Helix MP3 Encoder - Revision history

Maikmerten: /* Encoder switches */ Document -L switch

2024-06-11T17:24:32Z

Encoder switches: Document -L switch

← Older revision		Revision as of 17:24, 11 June 2024
Line 65:		Line 65:
	High frequencies will only be encoded if -V >= 80 or -B >= 96.		High frequencies will only be encoded if -V >= 80 or -B >= 96.
	\| HF2 for unrestricted high-frequency encoding		\| HF2 for unrestricted high-frequency encoding
			\|-
			\| -L \|\| Set ''per-channel'' bitrate limit (upper cap) in VBR mode. Defaults to no limit. \|\| -L80 to limit maximum VBR bitrate to 160 kbps for stereo files
	\|-		\|-
	\| -M \|\| Stereo-mode/Mono selection. 0: stereo, 1: M/S stereo (default), 2: dual channel, 3: mono \|\| -M3 to downmix to mono		\| -M \|\| Stereo-mode/Mono selection. 0: stereo, 1: M/S stereo (default), 2: dual channel, 3: mono \|\| -M3 to downmix to mono

Maikmerten: /* Encoder switches */ Document -T switch

2024-04-20T12:45:18Z

Encoder switches: Document -T switch

← Older revision		Revision as of 12:45, 20 April 2024
Line 71:		Line 71:
	\|-		\|-
	\| -SBT \|\| Threshold for short-block decisions. ''Lower'' values mean ''more'' short-block usage. Default is 700. \|\| -SBT500 for more short-blocks (more responsive to transients, might increase bitrate in VBR)		\| -SBT \|\| Threshold for short-block decisions. ''Lower'' values mean ''more'' short-block usage. Default is 700. \|\| -SBT500 for more short-blocks (more responsive to transients, might increase bitrate in VBR)
			\|-
			\| -T \|\| Bias for VBR quality scale. Ranges from -40 to 50, defaults to 0. In conjunction with the base VBR quality scale (-V, see below) this determines the bitrate and quality of VBR encodings. Can be used to reach bitrates usually not achievable with the base VBR quality setting alone. \|\| -V0 -T-40 for a lower than usually possible VBR bitrate
	\|-		\|-
	\| -U		\| -U

Maikmerten: /* -HF switch */ spectographs of HF1 and HF2 and description of encoding behavior

2024-04-20T08:35:32Z

-HF switch: spectographs of HF1 and HF2 and description of encoding behavior

← Older revision		Revision as of 08:35, 20 April 2024
Line 86:		Line 86:
	=== -HF switch ===		=== -HF switch ===

	Due to ta design decision in the MP3 ~~bitstream~~ format, encoding high-frequency content (frequencies beyond 16 kHz) can be comparatively costly in terms of bits, while these frequencies often do not contribute much to the perceived audio quality. For this reason, the Helix MP3 Encoder will not encode high-frequency content by default. This, however, can be controlled by the -HF switch.		Due to a design decision in the MP3 format, encoding high-frequency content (frequencies beyond 16 kHz) can be comparatively costly in terms of bits, while these frequencies often do not contribute much to the perceived audio quality. For this reason, the Helix MP3 Encoder will not encode high-frequency content by default. This, however, can be controlled by the -HF switch.

			With -HF1, only Mid-Side stereo (M/S-stereo) frames will contain high-frequency content, while other frame-types will not encode any high-frequency content. The encoder will automatically choose the most efficient stereo mode for each frame. Usually, around 80% of frames will be coded as M/S-stereo ("joint" stereo encoding), with 20% of frames being coded as simple-stereo (separate encoding of channels). With -HF1, the ~20% of frames not being encoded in M/S-stereo will not encode high-frequency content, which might be perceivable as "dropouts" in the high frequency bands in unfavorable conditions.

			With -HF2, all frame types can encode high-frequency content, so no "dropouts" should occur.

			The behavior of -HF1 and -HF2 can be observed in the following spectrographs:

			-HF1: [[File:Helix-pink-noise-v110-hf1.png\|384px]]
			-HF2: [[File:Helix-pink-noise-v110-hf2.png\|384px]]

			The rationale for having the -HF1 option in the first place may be that if M/S-stereo is not applicable for a given frame, many bits are already spent on encoding the stereo image with separate stereo encoding. If the bitrate cannot be increased (for instance, when doing CBR encoding), it might be preferable to spend the remaining bits on the usually more perceivable lower frequencies, omitting high-frequency content. This rationale appears less applicable to VBR encoding (as the encoder can choose a more fitting bit allocation) and thus using -HF1 in VBR encoding might not be recommendable.

	=== Magical Mystery Switch -TX ===		=== Magical Mystery Switch -TX ===

Maikmerten: /* Encoder switches */ new section regarding the -HF switch

2024-04-20T07:50:54Z

Encoder switches: new section regarding the -HF switch

← Older revision		Revision as of 07:50, 20 April 2024
Line 83:		Line 83:
	\| -X \|\| Control writing of Xing/LAME header information. 0: No headers, 1: only basic Xing information header, 2: Xing header with VBR-TOC and LAME header (gapless information) (default) \|\| -X0 to disable headers (in very rare cases of incompatibility)		\| -X \|\| Control writing of Xing/LAME header information. 0: No headers, 1: only basic Xing information header, 2: Xing header with VBR-TOC and LAME header (gapless information) (default) \|\| -X0 to disable headers (in very rare cases of incompatibility)
	\|}		\|}

			=== -HF switch ===

			Due to ta design decision in the MP3 bitstream format, encoding high-frequency content (frequencies beyond 16 kHz) can be comparatively costly in terms of bits, while these frequencies often do not contribute much to the perceived audio quality. For this reason, the Helix MP3 Encoder will not encode high-frequency content by default. This, however, can be controlled by the -HF switch.

