Photos of the Drama 2/3" CCD prototype

Status
Not open for further replies.
timbook2 said:
Thanx Otis,

I agree on the fan size and there are good silent ones but I think the fan itself is not going to make many people happy.....even Arri and other film cameras have blimped versions because of the noise issue....I know the REDone has a fan as well and gets some slack because of this....in the end I imagine other cooling options will win.
Click to expand...

There are no other dependable cooling solutions.

Check out the Penelope Delta prototype. Large fan as well. Red and Si-2K have small fans and we know where that leads, high noise, solutions that need to stop the fan and overheat while recording etc. Drama is below 10dB with constant fan operation and has a noise spectrum with very low frequencies, I'm not aware of another camera that matches that spec, and that's probably because it's physically impossible with a small fan.
 
Actually, the fan does have me asking a question - does it port directly into the chassis of the case or is the cooling chamber, so to speak, a separate self-contained area?

If the former is the case (which it appears to be), what do you suggest regarding the eventual buildup of dust in normal environments (as in virtually every computer)?

And what about safeguarding against dust/water/sand/organic matter in less hospitable environments?

These have traditionally been problems for anything with a fan, so I was just curious if you were innovating here as well.
 
Michael Olsen said:
Actually, the fan does have me asking a question - does it port directly into the chassis of the case or is the cooling chamber, so to speak, a separate self-contained area?

If the former is the case (which it appears to be), what do you suggest regarding the eventual buildup of dust in normal environments (as in virtually every computer)?

And what about safeguarding against dust/water/sand/organic matter in less hospitable environments?

These have traditionally been problems for anything with a fan, so I was just curious if you were innovating here as well.
Click to expand...

In all these aspects, the Drama is similar to the typical digital cinema camera. The only camera that could be superior to these is Alexa but I do have some reservations about that design, so only time will tell.

The Drama prototype is a very simple design. There are filtered inlets, some ducts and an large outlet with a fan. The sensor and analog CCD electronics part is fully sealed by design. The storage is also sealed by design. The Drama processing units and all related boards (psu etc) which are in the airflow path are industrial quality components designed for multiyear 24/7 operation in very harsh environments.

I do have a design for a heatsink that allows practical sealing of all the components to make it even better, and I will let you know when it gets tested.

Drama is not the type of camera you need to send to the manufacturer for simple cleaning. If you do use it in the desert for 1000 hours, or in humid environments beyond its specs, it will require good cleaning and I will make sure the design allows it to be done locally without problems.

Even if you manage to damage the parts that are in the air circulation path, which would require 100s of hours outside operating humidity range and in very harsh conditions, a short circuit due to water or something like that, the parts and service cost for a full replacement of these parts will be very low. Less than 15% of the package cost. So, don't think of digital cinema camera service costs where a broken connector costs $100s. Drama will be affordable in long term use due to its design and component quality.

One of the first Drama projects will probably be shot in the Alrgerian desert.

The warranty will be the most impressive in the industry.
 
Thanks for all the replies Otis. I really admire your stamina! Not many people announce and maintain....you obviously are one ! Hopefully I will be able to see it in action one day.
 
Otis,

I understand that your focus is cinema and 24p, but does the camera have any provision for 50/60p at either 1080 or 720 resolution? If so, a version with the sensor separated from the recorder by a cable would be an attractive option for high-end POV work. The current best options are 1/3" cmos block cameras with a nanoflash (cost is ~8000USD), I would imagine a system based on your design would compare favorably.
 
Mike Sertic said:
Otis,

I understand that your focus is cinema and 24p, but does the camera have any provision for 50/60p at either 1080 or 720 resolution? If so, a version with the sensor separated from the recorder by a cable would be an attractive option for high-end POV work. The current best options are 1/3" cmos block cameras with a nanoflash (cost is ~8000USD), I would imagine a system based on your design would compare favorably.
Click to expand...

On the basic 24/25p edition, 1920x800 scope is 1.5x slow motion (36fps vs 24fps).

I have tested a 50p 1920x1080p version of the board recording to ram for 20 second segments then flashing to storage. Very high end SSD can do unlimited 50p 12bit uncompressed.

