Lines Matching full:based
135 * Allow recode for all frame types based on bitrate constraints.
154 // Estimate the level based on quantizer and frame type
172 CDEF_PICK_FROM_Q, /**< Estimate filter strength based on quantizer. */
178 // Terminate search early based on distortion so far compared to
205 // More aggressive pruning based on tx type score and allowed tx count
214 // Set to larger Q based only on the detected slide/scene change and
276 // Threshold used by the ML based method to predict TX block split decisions.
289 // inter blocks. It enables further tx type mode pruning based on ML model for
326 SUPERRES_AUTO_DUAL, // Tries no superres and q-based superres ratios
327 SUPERRES_AUTO_SOLO, // Only apply the q-based superres ratio
355 * \brief This enumeration defines a variety of simple motion search based
364 QIDX_BASED_AGG_LVL1, /*!< Qindex based prune aggressiveness level, aggressive
365 level maps to simple agg level 1 or 2 based on qindex.
371 SIMPLE_AGG_LVL3, /*!< Total number of qindex based simple prune
388 * non-rd path based on sse value.
392 0, /*!< Early terminate inter mode search based on sse disabled. */
394 1, /*!< Early terminate inter mode search based on sse, index 1. */
396 2, /*!< Early terminate inter mode search based on sse, index 2. */
398 3, /*!< Early terminate inter mode search based on sse, index 3. */
400 4, /*!< Early terminate inter mode search based on sse, index 4. */
456 * is adjusted based on noise level of the current frame. The sf has three
459 * 1 and 2 : Reduce the number of frames based on noise level with varied
466 * 0: estimate bits based on a constant value;
467 * 1: estimate bits more accurately based on the frame complexity.
482 * 1: Conditionally allow motion estimation based on 4x4 sub-blocks variance.
531 // Skip motion estimation based on the precision of center MVs and the
545 // Prune starting mvs in TPL based on sad scores.
554 // Calculate rate and distortion based on Y plane only.
586 // Disable global motion estimation based on stats of previous frames in the
603 // Prune extended partition types search based on the current best partition
609 // Prune part4 based on block size
629 // Sets max square partition levels for this superblock based on
639 // Sets level of adjustment of variance-based partitioning during
649 // Thresholds for ML based partition search breakout.
652 // Aggressiveness levels for pruning split and rectangular partitions based on
654 // simple motion search based pruning. QIDX_BASED_AGG_LVL1 corresponds to
655 // qindex based and simple motion search based pruning.
701 // Prune extended partition search based on whether the split/rect partitions
708 // Prunt rectangular, AB and 4-way partition based on q index and block size
711 // 2 : prune all block size based on qindex
714 // Prune rectangular partitions based on 4x4 sub-block variance
716 // true : prune rectangular partitions based on 4x4 sub-block variance
726 // Prune rectangular partitions based on prediction mode chosen by NONE
735 // based on center angle of the prediction mode chosen by NONE partition. For
743 // For speed >= 7, variance-based logic is used to determine the partition
760 // 1 : pruning based on neighbour block information
764 // Prune rectangular split based on simple motion search split/no_split score.
815 // Skip evaluation of non-square partitions based on the corresponding NONE
832 // 0: disable, 1: if bsize >= BLOCK_32X32, 2: based on bsize, SAD and qp
869 // Reduce single motion search range based on MV result of prior ref_mv_idx.
883 // Accurate full pixel motion search based on TPL stats.
911 // Skips full pixel search based on start mv of prior ref_mv_idx.
934 // Bypass transform search based on skip rd at following stages
967 // Prune reference frames for single prediction modes based on temporal
984 // Skip the current ref_mv in NEW_MV mode based on mv, rate cost, etc.
