February 13, 2012

The Latest on the Chinese KineRAW Digital Cinema Camera

The KineRAW is a Chinese digital cinema camera that I called (perhaps too harshly) a "knockoff" of the ARRI ALEXA and RED cameras. This is a guest post by Wentao Marvin.

As China is celebrating the Lunar New Year (the Year of Dragon), Kinefinity KineRAW development has entered the final stage of internal user testing, and public testing is scheduled to roll out in February. This update focuses on the S35 (Super35-sized sensor) and S8 (Super8mm-sized sensor) model as they are being developed back-to-back.

Here are some images showing the S35 camera body design:

A few key specifications of the S35 model:

  • Super 35mm sized CMOS sensor
  • Dual raw recording, uncompressed 12bit linear cinemaDNG (800Mbps) or 10bit log90 GoPro Cineform RAW (100Mbps), uses regular 2.5 inch SSD or HDD (shoot and dump)
  • DCI 2K 1-24fps, 1080p 1-30fps, 720p 1-60fps. (Not yet finalised)
  • Interchangeable lens mount, supports PL (standard), Canon EF, Nikon F, OCT-19
  • PL mount version 280mm*135mm*155mm, approx. 4kg.

I've been told that the dynamic range of the sensor is likely to achieve or exceed 13 stops. However the S35 model is still being fine-tuned, the actual footage is estimated to be released in mid Februrary. Here is the latest information on the camera, in the form of Q&A:

FAQ

  • Q: Why there are HDMI and VF sockets on the front of the camera? A: They are intended for compact monitors or EVF, two SDI output on the back are for reference monitors.
  • Q: What is the function of that antenna? A: It will be used for remote control (eg. smart phones or tablets), also a wireless monitoring feature is being considered.
  • Q: How many recording buttons are there on the body? A: One on the left, one on the back, one remote trigger socket on the front.
  • Q: What power supply can I use? A: the S35 model uses 4 pin XLR, supports standard 14v battery, a 140Wh battery can last approximately 4 hours.
  • Q: How many accessory mounting options are there on the body? A: There are two 1/4-20 mounting screws on the left and right side respectively, one 3/8-16 screw and 3 rows of four 1/4-20 screws on the top.
  • Q: It was mentioned that the S35 model will have an EF mount version, is the aperture electronically controlled? A: Currently the sample unit cannot electronically control the aperture of EF lenses, however it's still being developed and might be able to do that upon release.
  • Q: Does this camera have Genlock? A: It doesn't have Genlock but it will feature a special interface for 3D or multicam sync.
  • Q: How much is the S35 model? A: Approximately $8000 USD body only.

The S8 model shares most of the key features of the S35 model, two test video and some uncompressed frames can be viewed below, please note that these footage do not represent the actual quality of the S8 model since it's still a sample unit. You can download high quality .avi files of each test here and here.

Sample uncompressed 1080p frame  grabs(processed from 2.5k cinemaDNG) can be found here: 1, 2, 3, 4.

I will provide Koo with more footage as soon as more is available, please stay tuned!

Your Comment

135 Comments

Looks interesting. The Super 8 footage does look quite filmic, but I have a question. What's a Super 8 sized sensor camera good for?

February 13, 2012

-1
Reply
Eugenio Fernánd...

Documentary where a shallow depth of field is normally more problematic than it's worth.

February 13, 2012

-1
Reply
D

Because of the high pixel resolution of S8's 2.5K, when shooting with good f/0.95 lenses you can also get shallow depth of field. There is a 200% zoom on the HDMI viewfinder that lets you focus with greater certainty.

February 21, 2012

0
Reply

Be sure to see the comments on page two as well as there are more than 90 now...

March 3, 2012

-1
Reply

Maybe you only see one page, you may need to re-fresh your browser after making a post...

March 3, 2012

-1
Reply

About the page at the bottom that says the camera is 'rolling out' in FEB, that's a bit ahead of what's going on, they got two revised prototypes back from the factory and are shooting with those now, I have been getting some calibration frames this week, and the camera seems to be working enough for that, I am told that its shooting both CinemaDNG and Cineform (tm) although we are working on the camera monitoring LUT because those are needed not only for the camera monitoring but also for making 3D-LUT for various workflows like Iridas Speed Grade (tm) and Cineform (tm) to have color matching.

If things go well sales of the KineRAW-S35 (tm) may start in a few months. You can email Kinefinity.com (tm) and tell them you are interested in being kept up to date about when the camera will be on sale outside China. If you google KineRAW (tm) from time to time you may be able to follow what is going on. I still plan to make some more KineRAW-S8p (tm) test demos for my Vimeo pages, and to do additional testing and calibrations as the camera designs get closer to their public release dates.

You should note that the AVI linked to here ARE NOT uncompressed AVI and show heavy losses over the 100% uncompressed samples, even though they show quite good results. The reason for that is that the frames for uncompressed AVI are 1920x1080x3bytes each, or 10.4MB each, so for 6 minutes of video it would be about 1.5GB to download. The camera Bayer frames are 12bits per pixel (1.5 bytes is 12bits packed two pixels in three bytes) so the do not take as much space as uncompressed BMP frames used in the uncompressed AVI file. The AVI I sent to Kinefinity.com (tm) and Vimeo (tm) were compressed MPEG4v2 to make them small enough to meat Vimeo (tm) limits and so have lost some of the crispness and fine detail that the camera can capture. If you want to see some of the uncompressed BMP frames used to make the demo-videos you can email me directly at: tempnulbox (at) yahoo (dot) come and put 'NOFILMSCHOOL reader BMP frame request' in the subject line, that way I know you are not junk mailm thanks.

If you have any questions about when the cameras will go on sale or when additional videos from the S35 will start to be shot, you can email Kinefinity.com (tm) directly since they are in control of those dates.

March 3, 2012

-1
Reply

It good for using Super8 C-mount lenses like the Angenieux 6-80mm f/1.2, as well as Regular 16mm C-mount lenses as they are lighter weight than 35mm lenses of the same type, in addition to that the total cost of the camera is lower, as the MSRP is forecast to fall between maybe $1500 and $2000. Because of the 2.5K resolution of the sensor, in the case of the demos, which should only be viewed 1:1 pixel on your monitor as any re-size will degrade the quality, it have oversample for making Blu-ray or DCP for projection in Digital Movie theatres, Arri Alexa (tm) shoots 2880x1620 and these samples were shot at 2592x1104 so there is only a pixel difference of 288 pixels in the sensor active area across the frame. That oversample helps reduce some de-Bayer artifacts. Its an alternative to using a HDSLR for making a micro-budget feature film, and should work for that function if used within its limitations.

February 21, 2012

0
Reply

Frame grabs look great. Samples I have seen (the ones you posted have been around a while) do indeed make it look analog and close to film.

Should be a worthy product if it hits the market place soon.

February 13, 2012

0
Reply
jordan carr

Thank you for the complements, after I get done with the anamorphic test reel, I'll try to shoot some shorter test reels that focus on particular looks and lighting types. If people have suggestions of what types of shots they would like to see they can email me with 'camera shot suggestions' in the email subject, my email is at the top of each page at my web site, google DANCAD3D to find that.

February 21, 2012

0
Reply

Very interesting indeed! Looking forward to more developments :)

February 13, 2012

-1
Reply

Feel free to contact Kinefinity.com (tm) directly as their email in on their web sites,

http://www.Kinefinity.com

and

http://www.KineRAW.com

If you have an interest in one of their camera models please tell them as that helps gauge the interest and project the sales volume so they can bargain for the best price with the factory orders. You can also get answers to specific questions about details that are important to you as a filmmaker, and they can also learn your concerns about matters that are important to you personally.

February 21, 2012

0
Reply

Cool!

Though the "super 8" model would have been better off as a 1/2 or 2/3 chip to use readily available lenses for Super 16. There are plenty of excellent and cheap 1/3 cameras on the market, though not 2/3 ! With few Super 16 digital cameras available there is a huge supply of lenses waiting for an inexpensive camera to mount them to.

M

February 13, 2012

1
Reply

They've got an S16 cam in the works too...apparently with a CCD sensor, meaning no rolling shutter!

February 13, 2012

1
Reply
Gabe

The sensor is a bit wider than Super8, when shooting 2592x1104 the image diagonal is close, the sensor works well with 1/2" and 2/3" megapixel lenses of good quality as well regular 16mm lenses such as Schneider and Kern Switar. Its also possible to relay optics to shoot with 35mm movie camera or still lenses, rather than just a DOF adapter, and in that way it should be possible to use Lomo anamorphic lenses with the 1828x1556 pixel mode, or even rent Panavision (tm) anamorphics and make a mount to fit those to the relay optics. If you have an interest in shooting with other than C, CS, and D mount lenses on the KineRAW-S8 (tm) you can contact me and I can tell you more about how that could be done. As for the KineRAW-S35 (tm) it has a changeable lens mount system and so you can use various 35mm movie or still camera lenses on it without relay optics.

February 21, 2012

-1
Reply

It will perform as good as a Chinese car!

