Sparta 2/3 RAW in development

[Eren Ozkural]

Maybe I'm not the target market but I (and quite a few others) are straddling the fence between buying an S35 Scarlet and the 2/3.
The ability to use stills & cine glass, shoot stills, DOF/FOV characteristics all make the S35 model superior. However it costs $7K for the brain alone.

For $3.75K+ we get 2/3 sensor/glass with 3K resolution and 120fps capabilities.

Spending almost $8K for 2/3 1080p 60 is not a selling point to me, Tom.

If this was a camera from the big players I could see the appeal for TV stations despite the crippled specs (integration into proven workflow, vendor relationships etc.) but this price point is clearly aimed at lo/no budget indies, i.e Scarletusers.

 

[Tom Lowe]

Well, good point, Eren. I guess whoever comes out first with 2/3 1080/60p will be sitting pretty. Plus, Red has like a thousand times more credibility than these guys. Scarlet 2/3 will probably have better ISO/ASA ratings as well, with CMOS.

 

[Eren Ozkural]

Plus, Red has like a thousand times more credibility than these guys.

We live in good times to be able to say that, don't we my friend? :beer:

 

[Cesar Rubio]

Can you post a link to the raw images from the KAI-02150 you speak of?

Here is an Uncompressed Raw image (debayered to Uncompressed RGB) made with a 1" Kodak CCD (KAI-2093) MVC:

1080x1920 pxs.
24fps
1/24th
Fujinon 12.5mm opened at f1.4

Tiff full resolution 1080p:

http://dr-3dcameraco.com/Uncompressed-Images/Me-24fps.tif

I did not corrected the image whatsoever in post.

Look at the mixed lighting, CCD's are good for that....

This is the "old" Kodak sensor, the new Kodak 2/3" CCD (KAI-02150) has better low light handling and smear.

The selling point for the Sparta?

Uncompressed Raw at up to 1080/60p!

When It gets out , I will buy one of those.

Cesar Rubio.

 

[Tom Lowe]

Welcome, Cesar. Thanks for the image.

 

[Cesar Rubio]

Welcome, Cesar. Thanks for the image.

Thank you Tom. :beer:

I hope my post will last, I did not come here to promote anything, just to "clarify" some things when needed...just like Jim does in other forums....

CR.

 

[Eren Ozkural]

Interesting capture, it does seem to handle the mixed lighting situation well (to my eye).
What applications and NLE's do you see working with uncompressed RAW?

EDIT: By the way, your participation is much appreciated The modern field of digital cinema is turning out to be quite an exciting arena, thank you for coming onboard Cesar.

 

[Cesar Rubio]

Interesting capture, it does seem to handle the mixed lighting situation well (to my eye).
What applications and NLE's do you see working with uncompressed RAW?

EDIT: By the way, your participation is much appreciated The modern field of digital cinema is turning out to be quite an exciting arena, thank you for coming onboard Cesar.

Thanks Eren.

I captured the image with Streampix 4. (SP-4)

I used an "industrial" camera (Machine Vision Camera= MVC)

Streampix 4 record Uncompressed Sequence files (SEQ), then in post you can debayer the images with different settings that you prefer.

SP-4 Is not like Red Cine, is more "primitive" (read basic) but it does the job well anyways.

Other options is to export the SEQ to Cineform Raw, which gives even more correction settings with Prospect/Neo 4k.

I really love the newly First Light included on Cineform, see here how it looks:


http://dr-3dcameraco.com/Cineform_Videos/First_Light-2.bmp


BTW, That image with my kids was my first recording win Streampix 3 directly to Cineform Raw almost 2 years ago ( the compression is made on the fly just like the Red camera does it for Redcode)


CR.

 

[Cesar Rubio]

Sorry I forgot.

When you debayer the SEQ files inside STP-4, you can do so to an AVI Uncompressed RGB file (my favorite), or use any other compression codec that you might have installed in your computer including for MOV files for the Mac platform.

I work with Premier CS4 now (on Windows XP), but I also have Vegas 8 and I use it sometimes.

FCP can work really well with MOV files as you already know.

