KineRAW-S8

[Chris Law]

Can anyone fill me in on some information about this camcorder. I want to be able to gain additional control's over colour settings when correcting in assimilate scratch. So that camera would be ideal for me.

[Sareesh Sudhakaran]

I'm not sure the camera is even in the market yet. Other than the info on the website and some PR, I haven't found a single review yet, of any of their models.

[Dan Hudgins]

I made some new photos of the KineRAW-S8p (tm) with different shooting configurations, with anamorphic lens and matte box, with follow focus gear on 12.5mm f/1.4 Fujinon 5MP lens and matte box with Mitchell follow focus, and using a 9.5-57mm f/1.6 reflex zoom that has optical view finder.

The color on the viewfinder looks better than how it shows up in the photos, its hard to take a photo of the LCD monitor and get the light on the subject the same contrast ratio and color balance, but to the eye the monitor image show more saturation and tonal range.

About when you can purchase one, email Kinefinity.com (sm) and tell them you have a serious interest unless they change the price etc. The more people that tell them they are interested the more they know about the interest.

The main question with the S8 is should they put the current design into limited production, or should they have more delays to make improvements?

The core of the camera has been working for more than a year, and they fixed some issues through firmware revisions, so its usable now I think for real filmmaking, a few minor changes would be good but it could be produced as is so far for people that are going to use it as intended. The proposed revisions would widen the market, but also maybe delay things for some time and possible inch the price up a bit?

I was all for producing the PREVIOUS version that had a sheet metal case, in place of the nice machined one this model has, so I think they could do both but they have their own ideas...

They are finishing up the KineRAW-S35 (tm) second prototype now, I have gotten some calibration DNG back and the ISO and tonal range and color look good, there are some issues with the OLPF selection.

Anyway, here are some photos, I hope to start shooting some hand held shots with the KineRAW-S8p (tm) soon, I purchased a 3.5mm f/2.4 Kowa megapixel HR lens, on the videwfinder it looks usable, I will have to shoot some tests because at 1920 wide on Vimeo its a challenge to be flawless, watch out for other camera makers only showing 1280x720 videos as there is a big leap from 1280 wide to 1920 wide in terms of de-Bayer artifacts showing up. I have some work to do on my up scaling features in my de-Bayer program, but if you stand back a 'safe distance' or transfer to 35mm print, they aren't as noticeable as putting ones noise on the 2K computer monitor etc.

Videos shot with it are here:

KineRAW-S8p (tm) lens test reel #A2 w/o links on Vimeo

KineRAW-S8p (tm) 2.5K lens test reel B1 on Vimeo

KineRAW-S8p (tm) anamorphic 2K (1828x1556) lens test reel C1 on Vimeo

you can email me at tempnulbox (at) yahoo (dot) com if you want some uncompressed BMP frames from those videos. If you are interested in the camera email Kinefinity.com (sm) and talk to them about your interests.

you need to look at this thread to see the photos,

http://www.dvinfo.net/forum/digital-...-8000-a-4.html

since the system will not let me link to them on this thread.

[Adrian Frearson]

Thanks for the updates Dan. I do like the image this produces, which is in the ball park of a very nice S8 transfer, superficially at least. I wonder who the market for this camera will be though? If the recent frenzy of the Digital Bolex is anything to go by, then maybe with a more retro designed body, it might do quite well. If the price can be kept around the 2 grand mark, then it might make it a decent alternative to shooting S8/DS8, or even 16mm for some people.
Right now, with a bit of shopping around, you could buy a nice used camera for 2k and enough stock and processing to shoot for a while. And, you'd have the fun ( and downsides ) of shooting on film. Don't mean to sound negative, just putting my thoughts down.

If film production completely stopped in a years time, there could be a few people looking for something like this, myself included. So unless there going to go for that retro/must have accessory market, then I don't think they should rush it.

[Sareesh Sudhakaran]

Originally Posted by Dan Hudgins

The main question with the S8 is should they put the current design into limited production, or should they have more delays to make improvements?