February 13, 2012

0
Reply

What a dumb thing to suggest. First, I'd be willing to bet a fair amount you've never actually driven a Chinese car. But plenty of the electronics you use -- including, in all likelihood, the computer or phone you used to make that comment -- are manufactured wholly or partially in China and they work just fine.

There are plenty of things wrong with China, and plenty of stuff that gets made here is crap, but there's plenty of good stuff too. Is this kind of comment really helpful in doing anything other than perpetuating a stereotype?

More importantly, did you actually look at the frame grabs or uncompressed footage?

February 14, 2012

-1
Reply

C. Custer, Pls tell him!

February 14, 2012

-1
Reply

the fact you are alive to type that comment proves you have never driven one either. Have you seen the crash test videos? Chinese labor is great, chinese quality control is not.

February 14, 2012

-1
Reply
kevin

As you type this on a low quality computer made in china. Remember they used to say the bullshit about Japanese products.

February 14, 2012

0
Reply
JEF

hahaha, ok first of all i find it funny i am being called narrow minded and controlling media while China is notorious for censoring. What I am stating is FACT. People take their designs, they build factories in china and they hire Chinese to work in them, this does not mean its a Chinese product. Like I said watch the crash test videos, the cars crumple up like paper. I dont hate Chinese people, in fact I am half Chinese! What I am saying is China has a LONG way to go before it catches up to the quality control and safety standards of the rest of the world. If you think different im sorry but you are just wrong and closed minded. I have my masters degree in mechanical engineering and have plenty of friends in the auto industry and this is simply a well known fact.

February 15, 2012

-1
Reply
kevin

@kevin

I do believe you are the one that is closed minded. The top Audiophile equipment in the world now comes from Chinese companies. I own a Shanling CD player (valve based), and a Melody (Australian brand, Chinese made) valve amplifier. They exceed anything in the market less than $10000. They look weird. Chinese styling is strange to say the least, but I would not go for any American or Japanese made equipment for twice the price (if money was no object).

In fact, I have actually stopped buying Japanese equipment because their quality control is so poor now. I threw all my Denon, Yamaha and Sony gear out of my second floor window in complete disgust. I then contacted the Japanese consulate to tell them that because of their poor quality and arrogant sales and service (including after sales service) that I would NEVER buy another Japanese product. I signed a contract with Canon two years ago stating that I would buy no more Canon products after I had a run of half a dozen dead Canon products (I have a loophole allowing me to buy more Canon lenses for my Canon cameras, but I have never used it as I prefer reliable equipment from China and Russia). In fact, the ONLY problem I have with my Volvo is the crappy Japanese made capacitors they used in the speedo. I need to replace them with reliable Chinese made Capacitors.

February 16, 2012

-2
Reply
boybunny

Speaking of Chinese Capacitors. My Schaub Lorenz TV packed it in a couple of months ago. It is Korean made with a Samsung (Korean) screen. Guess what needed to be replaced? Yeah, the idiots used Japanese Capacitors. It has reliable Chinese Capacitors now.

February 16, 2012

0
Reply
boybunny

@boybunny Ok you are clearly talking out of your ass. Toyota is the number one auto manufacturer in the world. In fact much of the word (i've seen this in america and Europe) are moving to adopt its business strategies and models in their companies. As for canon im not sure if you are talking about canon cameras and lenses but you care clearly making that up. Im also an avid photographer/videographer and have used many of their products extensively, and it and nikon are the most trusted brands in the industry as far as sports and wedding photography go. For shear quality you cant beat the Germans tho, leica lenses are the best of the best and German cars are so over engineered its insane. I see you didnt mention anything about saftey standards in general and cars. China is a great developing country and is improving daily. But they still have a long way to go before they catch up.

February 17, 2012

1
Reply
kevin

@Jeff

I believe you have dificulty with reality.

Toyota sells morte cars because they sell cheap and have large global advertising campaigns. More people eat Chuppa Chups lollipops than they do Beluga Caviar.... does that make Beluga Caviar inferior? In your mind, I suspect it does.

I had a serious issue with ALL my canon puchases. high end printers, corporate printers ($80000 POS trash heaps), my 20D had a list of issues (known issues in body and lenses), and professional image scanners. Not one of these Canon products worked when purchased, and not one was repaired by Canons incompetent repair teams, with the exception of the 20D and lenses. Do you want me to post the contract I signed with Canon that Canon wrote up on Canon letterhead where I agree to never buy Canon products for an unspecified payout from Canon? Do you actually believe that Canon would pay out a customer for their faulty goods for no reason? You may have the best of luck with your Canon equipment. But there is always a failure rate for every product, and somehow I got a long line of expensive faults.

Chinese cars almost a decade ago had known safety problems, but within months of the reports, all local Chinese made cars had safety features like airbags fitted. More improvents have been made in the past decade. Maybe you sould stop living in the past. I get the feeling that if you were alive in the 70's and early 80's, that you were the guy that said "Jap Crap" over and over until no one wanted to talk to you.

February 25, 2012

-1
Reply
boybunny

It's a fearful situation when people of such narrow perspectives get control of tools of media.

February 15, 2012

-1
Reply
Bobkat

I didn't say it was a good or bad thing, you made up your own conclusion which is interesting.

February 15, 2012

-2
Reply

What you meant was quite clear.

February 15, 2012

0
Reply

That's what I thought: most of the electronic devices we use today are already made in China, so they definitely got the materials to make a good camera. The question is will they be able to pull off the engineering and useability part of the equation.
The footage looks quite good, so the question will be: what will be the market price for these cameras, and will they be reliable tools?

February 17, 2012

-2
Reply
Heiko

My iPhone is manufactured in China, but that doesn't make it a Chinese product.
There's a huge difference between manufacture and design, and I think the point here is that Chinese design has a long way to go before it catches up, especially when it comes to reliability.
Chinese cars are garbage. They may look nice, feel nice and even drive nice, but hit them with anything harder than slippers and they crumple like sheets of paper. Same goes for the quality of every Chinese "knock off" I've tried or owned, so I think it's perfectly reasonable that people are skeptical about this camera.
Having looked at the images linked from here, I don't think established camera companies have anything to worry about. Quite frankly, they look like ass.

February 21, 2012

-1
Reply
Jack Frost

Do you have any particular points about the images? I'll be shooting some more and am interested in what camera's samples in particular you are making the comparisons to. If you would like to be constrictive, it would be useful for you to email me so we can talk about what subjects and grading styles you think would show a more attractive result.

February 21, 2012

0
Reply

People who are interested can make their own decision about if they are interested in using the cameras for their particular projects. Some people like one thing and others like another, some people like fast and some image quality. There are no best or worst in particular, even if you shoot 35mm there are many trade-offs regarding how you scan the film, and if you make a DI or contact print. 35mm film can vary greatly in how it ends up looking on the Blu-ray depending on how it is de-grained and filtered in addition to how it is graded and encoded. The point of RAW recording cameras is to give a wide range of possible work-flows and end results, with compressed HDSLR that only record 8bit data, you have to live with the way you recorded the images since any major changes to the 8bit data in grading will introduce histogram gaps. With a RAW recording camera you have the actual sensor data to archive and go back to if you want to change the grading. If that matters or not to the end user depends on what sort of end use he has in mind. For uncompressed transfer to 35mm release print and making DCP you are likely to see more from an absence of H.264 artifacts, perhaps.

February 21, 2012

-2
Reply

...Like American cars are so great.
Why did GM and Chrysler go Bankrupt.
Capitalism works with US Govt. Bailouts.

February 14, 2012

-1
Reply
sammy

they couldnt have asked that man to be more awkward haha

February 14, 2012

0
Reply
Robert

I guess you mean my Brother in the B1 reel? He was helping me shoot some test charts for calibrating the EI/ISO and K values of the camera's viewfinder. He has spine disease and is in constant pain. I was not sure if I should show those shots, but people were asking me for some flesh tone shots in various kinds of light and that was one of the few shots I have processed so I put it in the reel along with a bunch of other tests that I had been making.

February 21, 2012

2
Reply

Looking good!

February 14, 2012

1
Reply
Mark

Those were two of the most painfully dreadful camera samples I've ever seen. Ha! Good info, though

February 14, 2012

0
Reply

Huh...???...That footage with a S8 sensor looks better than RED.
RED RAW = Compressed Bayer

February 14, 2012

-1
Reply
sammy

Referring to the content and execution. Sorry for the confusion!

February 15, 2012

2
Reply

I'm sorry you liked neater the content or execution, what suggestions would you have for making future demo reels more to your liking?

February 21, 2012

-2
Reply

If you think this looks better than RED, you're either blind or trolling. Quite frankly, the TIFs look like crap if you view them at anything bigger than a cell phone. Compression artefacts out the wazoo. I remember Video for Windows looking like this about 15 years ago.

February 21, 2012

0
Reply
Jack Frost

As far as I know there is no compression in the TIF frames obtained from me in the first place, I don't know how they ended up here, maybe if they had asked me I would have suggested using some other frames as better examples. I have some BMP frames used to make the videos with that are the uncompressed source frames you can read instructions on how to get some on my Vimeo page.

