Creative Communities of the World Forums

The peer to peer support community for media production professionals.

Activity Forums ARRI ARRIRAW Questions

  • ARRIRAW Questions

    Posted by Zak Ray on January 31, 2011 at 2:35 pm

    If anyone happens to know the
    -Bitrate
    -Color space (RGB?)
    -Supported audio channels
    -Compression algorithm (DCT? Wavelet?)

    of ARRIRAW, I’d appreciate it. Thanks!

    Gary Adcock replied 12 years, 3 months ago 11 Members · 29 Replies
  • 29 Replies
  • Gary Adcock

    February 1, 2011 at 1:40 pm

    ARRIRAW bit rates for Alexa / D21 are similar to Uncompressed at 2880×1620.
    The .ARI frames are 7MB each, DPX files from the same output are 19MB per frame.

    ARRIRAW is recorded from the Tlink data stream over Dual Link HDSDI to specific recorders (Only Codex, Stwo and KG are supported today as far as I know)

    the Color space is RGB.

    ARRIRAW does not support audio, as it is captured as a sequence of frames just like DPX.

    I am not sure of the compression algorithm, I will check.

    gary adcock
    Studio37

    Post and Production Workflow Consultant
    Production and Post Stereographer
    Chicago, IL

    https://blogs.creativecow.net/24640

  • Zak Ray

    February 1, 2011 at 2:52 pm

    Thanks, Gary, I’d love to know.

  • Christy Mackarell

    February 1, 2011 at 3:59 pm

    Does this answer the compression algorithm question?:

    “An Uncompressed RAW signal is the direct electrical value feed from the photosites of the camera sensor without any processing or manipulation. Think of it as the “digital negative” from the camera”
    -Abel Cine Tech blog.

    Lots of folks can do an ARRIraw debayer: Baselight
    Codex, Nucoda etc. But right now, very few have implemented ARRI’s own SDK to do it; it’s not released generally at time of writing.

    Christy MacKarell
    Deluxe NY

  • Zak Ray

    February 2, 2011 at 2:53 am

    So… it’s uncompressed?

  • Fredrik Harreschou

    February 6, 2011 at 12:02 pm

    Actually, RED has patented in-camera raw compression (Redcode). So unless Arri wants to license Redcode to use in their products, they will have to do uncompressed raw.

  • Gary Adcock

    February 7, 2011 at 2:29 pm

    [Fredrik Harreschou] “Actually, RED has patented in-camera raw compression (Redcode). So unless Arri wants to license Redcode to use in their products, they will have to do uncompressed raw.”

    Arri has always regarded ArriRaw as an uncompressed capture- this is more because of it being a direct data capture as opposed to being a signal stream over T-link.

    for the record.
    The ArriRaw codec released to the public 3-years(?) before Red Digital was a camera company, at the time ArriRaw was released the only way to capture was as an uncompressed signal.

    gary adcock
    Studio37

    Post and Production Workflow Consultant
    Production and Post Stereographer
    Chicago, IL

    https://blogs.creativecow.net/24640

  • David Battistella

    February 7, 2011 at 5:49 pm

    I always thought ARRIRAW was based on wavelet compression scheme. That makes it 2.5:1 lossless compression if this is true.

    For the record REDCODE raw is also a wavelet based scheme. But at 2.5:1 it would be considered lossless which is mathematically exactly the same as uncompressed.

    David

    Peace

  • Fredrik Harreschou

    February 8, 2011 at 8:21 pm

    Hi!

    My comment was directed to Zak who asked if it was uncompressed. Which I assume it has to be, because of the RED patent.

    First, just to make it clear, I am no troll or Red “fanboy”. We actually received our first of two Alexas last week. But I think it is an interesting discussion.

    David, I am not sure if ARRIRAW is wavelet compressed. And Gary seems to confirm that. According to Jim Jannard from RED, the reason they were able to get their “in-camera raw compression patent approved was because the competition was asleep”, me paraphrasing. The thread was actually removed from reduser.net soon after he posted. I am *guessing* that it was for legal reasons.

    If indeed RED has patented compressed RAW in camera, as they claim, that will be a big issue for the competition when they want to go 4k with bayer pattern sensors. Something I hope and believe Arri has on their roadmap. The I/O requirements for getting uncompressed 4K RAW off the camera will be challenging (read: expensive) for years to come, the same with storage needs. And debayered RGB will be even higher data rates, or heavily compressed, and demand a lot of the in-camera processing to do a full debayer in realtime.

    Cheers,
    Fred

  • Fredrik Harreschou

    February 8, 2011 at 8:28 pm

    The reason the RAW file is smaller (approx 1/3) is because the RAW signal is one channel, not three as in RGB. The RGB file is made during debayer/demosaic. Or put another way, RAW is inherently a 3:1 compressed RGB file.

    EDIT: Sorry, I am not used to the different reply functions on this board (how do I quote?). This was a reply to Gary´s comment about file sizes in ARRIRAW vs DPX.

  • Robert Monaghan

    February 9, 2011 at 5:11 pm

    Hi Everyone,

    (A bit of background.. I am the first “ARRIRAW Partner” that ARRI signed on. I have (and continue to be) a developer supporting the camera and their image formats. My Info is first hand.)

    This information is already publicly available, but I will summarize it here:

    – There is no compression. Period. No wavelet, DCT, etc. Just uncompressed, untouched, unmolested bayer pattern image data. *Truely* raw data. Not a compressed bayer image, as others would use.

    – T-Link is a packing method to transmit the bayer data to a recording device. Think of HD-SDI as a “video specific” ethernet system. Not only does HD-SDI send video packets over its network, but it can be re-purposed to send data packets, too. In T-Link mode, the D-21 and Alexa send bayer data over the HD-SDI *AS DATA* — not as video.

    – The D-21 and Alexa produces a 12-bit Bayer Sensor image which is sent to the recording device.

    – There is no Audio support with ARRIRAW files. It is just bayer sensor data. Other means of recording the audio will have to be adopted. Consult with your recorder manufacturer. They may have a solution already.

    – The Colorspaces supported after the image is processed, will depend upon the tools you use.
    In general, any company that uses the ARRI SDK should get what ever the SDK offers. For example, you should get Rec709 and LogC. DCI P3 is also available. I imagine other colorspaces would be supported in the SDK in the future. If you had something specific that you needed, I would contact ARRI and put in a request.

    – ProRes that is recorded from an Alexa via the SxS card uses the very same ARRIRAW color processing engine. If an ARRIRAW image and a Rec709 ProRes movie are recorded at the same time, they should look pretty much identical, color wise. Naturally, the resolution and debayer quality will be different.

    Hope this helps.

    bob.
    https://www.gluetools.com

    Robert Monaghan, CEO
    Glue Tools LLC
    POBox 24124
    Santa Barbara, CA, 93121
    United States

    tel: +1 805 456 7997
    fax: +1 805 456 7998

    http://www.gluetools.com

Page 1 of 3

We use anonymous cookies to give you the best experience we can.
Our Privacy policy | GDPR Policy