ten individual mechanisms for movement.
I don't know if you've seen the inside of a hard drive, but I doubt you could squeeze that many components into todays current hard drive casings.
Not that it couldn't be done... but we'll probably all be using solid state hard drives long before they would try something like that. -->>>

Thanks for the info Matthew. I imagine somebody might have done it on a high end model, but that is all I wanted to know.

But Wayne, seeing your really advanced skills and capabilities, why don't you just post some good designs and make are work easier.
I mean not your proffessional stuff, just a simple prototype.
You could help a lot making a compression routine, for example.
Thank you.

I was wondering where you got to Jaun. Put it this way, Ben and whoever else=Bayer, Rob-capture, me used to BE = cases but a couple other people want to do it now. I have suggested doing a few credible things in times past, but no interest, so no go, you can guess, you reap what you sow. As far as my research it has been much, and has ended a while ago, so be content. As far as non-professional stuff, it is all professional stuff thats the problem. I get a bone and I don't let go till I got the best possible solution (though I admit, compression using 3D shape analysis is beyond me (maybe if I use Voxels)). It's called getting it and sticking it to the rest of the world. Finance is the avenue, so I got to pull a simple one to make the money for it, and I have just bought a stack of books to get back into games programming, and maybe generate some gravy. So I advise leaving me alone for now, I got a lot to do, and only am half way through the 35mm adaptor threads. Thats why I am into just answering questions at the moment.

I actually was discussing enclosures with somebody, but that person disappeared, and before I go to somebody else I would prefer to see what they are doing. Takes time, which I don't have so latter this year.

BTW I have the case under control...as I have said we will be doing the desing in-house in CAD

Juan?

News on the Altasense Steve N?

Fellows, please, let's keep everything cool and friendly around here. This is such a unique forum, just look at all the huge threads going on. There's an incredible amount of interest here and everything would go so much more smoothly if we keep it civil and professional. Please try to keep that in mind -- and no finger pointing please, it doesn't matter who started what, all that matters is that we all work together and get along as respectful and courteous people. Many thanks to all,

I agree..will do ;)

I got a really ANGRY email from Wayne ;(

Chris please send me my post about Wayne that you deleted off the list. I have someone here that wants to read it after seeing what Wayne sent me in an email today

Hey Obin, how much do you think it'll cost to machine a case?

As of right now I'm estimating around $1,500 with mill time costing from $50-$70/hr. ($70 is for a 5-axis mill).

I can do the Solidworks modeling myself (my brother :)

I am not sure Jason..I will be doing a model in LightWave soon ..I just got an email from my programmer..says he has a beta for me to test later today! can't wait ;) I will let you know how it goes and send a screenshot

BTW Obin,

are you planning any way to adjust the gains on the camera, i.e. the analog gains? How precisely can you set them (individual gains), and also how precisiely can you set the frame-rate?

case
 

my roommate works in a machine shop. he built us a steadycam arm out of aluminum for free. maybe we could help?

yes R G B gain and RGB MASTER GAIN and framerate from 1 60fps ( or whatever your hard disk can fly)

We will have white balance of the R G B analog gains so that we get as much color info as we can in the 10bit file