994 // If the speed feature >= 3, skip the current ref_mv in NEW_MV mode based on
1001 // rd-cost based on mode rate and skip the mode evaluation
1004 // Flag used to control the ref_best_rd based gating for chroma
1011 // prune wedge and compound segment approximate rd evaluation based on
1015 // prune wedge and compound segment approximate rd evaluation based on
1062 // Based on previous ref_mv_idx search result, prune the following search.
1102 // Prune inter modes based on tpl stats
1111 // Model based breakout after interpolation filter search
1113 // 1: use model based rd breakout
1127 // COMPOUND_DIFFWTD based on the availability. Levels 0 to 3 indicate
1159 // Prune warped motion search based on block size.
1187 // skip sharp_filter evaluation based on regular and smooth filter rd for
1211 // Prune intra mode candidates based on source block histogram of gradient.
1217 // Prune intra mode candidates based on source block histogram of gradient.
1226 // Prune UV_SMOOTH_PRED mode for chroma based on chroma source variance.
1228 // true : Prune UV_SMOOTH_PRED mode based on chroma source variance
1239 // 1 : Evaluate applicable filter intra modes based on best intra mode so far
1272 // Prune chroma intra modes based on luma intra mode winner.
1355 // Prune level for tx_size_type search for inter based on rd model
1363 // Prune transform split/no_split eval based on residual properties. A value
1405 // Method 1: Based on block error, try using transform domain distortion for
1407 // Method 2: Based on block error, try to compute distortion in transform
1483 // 2: Disable for blocks which turn out to be transform skip (skipped based on
1486 // stage except NEWMV mode. For high quantizers, prune conservatively based on
1487 // transform skip (skipped based on eob) except for NEWMV mode.
1525 // Prune RESTORE_WIENER evaluation based on source variance
1531 // Prune self-guided loop restoration based on wiener search results
1533 // 1 : pruning based on rdcost ratio of RESTORE_WIENER and RESTORE_NONE
1534 // 2 : pruning based on winner restoration type among RESTORE_WIENER and
1565 // Estimate motion before calculating variance in variance-based partition
1576 // 3 - check merge and prune checking final split based on bsize and qindex
1583 // The heuristics selected are based on flags
1625 // For nonrd mode: use hybrid intra mode search for intra only frames based on
1657 // For nonrd mode: Prefer larger partition blks in variance based partitioning
1666 // indicates larger window. If set to 0, step_param is set based on internal
1703 // Forces larger partition blocks in variance based partitioning for intra
1707 // Use fixed partition for superblocks based on source_sad.
1718 // Intermediate termination of newMV mode evaluation based on so far best mode
1729 // Prune inter modes w.r.t golden or alt-ref frame based on sad
1732 // Prune inter mode search in rd path based on current block's temporal
1738 // 1-2: Reduce precision to halfpel, fullpel based on conservative
1746 // 2: Switch to halfpel, fullpel based on low block spatial-temporal
1750 // Prune intra mode evaluation in inter frames based on mv range.
1753 // based partitioning. The minimum values should be 7 to avoid left shifting
1757 // Qindex based variance partition threshold index, which determines
1764 // Enable GF refresh based on Q value.
1802 // Early terminate inter mode search based on sse in non-rd path.
1805 // SAD based adaptive altref selection
1811 // Level of aggressiveness for obtaining tx size based on qstep
1814 // Avoid the partitioning of a 16x16 block in variance based partitioning
1824 // A qindex threshold that determines whether to use qindex based CDEF filter
1835 // increase encoding time. Select threshold based on speed vs quality
1848 // Prune H_PRED during intra mode evaluation in the nonrd path based on best
1858 // Enable pruning of intra mode evaluations in nonrd path based on source
1869 // Prune intra mode evaluations in nonrd path based on best sad so far.
1893 // Sets force_zeromv_skip based on the source sad available. Aggressiveness
2023 * based on the passed in speed setting. (Higher speed gives lower
2037 * based on the passed in speed setting and frame size. (Higher speed
2050 * based on the passed in speed setting and current frame's Q index.