Thanks for sharing the info and videos.

They should definitely go into limited production. Who knows, the feedback might be critical and more accurate as to what the market really wants. Reliability is the key here, and unfortunately the website does not inspire much confidence.

[Dan Hudgins]

They are finishing up the KineRAW-S35 (tm) for field testing now and say they will get back to the S8 production model design when they have some time after this is done. I've been helping out with the S35 calibrations making some LUT to use in the camera for monitoring at various ISO and K values etc.

This week with the KineRAW-S8p (tm) prototype I've been shooting some hand held tests to make a video to post on my Vimeo page, testing the Kowa 3.5mm f/2.4 HR megapixel lens.

Feel free to email Kinefinity.com (tm) and let them know your thoughts if you are interested in their camera development. It helps them gauge the interest in the various models and features etc.

[Dan Hudgins]

Originally Posted by Chris Law

Can anyone fill me in on some information about this camcorder. I want to be able to gain additional control's over colour settings when correcting in assimilate scratch. So that camera would be ideal for me.

As I understand it Scratch (tm) now supports CinemaDNG import, although I don't know what the quality might look like.

Its possible to use my free de-Bayer program, DANCINEC.EXE (tm) to make 48bpp TIF or DPX (48bpp or 30bpp) and import those into Scratch (tm) just as you would film scans.

You can see samples color corrected in my freeish/free DI (Digital Intermediate) and de-Bayer software on my Vimeo page.

There are losses in the compression to post the videos on Vimeo, the uncompressed frames show some additional detail that makes the KineRAW-S8p (tm) footage closer to 35mm Techniscope (tm) in detail, depending on the lens used and such.

[Dan Hudgins]

Originally Posted by Adrian Frearson

Right now, with a bit of shopping around, you could buy a nice used camera for 2k and enough stock and processing to shoot for a while. And, you'd have the fun ( and downsides ) of shooting on film. Don't mean to sound negative, just putting my thoughts down.

The cost of processing and scanning just a few minutes of finished 35mm film on a normal high shooting ratio can come to the full cost of the KineRAW-S8 (tm) if it gets on the market.

I was shooting color print stock since you can get it with B&H 1870 perfs, they don't seem to be an issue using it in a camera ment for B&H 1866 perfs at 24fps, the cost is low and the grain is fine, and if you scan the color print stock you can get a sort of HDR black and white image out of it because the red layer is slower than the blue layer by a few stops.

But shooting 35mm film means very heavy cameras and lots of camera noise. The KineRAW-S8p (tm) as not fan and make zero camera noise when the notebook hard drive in not in the second slot for dumping shots to. With just the SSD in it you can get close to the actors and use the sync sound, unlike any 35mm movie camera you are going to find in the same price range. Also on a low budget film you want to be able to shoot on high shooting ratios of 20:1 or higher. You can't do that on 35mm film even shooting short ends because the lab fees will push your budget way over.

Rather than give all your budget to the lab, its better to use that money on the actors, costumes, sets, and locations as well as catering, as its what in on the movie screen that counts in the end to the audience, they don't care what kind of camera was used as long as the images look film like in some way.

If you want to go the film way, I have free programs on my web site for 35mm (or S16,R16,S8,R8 vista-vision or 65mm etc.) scanning DIY and for making a 35mm film recorder to go direct to 35mm print stock, as well as my freeish DI system to do grading, editing, and sound mixing, to get to the output DPX frame set for making a DCP or to convert to a 35mm film print etc.

[Steve Mullen]

Originally Posted by Adrian Frearson

I do like the image this produces, which is in the ball park of a very nice S8 transfer, superficially at least.

I can play Dan's tests on my 63" HDTV using the latest AppleTV. They look very good with what I consider THE most important feature -- the ability to capture very very bright white WITH details. Of course, the BM camera is claims to do the same and one expects the DigitalBolex to do the same.