February 21, 2012

-1
Reply

So this camera is from China? I wonder why the test footage was shot in San Francisco... Guess it doesn't matter, but its going to bug me. Looks great though.

February 14, 2012

-1
Reply
Wayne

The footage was shot by Dan Hudgins, (http://www.dancad3d.com/) who is testing software with the camera for this company, and he's based in San Fran

February 14, 2012

1
Reply

Cool thanks! I live in SF so it was interesting to me.

February 15, 2012

2
Reply
Wayne

Holy Christ, the excuses people will make for the sake of it.

February 15, 2012

0
Reply
Bobkat

The cameras can shoot uncompressed CinemaDNG, and I have been developing a free de-Bayer program that supports CinemaDNG recording cameras, like Acan dII (tm), SI-2K (tm) (via their converter program), Aaton (tm), Kinor-2K (tm) and others. I offered to help calibrate the viewfinder and needed to shoot some footage to shake down my full uncompressed DI/NLE/CC/MIX feature film finishing system that is currently up for download on my web site for 'beta-testing'. Its only one of many possible work-flow options, but for the S8 its no-charge now, so it gives micro-budget filmmakers ca chance to get out from under H.264 like compression when making a feature film for theatrical projection.

The S35 and S16 models are also planed to support Cineform (tm) wavelet compression, like the REDCODE (tm) compression used in the RED cameras like Epic (tm), Scarlet (tm) and RED-ONE (tm).

For the alternative work-flows they have been working on 3D-LUT support so that color matching can be obtained with the camera's viewfinder LUT settings made while filming. In addition to that the DNG frames also support some color matching through the DNG tags and meta-data.

Since there were some technical details of the SMPTE time code and meta-data to work out as well as the color matching, I offered to do some testing and development of my support for their cameras. I also have an interest in using their camera for a project I have planed for some time.

February 21, 2012

1
Reply

There's something fishy about the shot of the bird in the second video at 0:26. It looks like it was keyed to give the impression of more dynamic range.

February 14, 2012

1
Reply
Luke

It's almost certainly just a very naive (temporal) noise reduction filter. The wood grain is smooth enough that small movements aren't actually treated as "movement," so you just get a static image that is the average of all of the frames. The transients (e.g. pieces of bread) have sharp edges so the filter doesn't average them out.

It is pretty weird looking, though.

February 14, 2012

-2
Reply
cows

Any of the de-Bayer settings in my programs can be changed or turned off. Most of the shots are what they are as tests, its not much more than that, people were asking to see something, so its something. I hope to produce some additional tests, and when the cameras get out for wider testing, who knows what you may see.

If you don't like what results I have made with my programs, there are an almost infinite other ways to process the camera data, and then you can talk about what you think of those results.

February 21, 2012

1
Reply

Yeah I noticed the same thing, especially the ground on the bird... looks key framed.

February 14, 2012

0
Reply

The camera shoots 12bit uncompressed so has more dynamic range than shows in the videos which are only 8bit (24bpp). In the case of the bird shot I darkened the background. I'm not sure what you are implying by saying 'keys' were used? It's not a matte shot, just a heavy grade using a S-curve to convert the long range and rather flat image to have more mid-tone contrast.

February 21, 2012

0
Reply

Picture seems to be ok, colors appear un-natural like small sensor cameras, but then again that could be fixed in post. A very strange selection of shots which don't indicate much. Would like to see nigh shots, low light shots, shallow depth,skin tones, not seeing anything impressive at this stage. But picture aside, I would need to see lot's of tests and reviews before I could trust the reliability of such a camera. the key'd bird, the moon, the milky microscopic shots, what's all that about? and I don't see the point of S8 & S16.
$8,000 for a Chinese camera is a little high for my liking. But I'll keep an open mind and wait for mid Feb.

February 15, 2012

0
Reply

The point of S35...S16....and S8 is it offers choices in terms of Depth of Field..camera body
size and weight..lenses..and cost of $$$. But all the cameras shoot True Raw 2.5K which is
about 100 MB/s....Red Scarlet is 55 MB/s...Common sense will tell what has better color.

Red gives you no choice. All S35 REDODE. Either first run EPIC sensor or SCARLET sensor
which is EPIC rejected sensors. And All RED Raw = REDCODE,...,,,No TRUE RAW.
Kinefinity will allow TRUE RAW and compressed Cineform Raw...so you have best of
both worlds.

Kinefinity makes able to choose both sensor size and what form of RAW recording. RED does not.

February 15, 2012

0
Reply
sammy

Not rejected sensors dumbass, ASIC boards.

February 17, 2012

0
Reply
Jack Frost

Sorry Mr. Square..obviously am artist than engineer.

February 17, 2012

0
Reply
sammy

A bullshit artist, sure.

February 21, 2012

0
Reply
Jack Frost

BS is Bachelors of Science....BA is for artists..
Proof is in the colors in the clip...
One of these years Red will get color right...
httpAprovideocoalition.com/index.php/aadams/story/red_color_comparison/

February 24, 2012

0
Reply
sammy

Another point is that its hard to see the additional resolution of shooting DNG raw from the compressed videos above and on Vimeo. I did not supply uncompressed AVI to Kinefinity (tm) because they would be around 200GB or more in size, and for most people they would have problems downloading them and having a computer fast enough to playback uncompressed footage.

I have done internal tests and the clarity of the uncompressed playback is noticeable, and would be of advantage for release to digital motion picture theaters as a DCP. The losses from DCP projection in digital movie theaters would be less than from the double compressed version you see on Vimeo, probably.

For making 35mm release prints going all uncompressed from the camera to the release print through a uncompressed DI avoids ALL compression losses. Its a bit sad that 35mm film projection will end before too long, probably, so DCP will be the only alternative, but it may be possible to playback uncompressed frames for digital theater type use with special venue/attraction type releases. Its mostly a matter of the frame server feeding the projector, so as SSD become faster distribution on SSD rather than HDD can make a difference in bandwidth for that application of uncompressed projection.

February 22, 2012

0
Reply

Thanks for the suggestions. The camera has gone through some firmware changes over the last few months and I have been working on various improvements to my software support. My Brother got involved in the anamorphic test reel, so he wanted to add more shots and try to edit it into something since its also a sync sound test reel, and to show a full DI with sound mix.

Once I get done with that, soon, I try to demo some other types of content.

February 21, 2012

0
Reply

The color balance in the demos was graded and processed for 'film like' results. In some cases the skylight adds some blue cast, and the green light bouncing off trees makes some green cast, so those combined when the contrast is increased on an S-curve makes for a off color in the shadows. On top of that the COF shots were made near sunset and I graded out the orange cast, pushing the shadows more toward cyan.

I felt it was better to show how I would grade the shots for use in a theatrical movie rather than to get test chart accurate results, since such test chart accurate results were never obtained by movie film, only when film scans are corrected digitally can they get close to looking like a studio TV camera, and who needs something that looks like a studio TV camera to make movies with?

The camera's shoot and record True RAW sensor data, so you can process that for 'accurate' color as long as you use any needed UV or IR cut filters to compensate for excess IR or UV in the light sources. You can also process the True RAW sensor data for many other looks as you please and use many work-flows as you are starting out with as much sensor data as you can get.

Kinefinity (tm) asked me to make the contrast high in the demo footage as they thought some of the other tests I sent to them holding as much dynamic range as I could were to flat looking. Later I may make some additional 'flat' looking demos so you can see the 12bit data with less heavy of an S-curve applied, but the results will be less 4th generation film like since the S-curve gets more exaggerated on each film generation printing so if you want the results to look like a 35mm release print which may be 4th or 6th generation printed film-to-film, then a corresponding S-curve needs to be applied to the True RAW data when it is processed to make the graded frames.

February 26, 2012

0
Reply

Thanks again to everone who voted for Parasite on YobiTV!

Thanks to your votes I made it to the round 3 finals.

If you voted before or you like the movie please vote again everyday at:

http://yobi.tv/yobifilm/​contestant/17928/dirknienaber/​parasite

Remember you can vote once everyday so the more you vote the better my chances of winning!

You can log into YobiTV quickly and easily with a Facebook or email account, it's really quick.

Don't forget to show Parasite to your friends, family and co-workers and get people voting.

Once again, a big thanks to everyone for thier support.

February 15, 2012

0
Reply

Have approximate prices been revealed?

February 17, 2012

0
Reply

Re S8 price ... following from Dan Hudgins on his Vimeo page for first video above https://vimeo.com/33997025

About the price, that is up to Kinefinity (tm) the makers of the KineRAW (tm) cameras, I told them people were asking and they said maybe about 1/4 the price of the S35 model, so that would maybe be between $1750 and $2000, but the currency exchange rates vary so by the time you order there may be some flux. If they could put a single LARGE order in with the factory they can probably get a price break and maybe pass the savings on to all the buyers, so let them know you have an interest and try to organize a volume purchase and maybe that way everyone can get a discount. Its hard to make something like this in small quantities and keep the price down, you understand that.

