xFile Products  FAQs

Below you will find answers to our most frequently asked questions for the xFile framework and related products. If you cannot find answer to your query submit your technical support case at support@s-a-m.com or call us at +44 (0) 118 921 4214 (09:00 - 17:00 GMT weekdays). Our support team looks forward to responding to your questions. 

    1. What other factors will affect Snell OD performance?

    Conversion timings vary based on a combination of: 

    CPU core quantity and speed 

    PCIe bandwidth capability 

    • PCIe buses have several iterations. Each newer version has approximately doubled the processing bandwidth capability with its predecessor. 

    File access bandwidth

    • Local disk read/write access 
    • Network storage read/write access 
    • Codec - compression ratios
    • File types – large single file or dpx 

    Memory 

    • Memory speeds - DDR2/DDR3 
    • Additional memory is required to cover 2k/4k conversions. Generally an additional 16 Gb memory per GPU. 
    • Not enough system memory will cause disk swapping, which will impact the conversion speeds.

    2. Which AMD driver can I use to run Snell OD product?

    For the up to date AMD driver information, please refer to AMD website here.

    3. Which NVIDIA driver can I use to run Snell OD product?

    For the up to date NVIDIA driver information, please refer to NVIDIA website here

    4. What file types does Snell OD and its products support?

    Snell OD has been optimized to support a wide range of file types. You can see the full in our Snell OD Supported file types document here

    If you cannot find the file you wish to convert on this list, please do not hesitate to contact us. We are committed to constant product development and capturing requirements is important part of this process.

    5. With AMD GPUs can I use the same GPU to drive a local display and operate Snell OD product?

    Yes this is possible as AMD GPUs can facilitate both an OpenCL environment and provide a graphical output to a monitor simultaneously, however we recommend not sharing a common resource in this way and promote the use of a dedicated resource to drive a display. From Version V1.1.0.10 Alchemist OD offers the ability to pick and choose which GPUs are used for image processing by the Snell OD Node.

    6. Can I use multiple, different types of GPUs with Snell OD?

    Yes this is possible, however the types of GPU available should be considered as a higher quantity of GPUs doesn’t always equate to faster performance.

    Example: 3 x K20 will be more effective than 3 x K20 + Quadro 2000. This is because in the latter the K20’s will operate at the same speed as the Quadro 2000. 3 x K20 is far more effective than 4 x Quadro 2000.

    This is why typically we recommend using matched (same specification, brand, name) GPUs.

    From Version V1.1.0.10 Alchemist OD offers the ability to pick and choose which GPUs are used for image processing by the Snell OD Node.

    7. How can I utilise a Firewall if I want to install a distributed system with multiple nodes?

    If the Snell OD node is distributed across several servers then the following information will be required if there are firewalls placed between the servers.

    Source Service Source Port Destination Service Destination Port TCP/UDP
    Snell OD Client
    Snell OD Client
    Snell OD Server
    Snell OD Server
    Snell OD Server
    Snell OD Server
    Snell OD Node
    HIGH PORT
    HIGH PORT
    HIGH PORT
    HIGH PORT
    HIGH PORT
    HIGH PORT
    HIGH PORT
    Snell OD Server
    Snell OD Watcher
    Snell OD Browser
    Snell OD Watcher
    Snell OD Node
    License Server
    License Server
    35061,35060
    35063
    35062
    35063, 35069
    35064
    5093
    5093
    TCP
    TCP
    TCP
    TCP
    TCP
    UDP
    UDP

    8. When I enable TCC mode the local monitor stops working?

    When NVIDIA Quadro GPU’s have TCC (Tesla Compute Cluster) mode enabled, the ability of the GPU to display a picture to a local monitor is disabled.

    When a single GPU is installed and is set to TCC mode, there will be no local display device available. Remote access will be required.

    If a local display is required then an additional GPU can be used, this could include an on-board graphics device or a dedicated GPU.

    9. Does Snell OD support Apple ProRes?

    Yes, however due to licensing guidelines there are some circumstances ProRes will be automatically disabled by the software. ProRes will be enabled when the Snell OD Node is installed on a Windows Server operating systems. ProRes will not be enabled when Windows 7 / 8 is used. Due to licensing requirements, the encoding and decoding of Apple ProRes file formats is not permitted when an X Window system is running a local session i.e. if an X server is running on the Snell OD Node host machine, Apple ProRes jobs will not be processed.

    10. What GPUs do I need to successfully install Snell OD?

    NVIDIA GPU support

    C1060* K10 Quadro 4000 K2000
    C2050* K20 Quadro 5000* K4000
    C2070 K40 Quadro 6000*   K5000
    C2075* K6000*

    AMD GPU support

    AMD FirePro™
    AMD HD6000 Series*
    AMD HD7000 Series
    AMD R7 Series*
    AMD R9 Series

    Conversion times will vary depending on specification. 
    Please note that multiple cards should be matched. 
    *These cards have not been verified by Snell Engineering

    11. How do I add additional licenses to increase the number of jobs I can run at any one time?

    Upon purchase of a new license you will be given instructions as to how to proceed. However, we will need the locking code of your license server to proceed with the license install.

    For details on the locking code retrieval procedure, see section 8.2 of the Product installation guide.

    Once the license is available simply update your license server and install the Snell OD node on your new hardware. The new node will need to be configured within the system page (see product user guide). Once configuration is complete multiple jobs will be able to be processed in parallel.

    12. Why do I get an error message every time when I try to run a conversion?

    If you are using NVIDIA GPUs and running Microsoft Windows Operating system, it is important that the GPU runs in TCC mode. Professional NVIDIA Quadro GPU cards can run in one of two modes, WDDM (Windows Display Driver Model) or TCC (Tesla Compute Cluster). NVIDIA Tesla GPUs will default to TCC mode. In TCC mode the display port is disabled. Enabling this mode will disable the graphical output of the specific GPU, but will enable the OpenCl environment required by our software. To learn more, see Appendix C in the installation guides.

    13. How can I set up or change the ARC (Aspect Ratio Conversion) parameters?

    Alchemist OD offers comprehensive aspect ratio conversion.

    Generally, where conversions are SD to SD, or HD to HD, with the ARC settings in default, the output picture will mirror the input with respect to aspect ratio.

    In the case of Up-conversion, or Down-conversion, care should be taken to ensure that the desired aspect ratio parameters are associated with the output file.

    To learn more, see Snell OD How to configure the ARC document here

A Quantel Company © 2013 - 2018 Snell Ltd. All rights reserved.  Our Cookie Policy