Chinese 2K RAW S35 Cinema Camera, 8000 dollars

Does anyone know how the Kineraw MINI gets from 4k at the sensor to the much lower resolution recording formats (2k, 1080p, 720p)? For example, are they using line skipping or cropping or interpolation or binning?

Kinerawm S35 which is the big brother to the mini uses pixel binning to go from 4k to 1080p resolution
 
New KineRAW-S35 (tm) EI ISO 160 to 5120 test reel on Vimeo

New KineRAW-S35 (tm) EI ISO 160 to 5120 test reel on Vimeo

I posted a new EI ISO test reel on Vimeo, it shows the image quality and noise level after finsh processing in my free de-Bayer program of the DNG frames shot. I pulled short segments of some few shots of the many I have made that represent various lighting conditions, the contents of that reel are:

1) Full range graded and de-noised samples EI ISO 160 to 5120

2) through 7) Range adjusted versions for people with LED or HD monitors that are ITU 601 limited and they don't have their viewing program set to compensate. These also help you see the shadow and highlight detail without clipping on any monitor.

8) Half size workprint with SMPTE burn in for TOD and frame in shot time code along with 2.39:1 guidelines. There are other fields that my de-Bayer program can burn in from the extra DNG meta-data such as the head and tail auto-slate markes, and slate data. Making burned workprint allows editing outside my DI system of 24.000fps cinema footage, even on 29.97 editing programs if you frame rate convert the burned workprint, and then you just conform the SMPTE frame number code (frame 00000000.DNG equals 00:00:00:00 time code, the TOD is time of day SMPET from the camera for conformation if needed to the camera log and Digital Slate is used) into the edit list spread sheet in my DI system and your uncompressed DI output will conform to your workprint edit on any NLE such as FCP (my program has a built in NLE, once you have entered the in and out SMPTE from the workprint, my programs NLE can playback the edit list's edit and you can confirm that the numbers were entered correctly, and make any adjustments needed as a double check, workprint for external editing is not required, but some people may want to do that because they can convert the workprint frames to H.264vfw and work quickly in any low cost NLE on a laptop, but retain the full 48bpp TIF or DPX version for the conform in my freeish DI system).

9) RAW images showing the actual sensor exposure level on each shot, this is only in full range so your monitor may clip the highlights somewhat more than the camera did.

10) These are still frames taken from the cameras reference BMP files, the camera saves a reference screen shot of the monitoring image for color reference to have as a record of how the image looked when you were shooting and it includes the zebras if they were turned on. It shows the signal levels, but if you are using a LCD monitor on the camera that monitor distorts the image differently than the monitor you will be viewing these images on, but it gives a general idea of how I set the exposure while shooting, and then later graded the results for the full range shot set #1. I had intentionally exposed some shots light and low contrast knowing that I would later reduce the mid-tones to darken the overall effect.

https://vimeo.com/67738022

You should use the D-LAD image before each shot group to calibrate your monitor, the gray scales included in the D-LAD image are described in the documentation for my de-Bayer program in its ZIP file. Those gray scales are netural, so if the darker parts look red or magenta its a problem with your monitor probably, I write those gray scales directly to the image file without grading it, so they are absolute signal levels, I put that procedure in my de-Bayer program to make working with DPX and CIN files simpler since you get a reference to adjust when grading even if no gray card was shot etc. so that the absolute singal levels are known throughout the workflow chain, after the templates are programed, the un-tagged frames are processed knowing that the transforms are accurate to target signal levels.

This footage was shot with a prototype KineRAW-S35 (tm) camera, the current production cameras may produce somewhat different results.

I should add that de-noise is part of normal finish processin for all modern cameras, if you shoot RAW through CinemaDNG its done as part of the post production workflow, if you shoot compressed H.264 cameras its done as part of the pre-compression signal prep or in the compression processing. RAW signals show more noise when viewed without de-noise because they hold the original data, and to prep them so they can be compressed for shoing on the internet they need to be "de-grained" just as 35mm film scans are now also de-grained, you don't see raw film scans over compressed broadcast as there is not enough bandwidth to carry the 24fps grain from film or CinemaDNG camera footage. But with CinemaDNG you can control the post processing and still retain the original 99 to 100% sensor data to re-do things if needed.
 
Last edited:
KineRAW (tm) cameras finally going to be shipped outside China?

KineRAW (tm) cameras finally going to be shipped outside China?

It seems soon people outside China will be able to get the KineRAW-S35 (tm) Mini?

There seems to be a discount if you order using the Cinema5D link,

Quote: [HDvideoshop gave us a special deal for our readers:
Free 128GB Samsung 840 Pro SSD with your KineRAW mini order. CODE: cinema5d
3800€ (inc. tax) available here: hdvideoshop.com
(shipping to US is 89€)]

S35/Mini Links:

http://www.hdvideoshop.com/en/271-kinefinity

http://www.cinema5d.com/?p=20838

==

I've postes some new videos made with the KineRAW-S8p (tm) on my Vimeo page, I'm working on a slow motion skate board video shot at 220fps and other rates with that camera, I hope to post that before too long, the editing is a bit complex with multi images so I'll get it up when I have time to finish shooting, processing and compositing the shots for that one. There is one up shot with an Ang 6-80mm f/1.2 Super8 format zoom lens now, and an EI ISO 100 to 800 test reel shot with various lenses. I still have to shoot with CS and D mount lenses for the S8 format options, I'll be getting to that when the skate board video is done...