February 17, 2012

0
Reply

The S8 is designed for side by side 3D stereo filmmaking at 2x 2.5K, those oversampled frames could then be processed into 4K over-under DCP for projection in digital movie theaters as 3D movies.

The side-by-side option makes for a more compact and portable 3D camera system than using a large mirror rig with two larger cameras.

In that case there may also be some interest for 3D TV program production as the cameras can support 23.976, 29.97, and 59.94 frame rates at various resolutions.

I have not tested the 3D shooting part yet, but my DI system can support the processing and formatting of such a 4K DCP for 3D release, as far as I know.

February 22, 2012

0
Reply

Thanks for all your comments Dan!

February 22, 2012

0
Reply
avatar
Ryan Koo
Founder
Writer/Director

Thanks for the thanks. Both myself and the people at Kinefinity.com (tm) can be contacted for email if there is some question about matters pertaining to my work-flow or the cameras. They are working very hard on support for various professional work-flows such as Go-Pro (tm) Cineform (tm) and Iridas Speed Grade (tm) as well as supporting work-flows based on the Adobe (tm) CinemaDNG SDK such as Adobe (tm) products, and free programs such as XNVIEW (tm) and VirtualDub (tm) etc.

If someone has a question about some camera feature or sees the need for an additional feature they can contact Kinefinity.com (tm). If you don't get a reply from them for some reason you can email me, as I have mentioned my email is at the top of each page at my web site, and its tempnulbox (at) yahoo (dot) com, that is my personal email and I try to check it a few times a day to see if there are any support questions and such, and I can pass on your questions to them for you.

Its best to get the facts about what is going on rather than speculate. For the most part Kinefinity.com (tm) has been very open about what they are working on. For my part, I hope that I have been clear about what I am working on, and that if there are any questions or if you have a problem getting one or the other of my programs working right that you can contact me directly for individualized instructions if you are serious about doing a film with my programs.

For support information about using programs other than mine with the KineRAW (tm) cameras if and when they go on sale, you will have to deal with the respective companies whose products you want to use, but if there are any color matching problems you should let me and Kinefinity (tm) know about those so everyone can try to work together to see what can be done to resolve any issues that come up with interoperability.

For those that are not clear, the KineRAW (tm) cameras can shoot True RAW sensor data into Bayer type DNG frame files, that captures 100% of the sensor data. Cameras like RED's (tm) that use compression ratios of 10:1 or more only capture maybe a fraction of the sensor data without errors at the 12th bit level.

Because the KineRAW (tm) cameras capture the actual sensor data, the DNG can be converted into just about any file format you may want, like, Prores (tm), H.264, Cineform (tm), uncompressed AVI, BMP frame set, DPX frame set, TIF frame set, JPEG2000 frame set, Motion JPEG, and just about anything you can name by use of one or more third party conversion programs.

Although the S8 is planed to record just DNG at full resolution at this point, I understand that there is a DNG to Cineform (tm) conversion program that could be used to make Cineform (tm) for you to edit with for that camera as well by adding one step to the workflow, such conversion would be fast since you are going from DNG True RAW to Cineform (tm) 'raw' without much image processing. As I understand there is also a reverse conversion so you can go from Cineform (tm) 'raw' to DNG 'raw', its not totally lossless all the time that way, but its wavelet compression like REDCODE (tm) 'raw' and so gives the benefit of being able to increase record times up to about 10x, so rather than 9 minutes on 60GB KineMAG-60 (tm) you might get 90minutes non-stop record time using Cineform (tm) record mode in the S16 or S35 cameras.

If you start with the 100% True RAW DNG recorded frames, then you avoid double compression losses going from them to a compressed file format for your (off-line) editing. Some people may find that the compressed formats meet their end use needs, but if you shoot DNG you can archive the original data to later do a high quality DI with later if your project sells and the distributor wants to get a higher quality DCP made. Much like shooting Super16 and having a 35mm blowup print made in a wet gate printer after you sell the rights, DNG of the Adobe (tm) CinemaDNG initiative are a true Digital NeGative format for cinematography use .

February 22, 2012

0
Reply

Dan, if you ever want to guest post about the camera and your tools here on NFS, you have an open invite.

February 22, 2012

0
Reply
avatar
Ryan Koo
Founder
Writer/Director

Thanks for the invite.

Right now the main DNG camera that might be of interest to filmmaking students seems to be the Acam dII (tm), but when the KineRAW (tm) cameras come out for sale and can get into the hands of filmmakers and students, then talking about minute details of the work-flow options may be of more interest to them, right now if someone wants to look into that aspect of the cameras they can maybe try out work-flows with the sample DNG frames on the Acam dII (tm) web site. I have some preset color corrections in my program DANCINEC.EXE (tm) in file DANCINEC.ZIP (tm) on my web site now, that can work with those sample frames.

One of the things I need to do is to develop additional presets for color correction in the KineRAW (tm) cameras, that way when you select a given EI/ISO and K value (and some other in camera options) you can get a first pass result that is a close match to what you saw on the camera's viewfinder. That works bt way of a special kind of color correction file authored in my programs that is loaded into the KineRAW (tm) camera using a USB flash drive memory stick, after those files are loaded they stay in the camera for future use and the camera remembers your last selection so when you turn it back on you are more or less ready to shoot again and see a corrected image on the monitor. The KineRAW (tm) cameras also have built in linear RAW views you can select to make sure you are not blowing out the highlights since you see the highlights expanded that way (everything else looks darker than normal so you see the highlights stand out with better contrast). In addition to all that you have Zebras and the waveforms can be set to sensor RAW or to processed view (like a video signal monitor). The Waveforms show the absolute sensor exposure level in sensor RAW mode, so you can know your exact exposure regardless of what the subject or lens being used is. With the Waveforms in processed mode (called Post) you see midtone at about 46% waveform no matter what the EI/ISO selected is, whereas in sensor RAW, the signal level in the waveforms ONLY depends on the sensor exposure, not the monitor LUT used to compensate for the EI/ISO selected, for instance, if you are at EI/ISO 100 and then you change to EI/ISO 200 and cut the shutter angle in half, on the sensor RAW waveforms the signal for 18% gray will drop to about half the value, but in processed 'post' view it will stay at about the same signal level. What the exact target signal level is for the 18% Gray card is depends on how the monitoring LUT are setup, but its about 46% for computer monitor display of the finished frames, and maybe 41% for Rec.709 end use.

As I get more experience with the camera's monitor calibrations, and people start using the cameras, then getting some feedback from camera users about how they are using the camera's monitoring features would be helpful and of interest.

The goal as I see it with these cameras is to make shooting True RAW not much different from using a camera that has built in color processing like a HDSLR through the use of various monitoring options, while at the same time being able to record and archive all of the sensor data, so that you have total control later in post production to ignore the monitoring LUT used while shooting if you want to, or to go with the original selections made at the time of shooting by way of the meta-data saved in each image frame etc.

The camera also make 3D-LUT for each shot made based on the in-camera monitoring options selected, so that you can have support for work-flows that do not yet support the in-frame meta-data.

That's an overview of what part of the camera's details I will be most familiar with, but if people have questions about issues relating to the use of the camera for filmmaking they can check out some of the behind the scenes photos that I'll try to include in the additional videos of the tests I am shooting. As has been said, a picture is worth a thousand words...

February 23, 2012

0
Reply

I welcome the addition of high quality cameras to the market. The idea that there is room for only one or two brands of high performance acquisition tools is nonsensical. Fanboyz beware, alternatives lurk.

February 17, 2012

0
Reply
Zan Shin

The camera does not surprise me. I went to China in 1981 with my Chinese wife and in Beijing we visited the first Chinese computer show open to the public and covered in Byte Magazine. The same day we drove in the countryside and saw a man walking a waterwheel to irrigate his crops and two girls singing and working ropes to lift water in a basket to accomplish the same thing. For us, that will be the day China began the transition to a technology country. China has already proven they can accomplish high quality technology. They just docked two craft in space and have shot down satellites.

What is lacking in China is the freedom to use this camera and create freely. There is a vast store of great literature just waiting to be put into feature film format. That and not a mere camera is the real serious threat that we ought to be concerned about ... and looking forward to. But I doubt it will be able to occur until the Chinese government and Communist Party learns to trust the creativity of the Chinese people. China's vice president who is slated to head the country in coming years is quite an interesting guy. He lived in Iowa studying corn farming and made many friends there. I wonder if he won't be the one to begin letting democracy happen? If he does, then watch out because the Chinese are capable of becoming awesome story tellers in film. They have over 5000 years of world class literature to draw upon. The camera sounds interesting but believe me it is not the threat to concern yourself with. If freedom ever begins to happen in China we are going to see a flood of films and they will have more quality than you probably imagine, because the Chinese have been expert story tellers for far longer than any civilization making films today.

February 17, 2012

0
Reply
ED

By freedom you mean cheap pornography of the western major motion pictures? Yeah, right, freedom. Don't you dare to change, China.

February 19, 2012

0
Reply
Natt

This is the project with the Russian Kinor 35mm movie camera
and the Chinese to developing a digital movie camera ?

