Selectable Dvcpro HD Recording

The AG-HPX370 series also supports the conventional DVCPRO HD codec. Because the AG-HPX370 series are designed to be used with the AG-HPX170 series/HVX200A series or a DVCPRO HD VTR, it adapts to a variety of system configurations.

48-kHz/16 bit, 4-Channel Digital Audio

The AG-HPX370 series can record full 48-kHz/16 bit digital audio on all four channels. You can freely select the audio source for each channel, choosing from mic-in, line-in and wireless receiver.

A V C - I n t r a T e c h n o l o g y

Sample Images of Intraframe Preduction	Left: Original image Center: Intra-frame predictive image Right: Difference image obtained from
	subtracting the intra-frame predictive image from the original image. This shows the high accuracy of
	intra prediction.

Intra-Frame (I-Frame Only) Compression Superiority

Motion-image compression can be divided roughly into two methods: I-Frame Only compression, which completes all processing within each frame, and Long GOP compression, which processes across multiple frames. AVC-Intra and DVCPRO HD use I-Frame Only compression, while HDV uses Long GOP compression.

The MPEG-4 AVC/H.264 standard encompasses both methods.

When the images of adjacent frames are similar, Long GOP compression achieves an advantageously low bit rate. However, this trait is not often seen in broadcasts like flash-filled press conferences, fast-action sports, and video with rapid move and electronic displays. Also, because processing is performed frame-by-frame in I-Frame Only, new-generation multi-core CPUs offer high-speed parallel processing. This makes I-Frame Only compression more suitable for nonlinear editing than Long GOP, for which parallel processing is difficult due to its inter-frame dependence. With its I-Frame Only compression, AVC-Intra produces remarkably stable images that are unaffected by adjacent frames, and meets professional needs in virtually all

Twice the Compression Efficiency of MPEG-2

By selecting the most effective compression techniques from among those in compliance with the H.264 standard, AVC-Intra has doubled the compression ratio of MPEG-2, even with I-Frame Only compression. Its intraframe predictive and context-adaptive entropy coding are particularly effective methods for boosting compression efficiency.

Intraframe predictive coding (intra prediction)

This process generates predictive images based on adjacent blocks of 8 x 8 pixels. Selecting the most suitable predictive mode from among nine luminance signal modes (see illustration) and four color signal modes, it generates accurate predictive images. The residual data (obtained by subtracting a predictive image from the original input image) is recorded together with the predictive image. Because the prediction accuracy is high, there’s minimal residual data, and thus high compression is achieved. This process is conducted within the frame, so prediction accuracy remains high even with fast-motion images.

situations and workflows.

mode 0

mode 4

mode 8

mode 1

mode 5

Average
mode 2	mode 3

mode 6

mode 7

Adjacent pixels Predictive pixels Copy of below pixel

Context-adaptive entropy coding

The entropy coding process used in MPEG-4 AVC/H.264 utilizes CAVLC (Context Adaptive VLC) and CABAC (Context Adaptive Binary Arithmetic Coding), both of which are context adaptive. MPEG-2 uses a fixed table when performing the VLC coding, with the result that compression efficiency is low with some types of images. In context-adaptive coding, on the other hand, operation varies with different kinds of images and high compression efficiency is maintained at all times.