Here is a screen grab of Streampix 4 (recording stereo 3-D):

Full resolution bmp image:

http://dr-3dcameraco.com/troubleshotting/Streampix4_preview.bmp

Thanks,
Cesar Rubio.

 

[Joseph Ward]

made with a 1" Kodak CCD (KAI-2093) MVC:

1080x1920 pxs.
24fps
1/24th
Fujinon 12.5mm opened at f1.4

Tiff full resolution 1080p:

The selling point for the Sparta?

Uncompressed Raw at up to 1080/60p!

Cesar Rubio.

Haven't seen you on these forums for a long time? Welcome back. You have to do some post on the 3D section here @ Reduser.:thumbsup:

Question? How does this camera compare to Ikonoskop A-cam dII?

Take care!

 

[Cesar Rubio]

Haven't seen you on these forums for a long time? Welcome back. You have to do some post on the 3D section here @ Reduser.:thumbsup:

Question? How does this camera compare to Ikonoskop A-cam dII?

Take care!

Thanks Joseph.

The Sparta camera?

They both have the same Kodak sensor, and record Uncompressed Raw (as DNG files) so the output should be pretty much alike.

The difference more appealing to me (on the Sparta), is that it will use relatively less expensive Solid State Drives (SSD) media, in contrast to the Ikonoskop more expensive for sure proprietary media (is a SSD too, with different connectors that's all).

Although I like the small form factor of the Ikonoskop a lot, the picture of one guy holding it in an awkward position looking at the viewfinder, is not very appealing to me.

Other than that, they are very good offerings that will compete in the Digital Cinema World.

The good thing of the Scarlet over these cameras is the 3K resolution, and honestly 3K downcorverted to 2K or 1080p looks better than straight 2k or 1080p, just like 720p does from 1080p sources.

But in the other hand I prefer Uncompressed Raw (UR) over any kind of on camera compression, if the Red team makes UR available in the future for the Scarlet, definitelly I would get one too.

And if they chose a 1" sensor over the current 2/3" design (not to downside the pixels more to pack 3K resolution on the 2/3" sensor) the better.

That would give almost the same pixel size than the Red camera.

The 1" sensor is 14.3mm wide /3000 (pixels)=0.0047

4.7 micron pixels is=22 micron sq pixels

The Scarlet 2/3" is 3.2x3.2=10.24 sq pixels

Double the pixels size in a same sensor design would give more light sensitivity and DR.

But then, if they make such a camera in that price range, who would buy the other more expensive cams?

I wish all the best to all camera competitors.

We can all have many cameras, not of just one brand...and still be happy! :thumbsup:

CR.

 

[Dan Hudgins]

Uncompressed editing for Sparta and Acam dII

Uncompressed editing for Sparta and Acam dII

What applications and NLE's do you see working with uncompressed RAW?

My "freeish" feature film DI system can be used with anything that can make TIF (48bpp 16bit) or BMP (24bpp or 8bpp), so it should work with frames from Sparta. Supports 10000 shots per project (0K to about 8K), and (256x10000)+256 audio tracks mix down to 16 channels, color corection, playback at 24,25,30 in sync with a second computer(s) (PC/MAC) through SMPTE sync.

We have been doing filmouts shooting 35mm Vision 2383 print film through a ROSCO #22 filter off a 1600x1200 LCD using my DANCINEL.EXE (tm). Tests on Canon footage using H.264 had some issues with frame rate conversion and macro noise from the compression, so Sparta being true 24fps uncompressed RAW would be a great advantage.

If anyone has questions about doing a feature using Scarlet, Acam dII, or Sparta or all three using my system they can feel free to email or PM me for whatever free help I can give them...

 

[Dan Hudgins]

Thanks, OLPF?

Thanks, OLPF?

Here is an Uncompressed Raw image (debayered to Uncompressed RGB) made with a 1" Kodak CCD (KAI-2093)

Cesar Rubio.

Thanks Cesar for pointing the link to your sample TIF file. I was wondering if the camera you made that image with uses an OLPF (Optical Low Pass Filter) over the sensor or not?

 

[Jose A. Garcia]

Hi Dan, hi Cesar and hi everyone,

I'm Jose A. Garcia, from Cineraw.

First of all I want to thank everyone for the interest in the Sparta.