Look like to Yes to me
I have a Konvas M2 film camera and a old ZenithE srl.
Yes i agree the best will be a camera for using the good super 16mm lens out there...
but a camera design for documentaries we have none at the present time.

February 17, 2012

0
Reply
Pierre Samuel Rioux

So far as I know, the Kinefinity.com (sm) camera were developed independently and are designed with modern components to make a state-of-the-art full bandwidth CinemaDNG and Cineform (tm) recording Digital Cinema Camera. They have been designing astro cameras and have experience with low noise camera design. They are interested in being contacted by people interested in their Digital Cinema cameras as this is planned to be a continuing development program, so feel free to email them with your thoughts about how the cameras would fit into your projects. The development of the KineRAW (tm) cameras has been going on for at least the last four years as far as I know.

February 21, 2012

0
Reply

If this camera gives me everything an ARRI does? For pennies? I'll take it. If I'm scraping my own pennies together to finance my first indie feature; and I'm using every trick in the guerilla movie making book...and this camera finally gets it going. I'll support a Chinese made product all the way.

February 17, 2012

0
Reply
MARK GEORGEFF

Given that most of the consumer electronics we use (including our lovely, hackable Pana GH2) come out of China, some of the above comments are plain stupid and parochial.

RED's Epic and Scarlet were initially going to be made in China by Foxconn until they had a falling out over manufacturing costs.

Maybe they would be more reliable if they had been...

February 17, 2012

0
Reply
skeptikal

I've been testing the prototype KineRAW-S8p (tm) while I shake down my software support for its CinemaDNG workflow. For those that don't know I have a freeish DI system on my web site that filmmakers can help beta-test, it has NLE/CC/MIX/DI/DIY SCAN and DIY FILM RECORDER, its a full uncompressed DI system that works higher than 4K so you can do a 'studio quality' DI if you want to at home, if you understand what places like Technicolor (tm) charge to do a DI for you, you will understand why that might be useful if you want to make feature films that show in movie theaters on a micro budget.

The KineRAW-S8p (tm) is of interest for me because the Panavision (tm) 2.35:1 format uses a larger area of the color print stock that will get exposed off a modified LCD monitor, so that retains more of the resolution the uncompressed frames can hold on the movie screen, depending on the lenses used and how well the projector is focused.

The camera shoots 2592x1104 for that format, as well as 1828x1556 (the standard 2K Panavision (tm) Cineon (tm) pixel resolution). If you have worked with 35mm movie film scans then you understand that movie film scans are not single pixel resolution at 100% MTF. In some ways the KineRAW-S8p (tm) can give results off a 35mm release print as projected that are quite close to having shoot 35mm movie film and then done a DIY film scan, except that you don't have to deal with the dust, muck, and scratches that movie film has.

Since movie film projection's days seem numbered, the results for DCP creation are a bit harder since that standard exceeds what is typical for 35mm release print projection (35mm film shows about 960 to 1280 horizontal lines at high MTF in typical projection a bit out of focus from a 4th generation release print, DCP for 2.39:1 aspect ratio is addressed at 2048x858 pixels).

Arri Alexa (tm) shoots 2880x1620 oversample for 1920x1080. The KineRAW-S8p (tm) shoots 2592x1104 so its horizontal pixel count is just 288 less than Alexa (tm). The S35 model will not have as much oversample, but the cinema lenses will give higher resolution in ratio to the image height so less sharpen should be required for the S35 footage.

There is no absolute 'best' digital cinema camera at this point, and one needs to think about how the final images will look on the movie theater screen. If you shoot with a HDSLR with 8bit recording you have no grading range without introducing histogram gaps. With 12bit True RAW data you have a wide grading range which makes it more like working with 35mm negatives in the CC of the frames used in the DI which would be at least 30bpp if not 48bpp. If you need speed, the Cineform (tm) option of the S35 gives you some grading range and smaller files and more recording time.

I hope to shoot some other subjects as I am able, I've been working with my Brother to shoot an anamorphic demo reel for the KineRAW-S8p (tm) I hope to have that done in a few weeks. Some people have contacted me about helping with the demos, and that would be good. I have to develop code to support new camera features as well as do some calibrations of the viewfinder, you have to remember that the camera shoots RAW, so someone needs to calibrate the color corrections in the camera's viewfinder so I'm helping with some of that for now, while shooting. I have needed to machine some parts for the camera as it came without accessories, and to make my own cables to access the external run button and power cable and battery so doing all those things takes some time.

After a few firmware updates, the camera seems mostly stable using the KineMAG-60GB (tm) SSD, I can put it into record and let it shoot the full magazine and it does not seem to have any glitches at all so far, and that is not an easy accomplishment for ANY company to record at over 107.3MB/s sustained.

The KineRAW-S8p (tm) does not have a fan, and so far as not overheated, although it is cool in San Francisco. It is totally silent, making it great for sync sound unlike the 35mm movie cameras I had been using. It has auto-slate beep and meta-data for the slate info, so you don't need to do a manual slate on each take. My software can read the meta-data slate info and burn that into the workprint frames along with date+TOD and SMPTE time code, as well as frame number per shot etc. and the head and tail slate markers.

My Brother has been doing a Off-line edit of the anamorphic test reel in Power Director 9 (tm), and then I have been conforming to that in my DI system for the result that will show up on Vimeo when done. In that way you can use any editing program even if it does not support 24fps content, so people who like FCP can do the off line using that, and then do the final audio mix in my DI system as well can conform the high resolution uncompressed frame based edit through my NLE. If you don't like some of my software you can use any part of it or not at all as there are many workflow options, but the end results may vary quite a bit depending on which one you pick.

The look of RAW cameras footage depends in large part on how you process it, so you should not judge the look based just on one workflow or filming style.

February 18, 2012

0
Reply

That free-ish software you mentioned got me interested, but wow. That site of yours is close to un-navigateable(!) I have no idea what I'm looking for and the page is filled with almost a nonsensical amount of texts that confuses me about what the site actually is supposed to be about. Is there some direct link to more info and screen-shots that I can use to get a feel for what your workflow is supposed to look like? Because, in the mess of links and random texts I cannot find anything.

A friendly tip: Most free blog-services (wordpress, blogger and the likes) are surprisingly easy to use and amazingly flexible for most any need. Freeing up time to do what you actually want to do. Whatever it is.

February 18, 2012

0
Reply

If anyone has a serious interest in using my DI system with any kind of camera, you can use 5DtoRGB (tm) with HDSLR to make DPX frames it seems, or VirtualDUB (tm) to make BMP frames from video files, they can email me directly and I will try to talk you through the menus. Making a feature film DI is a complex subject, and so that is the easiest way to get past the humps if you are serious about doing one. If you not serious then it does not matter what the web site looks like, its just a place to put files for download by people that would be serious about helping beta-test them. In a more perfect world I might be able to shoot videos, edit videos, code code, and find time to work on the web site, but right now the web site is at the bottom of the list as it does no good to have a useless program and a neat web site, I would rather have a messy web site and a program that does the job. Its free-ish so you get what you pay for.

February 18, 2012

0
Reply

Thanks very much for sharing, Dan!

February 18, 2012

0
Reply
avatar
Ryan Koo
Founder
Writer/Director

Thanks, I look at the KineRAW (tm) cameras for 35mm movie film replacement. The results may not be an exact match in all cases, but for me the balance is toward the True RAW DNG digital camera because of these reasons:

1) No dust, scratches and muck to remove from the film scans.

2) No need to scan the negative.

3) No film costs and no Lab fees, saving at least maybe $5000 to $125000 per project.

4) No need to manually spot the film scans that have dust missed by the dust busting software.

5) You can delete shots that are bad so you don't need to archive them making the storage cost even lower.

6) The camera is silent (other than the autoslate), unlike 35mm movie cameras, so you can shoot close to the actors.

7) You can grade shots up more than film scans maybe since there is no muck, scratches, and film grain to deal with. Its a pleasure to grade the 12bit sensor data as you can make up for lighting issues with curve adjustments in ways that trying with film would get very grainy probably.

8) With True RAW you can do the white balance and dark field calibration AFTER shooting so you can not limited to doing the dark field calibration just before shooting like in a compressed camera. That can save time and cover any pixels that went bad during shooting.

9) 35mm movie film is heavy to ship, and hard to take through airports without getting X-rayed. Using film makes shooting outside the US harder than doing things digital where you can FTP back the frames or ship the frames on light weight notebook HDD.

10) The KineRAW (tm) cameras have two slots and you can do shot selects and backups in the camera without having pack a notebook computer around. If you are shooting in a remote place, you just need some extra notebook HDD to dump shots onto and you can re-use the KineMAG-XX (tm) SSD over and over again. That makes the shooting maybe even more portable than shooting Super16 (tm) and you get 2.5K or 2K or HD etc. without the cost of telecine (tm) and other lab fees and film costs, as well as less weight and not having to use a changing bag and worry about dust in the gate and have to do a gate check after each shot etc.

February 19, 2012

0
Reply

what's with the weird movement of the crumbs at 00:38-40 on test reel B1?