S8 Links:

http://vimeo.com/72106324

http://vimeo.com/68791065

If you have not seen it yet, here is a S35 video from the previous sensor balance,

http://vimeo.com/67738022

The current cameras are said to have a different sensor revision, I don't have one of those to recalibrate with yet, so if you see some color shifts those should be able to be compensated for with revisions to the monitoring tables and white balance tables (LK5 and LT5 files that are user loadable from a USB flash thumb drive).
 
Last edited:
Quote: [Hi Dan

We will not attend IBC next week. But there will be a booth shows our
cameras, its number is 11C 80 (Lanparte Booth).

Jihua]

So at least you will be able to see the cameras there, the color on the monitor may not look right because the lights in the area may not be standard balance, that can be corrected with the manual analog gain override menu, you adjust the red and blue analog gain control to get the histogram peaks of the RGB post histogram display to align, you just adjust red and blue because green sets the EI ISO rating. You point the camera at something white, adjust the iris to get the green waveform to be 45% and then look at the RGB post histogram, first turn the red gain up or down to superimpose the red peak over the green peak, then go to the blue adjustment and adjust the blue to align with what now is a yellow peak, if you use the 800% zoom on the monitoring you can exclude other things in the frame and get a clean peak. The camera will reset the manual analog gain override if you change any of the other settings such as ISO or K, so you do the tweaking of the color balance last when setting the exposure. Its possable to add the numbers you see when you get the color balance right to the EQUAL.LT5 file manually to make another K plus Light Type setting in the cameras menus, you need to get the EQUAL.LT5 file they loaded and edit it with notepad.exe or some other text editor (watch out for ones that reverse or omit the CR/LF end of line marker) then load the edited EQUAL.LK5 file into the camera from a USB flash thumb drive, there is a prompt in the Config menu to "load user files" or something like that, take the USB drive out after it says "nnn user files loaded!" then reboot the camera, you should now see your custom white balance in the cameras menu. If green gain was larger than 1.0, you need to scale the red and blue gain values by it in order to make the LT5 entry work, so if green is 2x and red was 3.5x then you make a fraction 3.5/2 = 1.75 and you enter 1.75 as the red gain value, the camera will compute it back to 3x when green is set to 2x by the ISO adjustment when shooting, likewise for the other ISO gain changes which is why you enter all values in EQUAL.LT5 based on green gain is 1x at this time.

You should also check the monitoring adjustments menu in the camera, see if black is set to 16 and white to 235 and highlight to about 8 and see if that looks right on the monitor they have it connected to, you can also adjust the monitor output gamma level to get midtone to look right as many LCD monitors are not quite close to CRT monitor standard of gamma 2.22 etc, maybe.
 
It should be obvious that such a software on the PC/MAC doing DNG to Cineform (tm) Bayer "raw" (and also the reverse) can be used with the KineRAW-S8p (tm) and other DNG recording cameras to get high quality Cineform (tm) MOV for playback and editing, to finish for broadcast or non-critical use, and as a proxy in case you want more choice in the de-Bayer of the full DNG data for the finish conversion to DCP and IMAX etc..

Fast utility here for RAW to Cineform transcoding, including audio merge. Works great with BMCC DNGs already.

RAW 2 CF RAW
 
Miguel, I'm sure our program is not exactly the same as someone else's :) In general though, I'd love for someone to make and maintain a page that would compare different RAW converters and what they do.

Here's what I found from quick search on kinestation: http://www.bmcuser.com/showthread.php?5187-KineSTATION

Here's our current version of RAW 4 PRO : https://vimeo.com/album/2657825

We're moving towards the Batch Edition, and the next release will be adding popular video codecs on top of Cineform.
 
Thanks Alex. Actually I was going to give it a shot and make the donation. I'm just hesitating because I was left with the impression that it requires Cineform’s dpx2cf command-line utility, which is found with Cineform and GoPro Studio installations. Does this mean I need another app installed for RAW 4 PRO to work? How much would that cost?

My interest is because I intend to buy a Kineraw Mini soon... And if I can have a cheaper encoding solution other than KineStation, I'll take it.
 
Miguel, you are welcome!

Cineform: my understanding is they offer a full-functioning 15-day trial. Beyond that, you will need to license their encoder - please refer to their site for details/pricing.

However... we are adding MORE rendering options to RAW 4 PRO converter, this time free. These will be "popular codec" video files that work on both Mac and PC. Since we are offering more than Cineform starting with release v1.1.7, our app's name has been changed to RAW 4 PRO. More universal and better reflects what this helpful software does, we hope.
 
Back
Top