POV solutions using multiple heads and a central rackmount unit for remote recording, control, sync and monitoring are not very hard to do and they can compete with 8,000 usd for more than 4 units. A unit will be about 16x8x8cm including a prime lens, and less than 800gr. But depending on the project it might not be very practical due to the ammount of data. If it's a high end shoot for a very important event, it could make sense. Remember that it will still be a raw camera, there will be no high quality feeds for realtime use or broadcast. If you have a detailed description of the project requirements, I can tell you if it's possible and at what cost.
 
Congratulations on making the Drama! It looks very impressive!

I have a few questions.

You have been very open with the details. Will this be an "open-source" project?

Is your processing circuitry quantitizing the sensor data logarithmically into the final 12-bit, or does 12-bit correspond to the maximum, "linear," bandwidth capability of the sensor?

Do you have any plans for "full-frame" sensors?

Other than miniaturization, are there any plans for major changes in the "physicality" or configuration of the production models?
 
h-munster said:
Congratulations on making the Drama! It looks very impressive!

I have a few questions.

You have been very open with the details. Will this be an "open-source" project?

Is your processing circuitry quantitizing the sensor data logarithmically into the final 12-bit, or does 12-bit correspond to the maximum, "linear," bandwidth capability of the sensor?

Do you have any plans for "full-frame" sensors?

Other than miniaturization, are there any plans for major changes in the "physicality" or configuration of the production models?
Click to expand...

It's not open source. The main reason is because there has to be some margin to feed the handling of the project, stocking parts, general expenses and support. If it was open source I could offer the software for almost free (minus some software licensing paid to third parties) and offer a support contract for a fee. The problem with open source projects is that you can't depend on support when you need it. Some hardware parts need a 30day preorder. Someone has to store the parts because one month is not acceptable when you are using the camera professionally. Someone has to pay the engineers answering the phone and support email, etc. If it was a pure software project running in a generic computer it could work as open source and supported by an online community.

I store a linear 12bit bayer which covers any modern sensor. I can store 16bits if some future sensor needs 16bits. RAW dSLR cameras also store linear bayer so it's quite similar.

I haven't found any good global shutter sensor that would make sense for this project. The 2/3" makes sense for many reasons. If you are worried about shallow DOF, f1.4 lenses on this 2/3" sensor (10.56mm wide) is identical to f2.8 lenses on 35mm film shooting spherical 1.85:1/1.66:1 (21mm wide).

The production models will be a little smaller than the prototype but quite similar in configuration.
 
I want to come back on the design of the housing.
Totally understand that design follows function, I even more agree with Lawriejaffa that:

" if it could have some of the exterior aesthetic features of an older film cam.
Something that could let this product stand out in an affectionate quarter for indy filmmakers (aesthetics wise) is something venerable rather than gadgety?.... On an aesthetic level it would inspire cinema/film connotations which most of all would be appreciated by actual filmmakers - particularly indy filmmakers. It also allows the product to stand out in comparison to other video cameras."
Click to expand...

An exterior that reflects the aesthetics of the material it produces would be in order IMHO.
Just as you carefully shape the curves of your sensors for pleasing skin tones so you can shape a housing that says "this is a different camera with a different approach - aesthetics is spoken here, not pixelcount"

Having a distinct retro housing would be a statement of focusing to the traditional values of image quality like texture, freedom of artifacts and the "art of engineering".
Values like precision, quality, rock solidity, tradition, "works like a swiss watch" come to my mind.

It would be great to find a way to reflect all the inner values of the Drama on the outside.
Thou it would raise the price a bit and does noting for the material it produces, I'm sure, a lot of people would gladly spend some extra $$ for the option of a venerable housing.

The little mock up I did ( http://www.dvxuser.com/V6/showpost.php?p=2072775&postcount=46 ), was a quick and dirty, in 15 minutes just for fun thing.

But as a former industrial designer, I would love to help you with the design of something like that, if you want too go that route - even if it is just an optional housing. Heck I would even volunteer, just for the sake of it.