February 18, 2012

0
Reply
ben

At 24fps with a large shutter angle the motion blur is larger than what you would see in a 60p camcorder with a small shutter angle. The noise reduction tends to make the motion blur have less 'see through' that it would with the grain left in so it looks maybe somewhat featureless. I applied noise reduction to the bird shot, because of the hard grade applied. Those things can adjustable and can be turned off in my de-Bayer program, if you don't like the effects you can always use some other software to de-Bayer the frames or change the settings used. Many programs support DNG frames, its a universal open Digital NeGative format for storing RAW sensor data without Compression. Visit the Adobe (tm) web site for more information about their CinemaDNG (tm) project to promote DNG as a movie capture file format for frame based movie capture.

February 18, 2012

0
Reply

How many of the comments had anything to do with the camera in question, and how many were just ranting about stereotypes and super egos? It's a shame to have to weed through the crap to find the topic...

February 18, 2012

0
Reply

You're off topic...but so am I

February 18, 2012

0
Reply
bguest

Interesting reason. I love to see clearly Martha

February 19, 2012

0
Reply

Hey Everyone,

Thanks to your votes 'Parasite' made it into the final round on YobiTV!

The final round runs from Febuary the 20th-26th and I'll be competing against 15 other finalists for first place.

The 3 winners will be flown to Detroit, USA for spots on the new YOBI.tv original reality webseries “Aspire”.

Keep up the voting and don't forget to share this link on your Facebook wall, as well as with your friends, family and co workers.

You can vote once everyday for my entry with a YobiTV account which can be created in just minutes.

Don't forget to comment on the video as well and like it on Facebook or share it on twitter.

Thanks again to everyone who has voted and given thier support!

Cheers.

Dirk.

http://yobi.tv/yobifilm/​contestant/17928/dirknienaber/​parasite

February 21, 2012

0
Reply

It's past mid February. Where's the footage at???

February 28, 2012

0
Reply
Nick

Patience is a virtue...

February 29, 2012

0
Reply
sammy

Oh, is it?

March 1, 2012

0
Reply
Bob

I'm working on the anamorphic demo real now, my Brother wanted to make a kind of story film with sets and sync sound showing the type of thing one might make with the anamorphic lens, its about 95% done, I've been interrupted by processing some tests of the S35 for Kinefinity.com (tm) this week they wanted some monitoring look up tables made at ISO 400 and 800 for testing their two revised prototypes in China.

There was some back and forth on the skin tones but the last set of monitoring tables should work for Asian skin tones, at least in my mind, for some reason they want the skin tones quite green and its not something I think looks right so I asked them to turn the florescent lights off in their grading room as one's eyes can become accommodated to the green cast of florescent lights and that makes neutral gray seem pink .

Anyway, on my high quality CRT monitor, and by measurement, their reference image shot with a Canon DSLR is greenish and overexposed. So if they make some sample footage that looks greenish and over exposed its not the camera so much as their taste in grading, for myself the 'Matrix' look is something you add after shooting is done not in the camera unless its a known effect shot.

So if you wanted a reason for the delays that is one of many. By I have gotten photos of their calibration setup with the new S35 camera part way put together (for some reason it side panel was open). Shooting test images.

The noise level at what they think is ISO 400 look good, and at ISO 800 there is a bit in the shadows without de-noise, but as I have developed effective de-noise for the S8, applying that should work on the S35 as tests so far seem to indicate. If you have worked with 35mm negative scans at ISO 800, you would be surprised by how good the KineRAW-S35 (tm) looks.

I've developed two monitoring tables that give a nice long tonal range, and the camera converts those into 3D-LUT for use with Iridas Speed Grade (tm) and other compatible work flows. I have not seen the Cineform (tm) results yet using those LUT templates, but the conversions are similar so the colors should be a close match.

The main issue they seem to be having is with the zoom lens they have been using, I guess they have the Canon mount on the S35 for the test DNG they have been sending me this week. They have a quick change mount from Canon mount to PL and both parts have been fabricated at the factory.

As far as I can tell things are moving forward, but you need to email them and tell them if you are interested so they can know about how many to have made up for sale as that impacts the volume pricing they can negotiate with the factory.

The basic design of the cameras is working better than I could have hoped for 4 years ago, and is quite advanced. They just added anamorphic un-squeeze options in the viewfinder making using various ratio, 1.25x, 1.33x. 1.5x, 1.75x and 2.0x anamorphic lenses give a correct image on the viewfinder with various resolutions like 2.0x for 1828x1556, and 1.75x for 1920x1440 (for Panavision (tm) CinemaScope (tm) or 2.39:1 DCP), and 1.33x for 1920x1080 (for 16:9 and 1.85:1 DCP).

March 3, 2012

0
Reply

Vaporware

March 1, 2012

0
Reply
Chet

You can write to Kinefinity.com (tm) and maybe they can send you a JPG of the second KineRAW-S35 (tm) shooting the calibration reference DNG frames, they sent me a photo of their setup, they have a black cloth background and two small CF lights with barn doors.

They are getting enough light to shoot at f/4 for the ISO 400 and 800 tests, and the skin tones look usable, even though I'm not a fan of vapor lighting.

I wish the KineRAW-S8 (tm) was in production now, the prototype is working well enough to be used now it seems. There are a few minor issues requiring a re-boot sometimes, but I was shooting for about 5 hours the other night and it worked flawlessly during that shooting. The footage all seems fine and 100% glitch free, we shot about 130GB of DNG frames in the 1828x1556x12bit@24.000 fps mode.

You can write to Kinefinity.com (tm) and express your interest in making a purchase. From what I understand they want to get the KineRAW-S35 (tm) on sale in China first, then get back to the S16 and S8 models completion since that would give them some revenue to use for the additional prototypes as there is a higher cost making one-of camera prototypes by hand at the factory.

I for one hope they are able to bring the cameras to market, which is a more complex process than making the camera to the point of just working. If you have an interest in making a purchase, you should write to them and say so, as the number of potential buyers is an important variable in the whole process. There may be some willing to pay more to get the useful DNG workflow for film like results, but for my interests I would like to see the MSRP lower and within reach of people that would be interested in using my freeish DI system to make micro-budget feature films.

I was at a showing of the original 'Planet of the Apes' last night and the theater had a good turn out for a totally non-CG movie, and I also went to see the 3D converted Star Wars movie that was mostly CGed and the theater was mostly empty. We all know the the silent black and white non-CG movie got 'best picture'. So there may still be a market for real movies with real actors in them, made without CG using a camera to capture acting performances. The KineRAW (tm) cameras if the can come to market, may give a more film like freedom of capturing photons from the surface of the actors and converting that into graded 35mm film prints for projection in movie theaters (or DCP) that other cameras that process the sensor data in real time in the camera for fast but limited results. DNG is a Digital NeGative, the most 35mm move film like option in a industry standard file format.

March 3, 2012

0
Reply

I'm sure the Apertus project will come up with an affordable RAW solution sooner than the Kineraw. Hopefully their planned Kickstarter Campaign will be successful so they can implement the new sensor front-end...

March 7, 2012

0
Reply
Chet

Mybe they will...once they get a kickstarter going..

But first they need some decent footage cause
no-one will contribute based on what they shot so far.

March 7, 2012

0
Reply
sammy

The sensor they're interested in isn't even made yet.

March 7, 2012

0
Reply
Chet

Will Apertus still be recording 8bit JP4 compressed data?

What is the end use for a camera that is limited to JPG derived so called 'raw' data that is only 8 bits, you cannot hard grade 8bit data without introducing histogram gaps, and what about the block artifacts from the JP4?

Another issue is that if you record 8 bit data RAW liner, and the do normal gamma correction you only end out with about the top 6 bits being without histogram gaps just from that. If you record using the LUT in the camera then its not liner RAW data, and making that into a DNG is just not right since de-Bayer programs read linear 12 to 16 bit data not curve adjusted 8 bit data.

What is the bandwidth of the Apertus? What is the compression ratio, 100:1?

How are Apertus going to record 4K data through their JP4 bottleneck?

KineRAW (tm) cameras are the real deal, they record full bandwidth CinemaDNG in the camera, not some odd ball 8bit JP4 system that is not supported by industry standard editing software like Iridas Speed Grade. For compressed they support Industry Standard Cineform (tm) that works with many programs including GoPro (tm) software.

I spoke to them some time ago about adding a True RAW port to their IP surveillance camera that was the basis of the circuit they were talking about, but if you read on their forum you may see what their response was, in particular the insistence that no more than 8bit data is needed. RED (ONE (tm) used 12bit data for good reason, to argue otherwise just reflects on their position in the industry.

March 10, 2012

0
Reply

The KineRAW-S35 (tm) cameras are shooting now, they sent me some calibration DNG a few weeks ago and I made some monitoring LUT for their two test cameras to use in shooting, so who has what finished first?

March 10, 2012

0
Reply

Kineraw is unresponsive. I've emailed them for samples and release dates but have not answered me. Kineraw cannot even be purchased and has been in "testing" for quite sometime now, so I wouldn't consider that "finished".

