[libdc1394-devel] IIDC Proposal: Little-endian mode for 16-bit pixel data
Capture and control API for IIDC compliant cameras
Brought to you by:
ddouxchamps,
gordp
From: David M. <dcm@MIT.EDU> - 2007-05-24 06:02:07
|
IIDC Proposal: Little-endian mode for 16-bit pixel data Proposal revision: 1 Author: David Moore <dc...@ac...> Introduction: Prior versions of the IIDC specification use big-endian encoding (high byte first) for 16-bit per component pixel formats. This proposal adds a mode where these formats are transferred in little-endian encoding (low byte first). Specification changes: Addition to the table in section 4.3 "Inquiry register for basic function": Offset Name Field Bit Description 400h BASIC_FUNC_INQ LSB_First_Inq [9] Inquiry for least-significant- byte-first mode. Addition to the table in section 4.6 "Status and control registers for camera": Offset Name Bit Description 630h Data_Depth [8] Set to 1 to enable least-significant-byte-first mode for 16-byte per componenent pixel formats. Set to 0 to disable (the default) and use most-significant-byte-first mode. Additional text for section 4.6.1.2 "Data_Depth": "Bit 8 of this register, when set, enables least-significant-byte-first mode. In this mode, pixel encodings that are larger than 8-bits per component will be streamed with their least-significant-byte first (little endian) instead of the default big endian format. Changes to this bit will not affect isochronous data until the next time streaming is enabled." Change to section 5.1.3 "Video data payload structure", under heading "<Y (Mono16) format>": "When least-significant-byte-first mode is disabled: +--------------+-------------+ | High byte | Low byte | +--------------+-------------+ When least-significant-byte-first mode is enabled: +--------------+-------------+ | Low byte | High byte | +--------------+-------------+" Repeat this same change for every appearance of a 16-bit per component pixel format in sections 5.1.4, 5.2.2, and 5.2.3. |