Frank
 
Last edited:
Yep agreed - :) Hope you consider this idea Otis, put it focus group etc - you could test a poll here, with a couple of concept illustrations of various aesthetics.
 
CC/NLE/MIX/DI... de-Bayer under development

CC/NLE/MIX/DI... de-Bayer under development

Hi Otis,

I'm working on a de-Bayer and batch image processing program to go with the DI/NLE/MIX/CC program already in "Beta Test" on my website, the programs there are "free" to download right now I just ask people to comply with the EULA and report any bugs they may find.

I would like to look into adding some support for True RAW Bayer data from your "Drama camera" to the set-up menus in my de-Bayer program I am working on, I have several cameras in there now including Acam dII, SI-2K, and Kinor-2K.

Do you have any DNG output as Bayer DNG (not Linear DNG which is like a non-white balance TIF file)?

If you just have a sensor data 12bit per pixel without header, I have support for that if you use one of the supported 12bit formats. If you have another bit packing for the two pixels in three bytes format, then I would have to add your bit order to my code so that it could be selected in a set-up option for your camera models.

Support for meta-data may be an option later, but my code can color correct any RAW sensor data without the need for any kind of meta-data, its just good to shoot a few frames with a black paper, 18% gray, and 90% white so that the peaks on the histogram can be aligined easer.

Our DI system was developed for in house use with our 35mm movie film scanner made out of Oxberry optical printer parts, so we shoot near 4K RAW, make 16bit TIFF files uncompressed, CC using those, make the smaller PIX frames for the NLE to use for sync playback, edit, then output a set of full size frames in order, or a shoot list for my DIY film recorder program (DANCINEL.EXE (tm)). The scanner is run by DANCINES.EXE (tm), and the editing can be done with DANCAD3D.EXE (tm) or DANCAD87.EXE (tm), the 87 version uses the math co-processor in DX processors and runs a little faster for some tasks.

We are now also interested in Digital Cinema cameras that shoot True RAW sensor data, so I have been working on a de-Bayer program that also does other image processing on 16bit TIF, as well as DPX, CIN, and BMP frames, and reads and writes DNG and RAW data.

Anyway, if you are using the same sensor as Acam dII, I know what that looks like, are you using an OLPF on your camera?

Is your IR cut filter interferece or "green heat absorbing glass" type?

I'm asking about the filters since they change the de-Bayer and color matrix settings needed, as you know, so would not match the TAG 50721 settings in the DNG I/O.

Anyway, if your camera is going to save true RAW sensor data, it should be possable for me to work out what is needed to support it if you have some sample frames with or without headers.

We have tested our system to some extent from scan to projecting 35mm print stock exposed off a 1600x1200 monitor using my gamma spread code, using both 35mm negative scans we have done, some from another scanner, and frames from REDONE via REDCINE to 4K TIF. The sound mixing should work with up to about 512 tracks (256 for the film length and 256 for each of 10000 shots) and 16 channels output (making 8x stereo WAV or 16x mono WAV uncompressed image and uncompressed sound mixing). Its desigined for no-budget people to make feature films with a high quality DI all uncompressed, up to 10000 shots per project. Sync playback at 24fps is done with two computers and SMPTE time code (to MTC) to sync the image playback and any good sound editing program used to playback the mix file made in DANCAD87.EXE (tm), such as Magix DeLuxe (tm) which costs maybe $10 on ebay. Its like a 35mm print film and 35mm magfilm interlock system, but all digital so the sync is ment to be sample accurate. Its a film editing system, but does support 24/25/30fps through the non-drop-frame SMPTE time code, you can adjust the WAV to NTSC speeds after the mix, or maybe using another sample size for each frame, like 2002 samples rather than 2000.

If anyone has questions about using "Drama cam" with my software they can email me and include "Drama cam 24/7 software support submission" in the email subject line, thank you, I hope to hear back about the sample RAW data when you have some to share for testing...

Dan Hudgins
tempnulbox (at) yahoo (dot) com
http://www.DANCAD3D.com
San Francisco, CA, USA
 
Last edited:
Status
Not open for further replies.
Back
Top