As for the new sensor, the Apertus Project is going to use (once their Kickstarter Campaign is succesful) the CMV12000.

The specs are for the CMV12000 are:
SIZE: 12M (4096 x 3072)
MAX FPS:300
DIAGONAL (MM): APS-C
TECHNOLOGY: CMOS
MANUFACTURER: CMOSIS
PRICE: ~1000€
COMMENTS: 8,10 or 12bit output, Global Shutter

But if you insist the Kineraw is finished, then give me a link to where I can purchase it/see footage/see DNGs etc etc etc.

March 10, 2012

0
Reply
Chet

Thanks for the information Chet,

Quote: [Kineraw is unresponsive. I’ve emailed them for samples and release dates but have not answered me. Kineraw cannot even be purchased and has been in “testing” for quite sometime now, so I wouldn’t consider that “finished”.]

I was told today that field testing should begin in April if things go well, as I mentioned I have been doing some monitoring LUT for them, and yesterday they send a bunch of DNG frames to have checked for exposure levels, I can make some measurements in my programs so I was doing those and sent back the measurements, they were checking so exposure settings for consistency.

They may not be sending out samples yet, because they want to wait for some good looking shots, not just a gray card and color checker. If you have a specific question you want me to pass on you can email me at tempnulbox (at) yahoo (dot) com and tell me to pass on your question and email address to them directly, as I would need your permission to give you email address out to them for a reply, put 'Chet from nofilmschool' in your reply.

Quote: [As for the new sensor, the Apertus Project is going to use (once their Kickstarter Campaign is succesful) the CMV12000.]

As that part is going to be retail, there should be no reason that other camera makers cannot also base cameras on it?

Quote: [The specs are for the CMV12000 are:
SIZE: 12M (4096 x 3072)
MAX FPS:300
DIAGONAL (MM): APS-C
TECHNOLOGY: CMOS
MANUFACTURER: CMOSIS
PRICE: ~1000€
COMMENTS: 8,10 or 12bit output, Global Shutter]

Well if they can retail such a camera with Uncompressed True RAW CinemaDNG output in the US that would be great, but SSD capable of recording 4K are not easy to find maybe?

4096x3072x1.5x24.000 = 453MB/s

So if they are going to RAID 4x SSD like Aaton (tm) does, it would indeed be impressive if they could do such a camera for MSRP of USD $1310?

Is that what you are saying, they it will record CinemaDNG full bandwidth, or will it record JP4 type JPEG code at compression ratios of 100:1 or more for 4K use?

Where are their sample DNG that I can download to process in my free de-Bayer program now? Have they even shot a single frame?

If they are going to make a True RAW uncompressed recording CinemaDNG camera then I would be interested in it for the price suggested, if its a JPEG derived compressed camera, then will it have the block artifacts issues or how have the removed those while still getting 100:1 or more compression ratios?

Quote: [But if you insist the Kineraw is finished, then give me a link to where I can purchase it/see footage/see DNGs etc etc etc.]

You know the S8p (tm) sample footage in on my Vimeo page now, and I can send you some BMP frames from those, I'm not sending DNG out yet because the ones I have don't have the color matching tags adjusted well enough yet, as my de-Bayer program does not use those it was not a priority for me. I am told that sample S35 footage will be shot before too long, there was a sample video up in China of the first prototype S35 made some time ago that they did before, I put a link to that up on DVXuser for some time but deleted it since they were making improvements to the camera in the re-design. Kinefinity (tm) may still have that video up on the Chinese version of youtube, it was 1280x720 because that is their limit for that site. I thought the grading was a bit off balance, but they did not ask for my input when they shot that S35 demo. The 3D-LUT work has progressed since then.

You can see there is an image in the viewfinder on the trade show video Dan Chung made,

http://www.dslrnewsshooter.com/2011/08/27/birtv-2011-chinas-answer-to-re...

I hope that Apertus makes something that records without JP4 or other forms of lossy compression, as that would be an advance over cameras like REDONE (tm) and HDSLR (tm) but if it is not such an advance, then how it compares with other lossy compressed cameras will depend on how the compression deals with the noise levels in the sensor they have selected and such.

March 12, 2012

0
Reply

What fps were those DNG shot with?

And why are they not 2592x1940 rather than 2592x1936 pixels?

As you know the 8bit JP4 compression used by Apertus (tm) would require higher compression rates for higher resolutions and frame rates, so what is the compression ratio for 4K at 300fps?

Do you have any sample DNG from the Apertus (tm) demo video on YouTube?

What was the data rate in MBytes/second while shooting the YouTube demos?

March 14, 2012

0
Reply

I just grabbed them off their site where they are readily available. If you want more details you can go over to apertus.org and hop in their forums.

March 15, 2012

0
Reply
Chet

As far as I know that Elphel sensor does not shoot 2592x1936 at above about 15fps so maybe its not a good test of the compression losses for shooting 24fps at 2592x1150. Also the lens used does not seem very well corrected or something, so its not saying much about what block artifacts may be visible in actual movie shooting as those sample frames could have been made at 1fps or something slow to reduce the compression ratio of the Apertus 8bit JP4 of JP46 compression based on JPEG compression methods it seems.

The whole issue of actual bandwidth to the SSD stuck on the bottom of the Elphel camera seems a bit vague as they don't seem to make clear what the compression ratio is or that they are only recording 8bit data. With so called 'raw' data not being white balanced, even with gamma correction in the camera you don't have enough bits to clip the data after recording and not get histogram gaps.

The JPEG quality factor is not a direct relationship to compression artifacts, I know that even at 100% 'quality' JPEG can show artifacts on synthetic images in some programs I have tried, its not a true faultless image like uncompressed BMP or TIF.

Also digital cinema cameras need to have the sensor underexposed about two stops in order to hold the highlights in a more film like way, if you only start with 8 bits and record linear raw data the blue pixels may only have 5 bits above the 90% white clip point, and at just 5 good bits banding and histogram gaps can show up as an issue.

There are reasons that REDCODE (tm) and Cineform (tm) use 12bits or more and wavelet compression, and those same reasons for using wavelet compression in digital movie theaters showing DCP.

Using JPEG based compression at 8bits for 'raw' sensor data, is unique to the Elphel camera as far as I know and one wonders if the people on the Apertus forum understand what exactly those issue mean for the ultimate image quality of their project? There response to my inquiries about adding a True RAW port to the Elphel would seem to show that there was some understanding by those who said that the Apertus project was not solely linked to the Elphel camera, as far as I can make out, but what ended up it seems is the opposite of that trend.

If the LUT in the camera does white balance, gamma correction, and ISO S-curve, then what exactly is 'raw' in the recorded frames?

March 16, 2012

0
Reply

Looks like the Kineraw not only has to look out for the Apertus Project, but the Digital Bolex as well....

March 14, 2012

0
Reply
Chet

At least Digital Bolex seems to be a True RAW uncompressed recording camera design, their demos seem to have been shot using this stock industrial camera type,

http://www.alliedvisiontec.com/emea/products/cameras/gigabit-ethernet/pr...

You can see it in a photo at:

http://www.eoshd.com/comments/index.php?topic=399.15

Look at the image on both pages in the comments.

March 14, 2012

0
Reply

Yes I saw these this morning. I only wish the best for them, and praise them for taking initiative.

March 15, 2012

0
Reply
Chet

Camera rental house like us are caught in surprise always how technology changes with lower pricing it offers. Really look forward to test this camera against the branded ones.

March 17, 2012

0
Reply

March 19, 2012

0
Reply
Chet

The main problem with such cameras for filmmaking is that you need a very fast computer to try to get glitch free recording, I have someone contact me about a Sumix (tm) camera he got used, he said that it uses the same 2K sensor as the SI-2K (tm) if I recall right, but we was unable to get it to record well about 960x480 while his goal was 1920x1080, so maybe it works but how big of a computer do you need.

I contacted several camera makers years ago asking about if they could make a USB camera with a built in buffer so the camera could shoot 15 seconds or more to the buffer, and then dump at whatever speed the cable would allow, in that case the computer speed would have NO impact at all on the resolution the camera could shoot at and the system would be totally free of glitches in the recording even if you use just a net book with an external harddrive to store the shots onto. Such as camera can be made for about $250 including the sensor, OLPF_IRcut, DRAM for the buffer, FPGA, and case with C-mount for 2.5K resolution. If the buffer is large enough to shoot one to four minutes, then you can have the camera download the data while it is idle so that the buffer would hardly ever fill up, and you can also shoot with the cable un-plugged just to the buffer making it better for filmmaking than a camera that cannot be used unplugged from a computer.

The limited dynamic range of the Kodak chips is an issue for image quality, as is the vertical streak issue of CCD sensors. CMOS do not have the vertical streak issue, and the sensor in the KineRAW-S8p (tm) is rated for more db dynamic range than the Kodak chip if I understand the spec sheets right.

