Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
ISO 26428-9:2009 Digital Cinema Distribution Master — Image Pixel Structure Level 3: Serial Digital Interface Signal Formatting
SDI container format, timing parameters, and synchronization for DCDM Level 3 image transmission
1. SDI Container Architecture for DCDM Level 3
ISO 26428-9 defines the formatting and constraints of DCDM Level 3 image pixel structure for transmission over 1.485 Gb/s dual-link (SMPTE 372M) or 3 Gb/s (SMPTE 424M) serial digital interfaces. This standard is essential for ensuring that digital cinema projectors and other display devices receive image data in a consistent, predictable format that guarantees accurate color reproduction and spatial fidelity. The standard specifies a fixed 2048×1080 pixel SDI container operating at 24 fps with a 74.25 MHz sampling frequency and 2750 total samples per line across 1125 total lines per frame.
The SDI container architecture is designed to be format-agnostic with respect to the actual image content. The container always carries 2048×1080 pixels, but the active image area within this container can vary. This design choice was deliberate, allowing a single SDI infrastructure to support both flat (1.85:1 aspect ratio, 1998×1080 active pixels) and scope (2.39:1 aspect ratio, 2048×858 active pixels) content without any reconfiguration of the digital interface. The unused pixels, called padding pixels, are simply set to a fixed value of 010h and ignored by the display device.
The 2048×1080 SDI container is independent of the actual image area padding pixels (set to 010h) fill unused space. This means the 2048-pixel container width supports both flat (1998×1080) and scope (2048×858) image formats without changing the SDI infrastructure, simplifying projector interface design.
2. Key SDI Timing Parameters
| Parameter | Value | Description |
|---|---|---|
| Samples per Active Line (S/AL) | 2048 | SDI container width in pixels |
| Active Lines per Frame (AL/F) | 1080 | SDI container height in pixels |
| Frame Rate | 24 Hz | Progressive scan |
| Sampling Frequency (fs) | 74.25 MHz | Interface sampling rate |
| Samples per Total Line (S/TL) | 2750 | Total line length in reference clock periods |
| Total Lines per Frame | 1125 | Vertical timing including blanking intervals |
The timing parameters define a precise schedule for data transmission. Each line of 2750 reference clock periods contains 2048 active samples (the SDI container), preceded by 694 reference clock periods of ancillary data and blanking space, and followed by 8 reference clock periods for the EAV sequence. This timing is derived from the 74.25 MHz sampling frequency, which itself is a standard broadcast video frequency that allows cost-effective implementation using existing chip sets and interface components.
3. Synchronization and Data Format Engineering
The standard defines a virtual interface data format with three data channels (X’, Y’, Z’) each having identical data structures. These three channels correspond to the DCDM-specific color space defined in SMPTE 428-1, which provides a device-independent color representation optimized for digital cinema projection. Synchronization is achieved through SAV (Start of Active Video) and EAV (End of Active Video) sequences, each comprising four consecutive code words with F (Frame), V (Vertical), H (Horizontal), and parity protection bits (P3-P0). The parity bits provide single-bit error detection, allowing the receiver to identify corrupted synchronization signals.
Values outside the valid range of 010h to FEFh must be clipped at the interface. This is not optional failing to clip prohibited values (FF0h-FFFh and 000h-00Fh) can cause synchronization errors or equipment damage. The 12-bit code values provide 4096 possible levels, of which only the range 16-4079 is valid for image data.
The progressive segmented frame (PsF) system provides a bridge between progressive and interlaced domains, which was important for compatibility with existing broadcast infrastructure when the standard was developed. In PsF mode, a progressive image is transmitted as two segments mapped onto an interlace interface, with the first segment occupying lines 21-560 and the second occupying lines 584-1123. The PsF system maintains the full vertical resolution of the progressive image while using an interlaced transmission format, making it compatible with legacy SDI equipment. The relationship between progressive line N and PsF line M is given by the equations M = N/2 for even N and M = (N+1125)/2 for odd N.
For new D-cinema projector designs, supporting both progressive (F=0 on all lines) and PsF modes ensures compatibility with the widest range of DCDM sources. The dual-link (SMPTE 372M) and single-link 3G (SMPTE 424M/425M) interface options provide deployment flexibility based on existing infrastructure, with single-link preferred for new installations due to simpler cabling.
4. Frequently Asked Questions
Q: What is the purpose of padding pixels in the SDI container?
A: Padding pixels (value 010h) fill the portions of the 2048×1080 SDI container that are not occupied by the actual image. This allows different aspect ratios to use the same SDI infrastructure without reconfiguration. The display device simply ignores padding pixels and displays only the active image area.
A: Padding pixels (value 010h) fill the portions of the 2048×1080 SDI container that are not occupied by the actual image. This allows different aspect ratios to use the same SDI infrastructure without reconfiguration. The display device simply ignores padding pixels and displays only the active image area.
Q: How does the 1.485 Gb/s dual-link interface differ from the 3 Gb/s single-link?
A: Dual-link splits the signal across two cables (each carrying half the data, with one link carrying X’ and half of Y’, the other carrying the remainder), while 3 Gb/s single-link carries all data on one cable with higher serial bit rate. Both transport the same image format the choice depends on available infrastructure and cable plant.
A: Dual-link splits the signal across two cables (each carrying half the data, with one link carrying X’ and half of Y’, the other carrying the remainder), while 3 Gb/s single-link carries all data on one cable with higher serial bit rate. Both transport the same image format the choice depends on available infrastructure and cable plant.
Q: What is the significance of the X’, Y’, Z’ color coding?
A: X’Y’Z’ is the DCDM-specific color space defined in SMPTE 428-1. It differs from conventional RGB or YCbCr and provides a device-independent color representation optimized for digital cinema projection with a wider gamut and gamma encoding suited for high-brightness projection environments.
A: X’Y’Z’ is the DCDM-specific color space defined in SMPTE 428-1. It differs from conventional RGB or YCbCr and provides a device-independent color representation optimized for digital cinema projection with a wider gamut and gamma encoding suited for high-brightness projection environments.
Q: Can ancillary data be carried alongside the image data?
A: Yes, both HANC (horizontal ancillary) data in the interval between EAV and SAV, and VANC (vertical ancillary) data in lines outside the container boundary, are supported per SMPTE 291M. This allows metadata, audio, caption, and other data to accompany the image stream within the same SDI signal.
A: Yes, both HANC (horizontal ancillary) data in the interval between EAV and SAV, and VANC (vertical ancillary) data in lines outside the container boundary, are supported per SMPTE 291M. This allows metadata, audio, caption, and other data to accompany the image stream within the same SDI signal.
