Forum Replies Created

Page 7 of 97
  • Chris Kenny

    October 24, 2013 at 9:31 pm in reply to: The new MacPros

    [Juan Salvo] “RedRockets work inside of a thunderbolt to PCi enclosure. But you’re limited by the thunderbolt speed. For RedRocket legacy it’s not such an issue, but the RedRocket-X will be limited by this.”

    If Red offers the option to feed a 4K signal back to the host system with production-quality down scaling then bandwidth issues on Thunderbolt 2 would only come into play if you needed to actually work at 6K in real time, which will be pretty rare (since you can’t actually monitor it, etc.). It’s not clear if they’ll do this, however.


    Digital Workflow/Colorist, Nice Dissolve.

    You should follow me on Twitter here. Or read our blog.

  • [michael stirling] “This would mean in practice that we’d have the client media on the internal RAID but may (in the case of features at least) have to render to the file server RAID”

    What format(s) are you generally rendering to and how are you connecting to the file server? Just gigabit ethernet, or something faster?


    Digital Workflow/Colorist, Nice Dissolve.

    You should follow me on Twitter here. Or read our blog.

  • Chris Kenny

    September 24, 2013 at 9:08 pm in reply to: OT iMac updated, Macbook Pro soon?

    [Jeremy Garchow] “I just don’t think it’s a format that Apple would be interesting in supporting because it’s not going work on a Mac Mini. Unless, of course, there’s an easy way to transcode to ProRes.

    I hope I’m wrong. “

    It’s easier to decode than Epic footage (lower resolution, no wavelet compression), and they support that. Presumably FCP X’s existing features for generating proxy/optimized ProRes media would be available.


    Digital Workflow/Colorist, Nice Dissolve.

    You should follow me on Twitter here. Or read our blog.

  • Chris Kenny

    September 13, 2013 at 3:52 pm in reply to: confused over arri alexa log c behavior

    [Joakim Ziegler] “In general, it’s often not necessary to use the LUT, you will get better results just grading with custom curves directly on the LogC material, and you have much more control.”

    You can also have the Alexa LUT generator give you a LUT that applies the proper color matrix but retains the log gamma so nothing clips.


    Digital Workflow/Colorist, Nice Dissolve.

    You should follow me on Twitter here. Or read our blog.

  • [Eric Michael Cap] “Just had a conversation with Dave Helmly about this, and while the PCIe Thunderbolt card does feature Displayport IN as per Intel’s Thunderbolt specs, it’s not required to be connected – unless you want the video looped through.”

    Interesting. Didn’t Intel decline to certify an ASUS motherboard that tried to satisfy the video requirement with a DisplayPort input last year? I wonder if they’re opening up with this in general, or if they’ve got some special arrangement with HP. And I wonder how HP is handling the cabling here. Just externally, with a cable looping up from the graphics card to the Thunderbolt card? Or are they doing something more interesting, like OEM GPUs with internal DP connectors?


    Digital Workflow/Colorist, Nice Dissolve.

    You should follow me on Twitter here. Or read our blog.

  • [Herb Sevush] “Is there any reason to believe that you couldn’t use more cards if you needed more ports? Also under what configuration would you need more than 1 daisychaining Tblt port if you also have multiple PCIe slots available?”

    I’d assume each card will need to have DisplayPort looped into it, which might get annoying with multiple cards. Unless Intel has dropped the ‘all Thunderbolt ports must carry video’ requirement as well, or there’s some other scheme to get video on these ports.

    As for why you’d want more than one port… there are enough Thunderbolt devices out there that don’t have a passthrough port that just having a single port will be a bit annoying on a workstation. This includes a bunch of stuff that the presence of PCIe slots doesn’t really reduce the need for, like external drives.


    Digital Workflow/Colorist, Nice Dissolve.

    You should follow me on Twitter here. Or read our blog.

  • If you read the details, it says they’ll offer a PCIe card for their desktop workstations with one Thunderbolt 2 port in ‘early 2014.’ That’s notable in that Intel hasn’t previously allowed Thunderbolt to be offered via PCIe cards (or on any Socket 2011 Wintel system). But they’re not going to beat Apple to market, and only offering a single port is not especially impressive.

    Still, though, the fact that this is being offered at all suggests HP is seeing some demand for Thunderbolt among Wintel workstation customers, which is interesting.


    Digital Workflow/Colorist, Nice Dissolve.

    You should follow me on Twitter here. Or read our blog.

  • [Santiago Martí] “I’m sure it will be enough for a lot of people, but it will be expensive enough to make the Imac a better OSX choice.”

    The Mac Pro has traditionally been price-competative with similarly specced Wintel workstations immediately following model refreshes. I don’t see why that would change with the new design.


    Digital Workflow/Colorist, Nice Dissolve.

    You should follow me on Twitter here. Or read our blog.

  • Chris Kenny

    September 1, 2013 at 2:20 pm in reply to: Looking Ahead to 10.1

    [Andrew Richards] “They aren’t using SQLite directly, but rather via Core Data. Core Data has a method for hooking up to other RDBMSes, so it is technically feasible. “

    There’s no need to hook Core Data up to a multiuser RDBMS to enable collaborative features. Core Data itself has live syncing features, mediated through iCloud. In this scenario, each client system would retain a local, single user SQLite store, but Core Data would update those stores as other clients pushed incremental project updates to iCloud. This technology is designed to implement precisely the sort of real-time collaboration being discussed here — User A and User B both have the same sequence open, user A moves a clip, and User B sees that clip move almost immediately.

    Shared access to media (which is actually an easier problem) would be handled through a separate mechanism.

    Mind you, Core Data syncing is presently considered by many developers to be too buggy to actually, well, use for anything. But it’s hard to imagine Apple hasn’t been putting in a lot of work to fix it, since it’s a key enabling technology for the multi-device vision Apple has for computing. iOS 7 will probably be our first chance to see if they’ve fixed it.


    Digital Workflow/Colorist, Nice Dissolve.

    You should follow me on Twitter here. Or read our blog.

  • Chris Kenny

    August 24, 2013 at 2:02 pm in reply to: I know that some lust after this….

    [Jason Brown] “Thanks for posting, nice to know that Apple is still completely out of touch with what their customers want. Information.”

    This product is known to contain at least one unannounced component (the CPU) and could contain others (the GPU specs don’t match up with any current product, though Apple could just be clocking a current product differently). Apple may consequently be contractually prevented from discussing specs or prices in too much detail at this time.


    Digital Workflow/Colorist, Nice Dissolve.

    You should follow me on Twitter here. Or read our blog.

Page 7 of 97

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