March 21, 2012

0
Reply

The way you find the bits of useful data is to divide the rated dynamic range by 6 because when the sensor is at a gain setting of 1 bit for each T/ stop change in brightness you get 6db per T/ stop so,

64/6 = 10.6 bits

70/6 = 11.6 bits

So a sensor with 70db dynamic range gives about 1 stop more range than one with 64db, all things being equal. Since the Kodak sensor is multi-readout there can be some half or quadrant seam artifacts showing, and careful FPN masking is needed, and even with masking, the gain matching needs to be very close as well.

The sensor in the KineRAW-S8p (tm) has just one readout so does not have "splits" issues or seams in the image even without FPN masking being active.

March 21, 2012

0
Reply

What is the MSRP of the Tauri-HD 02150 SDI ?

And then what is the cost of a computer fast enough to record to without glitches, and how do you power your computer when shooting outdoors, as well as how to you pack everything around?

March 21, 2012

0
Reply

I posted the third test reel for the KineRAW-S8p (tm) last night.

It stutters a bit from the Vimeo playback on my Brother's computer, and there are some spots left behind in the scrolling titles from the compression it seems, but you get the general idea of what shooting at f/0.95 with a 2:1 squeeze lens gets you,

http://vimeo.com/38907938

March 21, 2012

0
Reply

Work on the KineRAW-S35 (tm) prototype "B" has been moving forward, one of the units in in field testing in China now and details are being tuned to ready the design for planed sales in China to start before too much longer, perhaps.

May 19, 2012

0
Reply

I'm also working on two new videos from the KineRAW-S8p (tm) prototype, testing lenses on some other of its many resolution modes. There is a fourth hand held video up at my vimeo page,

http://vimeo.com/40921094

May 19, 2012

0
Reply

Sample MOV and Cineform (tm) along with matching 3D-LUT are free to download at links in this posting, as are photos of the KineRAW-S35 (tm) second prototype in field testing in China now,

http://www.dvxuser.com/V6/showthread.php?260134-Chinese-2K-RAW-S35-Cinem...

Quote: [There are 720p H.264 mov for downloading:
http://115.com/file/e7jl12ad#final-422hq-QuickTime-H.264.rar

There are some Cineform RAW footages for downloading:
Clip1: http://115.com/file/c2exeutq#clip1.rar
Clip2: http://115.com/file/bemsmxgt#clip2.rar
Clip3: http://115.com/file/dpk7t4vt#clip3.rar

The Cineform (tm) player is here,

http://cineform.com/gopro-cineform-decoder]

Quote: [Here are links for Low-Contrast Rec709 LUT.

1. daylight-160.look: http://115.com/file/c2elxws7#daylight-160.look
2. daylight-160.cube: http://115.com/file/dpkn5su4#daylight-160.cube

Speedgrade NX, Firstlight can use .look file, while Da Vinci Resolve uses .cube file. People can use the LUT as monitoring LUT, but also use it as a starting point of grading .]

May 30, 2012

0
Reply

http://vimeo.com/44417714

http://vimeo.com/44424450

--------------------------

I have posted two new KineRAW-S8p ™ videos on Vimeo, if you would like to see the image quality I have some selected frames I can send, you can email me at tempnulbox [at] yahoo [dot] com with "nofilmschool user wants frames" in the subject line.

The E1 video is shot 1920x1440 for 4:3 black and white cinematography and processed for maximum monochrome dynamic range, you should request the sample frames to see what the uncompressed detail looks like.

The F1 video is shot at various over-sample modes for 4:3, 16:9, 2:1, and 2.35:1 aspect ratios. The shooting resolutions are 1920x1440, 2228x1252, 2400x1200 and 2592x1104.

The videos were de-Bayered in my free de-Bayer program v0.06 and v0.07 (v0.07 is not posted on my web site yet as I have some things to finish up yet) and the DI was done in my "freeish" DI software.

I have a KineRAW-S35 ™ here for testing and calibration now and hope to post some videos shot with it if anyone has in interest in S35 sensor cameras. You can request some Cineform ™ sample clips from Kinefinity.com (sm) by writing directly to the email on their web site, be sure to also request the 3D-LUT that go with each clip so you can see the clips playback in Quicktime ™ or other programs with the color corrected.

July 6, 2012

0
Reply

Kinefinity (tm) has posted some more Cineform (tm) clips on their web site for people to download.

For some reason when I try to post those links here the comment does not show up, but you can find the links on DVXuser or you can email me or them and get the links that way. Be sure to also download the Cineform (tm) codec and the 3D-LUT that go with those clips for color correction when you play them in Quicktime (tm).

July 23, 2012

0
Reply

Well that post worked, I'll try line link at a time,

http://download.kinefinity.com/download/PRJ-0001-011-A1.mov

July 23, 2012

0
Reply

OK, more than one did not work, so here is another one,

http://download.kinefinity.com/download/ALVIN-0001-099-A1.mov

July 23, 2012

0
Reply

July 23, 2012

0
Reply

July 23, 2012

0
Reply

July 23, 2012

0
Reply

July 23, 2012

0
Reply

Seventh CIneform (tm) clip shot in the KineRAW-S35 (tm) sample camera,

http://download.kinefinity.com/download/PARK-0001-071-A1.mov

July 23, 2012

0
Reply

If you email me I can send you some sample frames and the other links.

July 23, 2012

0
Reply

July 23, 2012

0
Reply

Here is the link to one of the 3D-LUT,

http://download.kinefinity.com/download/XD48K04M.cube

July 23, 2012

0
Reply

Here is the link to the other 3D-LUT,

http://download.kinefinity.com/download/XD48K04M.look

The 3D-LUT are mated to the ISO and K value used when shooting so need to be installed for the clips above to play with the right "one light" color correction, the look group used is intended for having the shots graded, so graded results would look better. Various monitoring LUT can be loaded into the camera and selected with the encoder knob from a quick button menu labeled "look", and depending on which "look" is selected the camera makes a corresponding 3D-LUT in each shot folder, in this case the same "look" was used for all my test shots so far, that are linked to above for those MOV files.

You will also need to drag and drop the 3D-LUTs onto the Cineform (tm) LUT tool in the Cineform (tm) program folder.

July 23, 2012

0
Reply

Here is the link to the Cineform (tm) coded for MAC and PC,

http://cineform.com/gopro-cineform-decoder

Quote: [Cineform Firstlight/Davinci Resolve/Speedgrade/Scratch could be used
for grading/rendering. May use the Cineform decoder
settings-->Demosaic Type-->advanced details to make image look
sharper.]

July 23, 2012

0
Reply

My email is tempnulbox [at] yahoo [dot] com and put "nofilmschool reader wants samples or links" in the email subject line, you can also email Kinefinity.com (tm) directly to get samples and links from them.

July 23, 2012

0
Reply

Get your priceline worked out...

Just got word about the booth number where history will be made:

BIRTV2012, Aug 22nd~25th, KineRAW-S35 (tm) is at booth number is 2B222

August 3, 2012

0
Reply

Kinefinity.com (sm) have booth 2B222 at BIRTV 2012,

http://www.birtv.com/en/Eindex.aspx/

Quote: [The Must-Attend Industry Exhibition in Asia

BIRTV is China's most prestigious exhibition in the industry of radio,film and TV and a key part of China International Radio Film & Television Exposition.It is also the only one of such exhibitions which gains support from China government and is listed number one among the supported exhibitions in China's 12th Five Year's Developing Plan of Culture.]

They are also revising their web site now, the new web site will have sample images and footage you can download for compatibility testing with your own work-flow.

Here is a map showing the booth location,

http://www.birtv.com/upfiles/201208/2B0801.pdf

They will maybe have two KineRAW-S35 (tm) bodies there to show and be able to answer questions in person. If you are interested in getting one they will be there to talk about that as well. The price for the body may be about USD $6000, they are still working out the details.

August 11, 2012

0
Reply

Got word from Kinefinity.com (sm) that they will be going to NAB 2013 in Las Vegas (in April),

http://www.nabshow.com/

Quote: [Our booth number in this coming NAB is C12813.]

So if you have wanted some 'hands on' time with their cameras, this will be the first time they are showing them outside China as far as I know.

I've been shooting some additional test footage with both the KineRAW-S35 (tm) and the prototype KineRAW-S8p (tm) and may post that on my Vimeo page when I get it edited together. They are working on a demo reel for showing at NAB 2013, so if you have any suggestions of shots you would like to see you can email them and ask them to have one of their field testers shoot something like what you would like to see maybe so that when you go to NAB 2013 you can see how it turned out. They said they will have cameras there you can look over, and maybe the KineRAW-S35-mini (tm) they are still working on, its a smaller version with the same high quality 100% true RAW DNG recording and S35 sensor with sub-PL mount for using a wide range of movie and still lenses, as I understand it, you can email them for detailed specs if you are interested.

Also the MSRP and shipping should be lower for the KineRAW-S35-mini (tm). So jot down booth C12813 if your going to NAB this year and say high, they have been developing these cameras for many years, and the larger KineRAW-S35 (tm) has been in field testing in China (in addition to the one I got to do calibrations in the monitoring with). They said they are very busy putting cameras together to ship out, so, so far things seem to be going well for them.

February 1, 2013

0
Reply

That insight's just what I've been looking for. Thanks!

April 8, 2014

0
Reply

Pages