About the first capture posted, you don't have to worry at all. These images were just one of the tests we shoot everyday, recorded directly to TIFF 8bit. We're in the process of testing our hardware and software so for now we're not concentrating on giving the best possible image, but on making progress as fast as we can. These daily tests confirm us that every step we take is right and the camera's working ok.

For those worried about compression and noise, we're proud to say we have one of the best sensors in the market in terms of noise levels and also we'll be recording to uncompressed RAW DNG 12bit as well as PNG 8bit so you'll always have an uncompressed option when shooting with the Sparta.

Of course we expect to see you all in the Sparta Week this September in LA. In fact, if it's possible and more cameras are released by those days (Scarlet, A-Cam dII...) we'd love to have a side by side comparison during the week.

Dan, we got your mail. You'll get an answer in these days but of course we'll count on you.

Cesar, as always, thanks for your support.

And for everyone else, if you have any question, please don't hesitate to ask.
__________________
Jose A. Garcia. Official Spokesman for the Cineraw Sparta 1080p60 RAW DNG Project.
http://www.cineraw.es/

 

[Cesar Rubio]

Thanks Cesar for pointing the link to your sample TIF file. I was wondering if the camera you made that image with uses an OLPF (Optical Low Pass Filter) over the sensor or not?

Good question Dan,

No those images don't have an OLPF yet.

I am in the process of researching some OLPF for my cameras, because indeed MVC need them (as any other cinema camera or camcorder BTW).

Thanks,
Cesar Rubio.

 

[Cesar Rubio]

Hi Jose:

I was wondering how you guys are going to make the camera output Raw 1080/60p at 12 bit.

With my cameras when I record 12 bit (Raw-12) or 14 bit ( over Raw-16) the maximum sensor frame rate is reduced accordingly...-1.5 times for Raw 12 and -2 times for Raw 16.

Since the Kodak 2/3" CCD sensor is capable of 1080/60p at 8 bit (Raw-8), I think that the most it can do at 12 bit (over Raw-12) is 30 fps (or 40 fps but who uses that odd number?).

Thanks,
CR.

 

[Jose A. Garcia]

They both have the same Kodak sensor, and record Uncompressed Raw (as DNG files) so the output should be pretty much alike.

Sorry to correct you here Cesar but we considered a Kodak sensor way back when this was the BLUEcam project. Now everything's different and we've said many times we're not using a Kodak CCD anymore.

 

[Cesar Rubio]

Sorry to correct you here Cesar but we considered a Kodak sensor way back when this was the BLUEcam project. Now everything's different and we've said many times we're not using a Kodak CCD anymore.

Oh I apologize then Jose....

I would be very interested in knowing what company makes the CCD sensor for your camera then, but I guess that info wont be released to the public just like in the Red cameras...

It is a Sony?

Probably not, AFAIK the only ones fast enough for video that they sell externally, are capable of 720/24p (2/3" CCD)...it seems they don't want to give the "best weapons" away to protect their cameras line.

CR.

 

[Chris Nuzzaco]

My first thought about their web page was: help.

I have a hard time trusting people who don't seem to care about their public image. Just me.

 

[Dan Hudgins]

OLPF info.

OLPF info.

I am in the process of researching some OLPF for my cameras,
Cesar Rubio.

At f/1.4 most of your tif looked good as far as aliasing and chroma moire, but there did seem to be a little aliasing on your nose which was the most in focus, I can see at smaller stops bigger issues could show up.

Have you seen these links, there is some info in them on OLPF,

Link about Epson's new OLPF filter,

http://www.epson.co.jp/e/newsroom/2007/news_20070906.htm

http://www.epsontoyocom.co.jp/english/product/Optical/set08/olpf_4/index.html

Here is another type, they have some files you can read on their site,

http://www.optics-online.com/lpf.asp

==

The OLPF needs to be larger than the sensor so that the edge rays from the lens do not get cut off.

Since these are about 2mm think they add negative spherical aberration when fast lenses are used and introduce some focus shift issues between working full open and stopped down, many lenses are not corrected for having such a thick plate between them and the focal plane. To combat the loss of focus the sharpen can be made stronger, but that adds noise etc.