Originally Posted by jruser
Can someone link to or explain this IEC61937 that keeps getting mentioned in something easy to understand? Google isn't being friendly to me on this topic.
With the increase in media storage devices that require compressed audio formats, applications must identify, describe, and use a variety of new encoded audio content for transmitting content from PCs to devices such as HDMI or DisplayPort receiver.
To represent an encoded audio stream to be transmitted over an IEC 61937-compatible interface, an application must provide:
The characteristics of an encoded audio stream to be transmitted.
The characteristics of a decoded audio stream on the target device.
In Windows Vista and earlier Windows operating systems, an application can infer the quality level of an audio format from the number of channels, the sample size, and the data rate of an audio stream that uses the format. For a PCM format, this information is available from the nChannels, nSamplesPerSec, and nAvgBytesPerSec members of the WAVEFORMATEX structure that specifies the format. For a non-PCM format, these three members have been commandeered to store information about the compressed data in the audio stream. Thus, the WAVEFORMATEX structure lacks information about the effective number of channels, sample size, and data rate of the non-PCM audio stream after the stream is decompressed and played. Based on the information in this structure, a user or an application might have difficulty inferring the quality level of the non-PCM stream.
The WAVEFORMATEX was extended to the WAVEFORMATEXTENSIBLE structure to provide the extra stream characteristics. However, this structure is also not adequate in describing the stream for IEC 61937 transmissions because it was intended to represent a single set of characteristics and used for uncompressed, multi-channel PCM data.
In Windows 7, the operating system addresses this problem by providing support for a new structure, WAVEFORMATEXTENSIBLE_IEC61937 which extends WAVEFORMATEXTENSIBLE structure to store two sets of audio stream characteristics: the encoded audio format before transmission and characteristics of the audio stream after it has been decoded. The new structure explicitly specifies the effective number of channels, sample size, and data rate of a non-PCM format. With this information, an application can infer the quality level of the non-PCM stream after it is decompressed and played.
The current Windows audio system did not have the necessary structure to define the advanced audio (It only supported SPDIF). The new structure adds support for the advanced HD audio codecs like TrueHD and DTS-MA to be bitstreamed over the HDMI and DisplayPort interfaces. The changes must be programmatically implemented in two places: The output device (like the 5870) and the decoder filter (like ffdshow).
Basically, if the Windows drivers for a particular audio device supports it (Windows 7 only), Directshow filters like ffdshow can be used to bitstream TrueHD and DTS-MA. You could then get bitstreaming from pretty much any player you want.
Right now, we are speculating (read as "hoping") whether or not it has been implemented in the drivers for the new ATI cards.