I'll be getting the 4K JVC and the BM for review. But, these developments will play-out over the next 6 months -- at least.

So I've bought a Bolex P1 8mm camera. I'm going to use these months to experiment with shooting Tri-X B&W -- as I used to do, because I love the grain.

If Chris has a place, I'll blog my experiences. My main focus is on HOW to get the 24fps film transferred to ProRes 24p -- at a reasonable price.

If this topic interests folks, please post.

[Steve Mullen]

Since there are several cameras arriving with 2K and 2.5K rather than either 4K or QuadHD -- I wondered what the image quality might be of these "lower than 4K" cameras.

Since most lower resolution content will become FullHD -- unlike 4K which often is converted to film or 4K DigitalCinema -- the question is how the higher is converted to the lower resolution. I've seen references to "oversampling." While oversampling sounds great, if we change the word to "filter" or "downscale" we see why a benefit might not be obtained.

If we a use another term -- "interpolation" -- the question is the complexity (and hence speed) used to convert X columns and Y rows to X' column and Y' rows. As I understand it, downscales that: are not even multiples (like 2 or 4) or are very small (2.5K to 2K) are problematic.

To learn more, I read BBCs Alan Roberts (Index of /WHP034/pdf) test of a RED. (If you don't like math, skim his reports.) He says this about 2K (2048x1152) mode:

'The performance in 2K mode is little different than from any consumer camcorder that has a 1920x1080 Bayer-patterned sensor and the coloured alias level is very similar. The result is not pleasing and should not be considered suitable for use in HDTV production. The down-conversion from 2k to 1920x1080 did not involve any scaling, the Redcine software simply cropped the image, which is probably the best thing to do.'

Alan answers another question. How much luma resolution should we expect. I assume, based on the sophistication of the deBayer process, that luma is 70% to 78% of the active H. photosites.

'... in a Bayer patterned single-sensor camera, the
maximum resolution that can be expected to be delivered is about
70% is the pixel count, both horizontally and vertically, since the coloured pattern of pixels has to be decoded in order to generate the R, G and B signals.'

On the other hand, visual inspection of rez. charts show that RED software obtains about 78%. Beyond that artifacts appear.

Another way to look at this, to obtain the same luma resolution as that from a 3-chip camera, the number of pixels must be 1.6X -- for FullHD -- a Bayer chip must be about 3.4MP.

Not to detract from the advantages of RAW, but resolution will not be great from any of these new cameras even though they are called 2K/2.5K rather than FullHD. Or, put another way, although they are at the original $3K price of the Scarlet, none of them may come close to delivering RED resolution.

[Dan Hudgins]

You can see some newer test videos links and some information about the S35 and Mini models being on sale at this link,

http://www.dvinfo.net/forum/digital-...ml#post1812325

Look for my post about IBC on that page.

As for capturing very bright white subjects, after conversion to Rec.709 or DCIP3 for viewing the highlights are crushed, so there is very little to see for the most part, and if you shoot with the zebras on set to 90% raw sensor signal, and adjust the framing and exposure and use a Cir.Pola flter to control the specular reflections I have not seen it being an issue under the various lighting conditions I have shot under so far, its more of an issue at night but people expect car headlights to be burned out and flare anyway.

Its more of a problem with monitors.

As for 2.5K going to 1920x1080 for blu-ray, which may be 1440x1080 and such, the resolution of a 4K pixel wide digital image file cannot be 4K anyway as that would produce aliasing and moire. The oversample of doing 2592 or 2880 down to 1920 or 2048 gives enough loss that one can know that the original file was above the digital image limit and the Bayer filter plus OLPF did not impact the image resolution in a serious way because any reduction from those resolutions would have to produce a similar loss just because they are pixel sampled images and subject to the same issues no matter if they are film scans or sensor sourced images. You get less of the noticable de-Bayer artifacts in the end result if you oversample the Bayer pattern resolution then reduce the RGB result with the right methods.