	=== Magical Mystery Switch -TX ===		=== Magical Mystery Switch -TX ===

Maikmerten: /* Encoder switches */ Extend description for -HF

2024-04-08T12:49:52Z

Encoder switches: Extend description for -HF

← Older revision		Revision as of 12:49, 8 April 2024
Line 57:		Line 57:
	\|-		\|-
	\| -HF		\| -HF
	\| Controls encoding of high frequency content (> 16 kHz). Disabled by default. Valid values are 0 (disabled), 1 (partial, only "mode-1 granules"), 2 (full, "all granules"). Note that high-frequency content will only be encoded if the psychoacoustic model deems encoding high frequencies as beneficial for the given bitrate/quality settings.		\| Controls encoding of high frequency content (> 16 kHz). Disabled by default. Valid values are 0 (disabled), 1 (partial, only "mode-1 granules"), 2 (full, "all granules").

			A value of "1" means that only frames with M/S stereo can include high frequency content. A value of "2" allows high-frequency content in all frame-types.

			Note that high-frequency content will only be encoded if the psychoacoustic model deems encoding high frequencies as beneficial for the given bitrate/quality settings.

	High frequencies will only be encoded if -V >= 80 or -B >= 96.		High frequencies will only be encoded if -V >= 80 or -B >= 96.

Maikmerten: /* Magical Mystery Switch -TX */ typo

2024-03-29T12:21:31Z

Magical Mystery Switch -TX: typo

← Older revision		Revision as of 12:21, 29 March 2024
Line 82:		Line 82:
	=== Magical Mystery Switch -TX ===		=== Magical Mystery Switch -TX ===

	During encoder-development, apparently different experimental settings were exposed for easy tweaking during testing. The [https://github.com/maikmerten/hmp3/blob/0e895bf47fdb0de3093622465031c8dbc4acc0b6/hmp3/src/pub/bitallo.h#L65 BA_CONTROL data structure] includes the wonderfully named int variables "test1", "test2" and "test3". While "test2" and "test3" ~~and~~ unused, "test1" is used in in the function [https://github.com/maikmerten/hmp3/blob/0e895bf47fdb0de3093622465031c8dbc4acc0b6/hmp3/src/bitallo3.cpp#L1069 startup_adjustNT1B of bitallo3.cpp]. The value of "test1" can be set via the -TX command line parameter. The default value is 6.		During encoder-development, apparently different experimental settings were exposed for easy tweaking during testing. The [https://github.com/maikmerten/hmp3/blob/0e895bf47fdb0de3093622465031c8dbc4acc0b6/hmp3/src/pub/bitallo.h#L65 BA_CONTROL data structure] includes the wonderfully named int variables "test1", "test2" and "test3". While "test2" and "test3" are unused, "test1" is used in in the function [https://github.com/maikmerten/hmp3/blob/0e895bf47fdb0de3093622465031c8dbc4acc0b6/hmp3/src/bitallo3.cpp#L1069 startup_adjustNT1B of bitallo3.cpp]. The value of "test1" can be set via the -TX command line parameter. The default value is 6.

	The function startup_adjustNT1B adjusts (increases) the per-band "Noise Target" (NT). The value of "test1" provides a multiplier "f", which factors in how much the noise target is adjusted. A [https://github.com/maikmerten/hmp3/blob/0e895bf47fdb0de3093622465031c8dbc4acc0b6/hmp3/src/bitallo3.cpp#L1075 per-band threshold table] decides which bands receive adjustment.		The function startup_adjustNT1B adjusts (increases) the per-band "Noise Target" (NT). The value of "test1" provides a multiplier "f", which factors in how much the noise target is adjusted. A [https://github.com/maikmerten/hmp3/blob/0e895bf47fdb0de3093622465031c8dbc4acc0b6/hmp3/src/bitallo3.cpp#L1075 per-band threshold table] decides which bands receive adjustment.

Maikmerten: /* Encoder switches */ Some words on -TX

2024-03-29T12:18:12Z

Encoder switches: Some words on -TX

← Older revision		Revision as of 12:18, 29 March 2024
Line 79:		Line 79:
	\| -X \|\| Control writing of Xing/LAME header information. 0: No headers, 1: only basic Xing information header, 2: Xing header with VBR-TOC and LAME header (gapless information) (default) \|\| -X0 to disable headers (in very rare cases of incompatibility)		\| -X \|\| Control writing of Xing/LAME header information. 0: No headers, 1: only basic Xing information header, 2: Xing header with VBR-TOC and LAME header (gapless information) (default) \|\| -X0 to disable headers (in very rare cases of incompatibility)
	\|}		\|}

			=== Magical Mystery Switch -TX ===

			During encoder-development, apparently different experimental settings were exposed for easy tweaking during testing. The [https://github.com/maikmerten/hmp3/blob/0e895bf47fdb0de3093622465031c8dbc4acc0b6/hmp3/src/pub/bitallo.h#L65 BA_CONTROL data structure] includes the wonderfully named int variables "test1", "test2" and "test3". While "test2" and "test3" and unused, "test1" is used in in the function [https://github.com/maikmerten/hmp3/blob/0e895bf47fdb0de3093622465031c8dbc4acc0b6/hmp3/src/bitallo3.cpp#L1069 startup_adjustNT1B of bitallo3.cpp]. The value of "test1" can be set via the -TX command line parameter. The default value is 6.

			The function startup_adjustNT1B adjusts (increases) the per-band "Noise Target" (NT). The value of "test1" provides a multiplier "f", which factors in how much the noise target is adjusted. A [https://github.com/maikmerten/hmp3/blob/0e895bf47fdb0de3093622465031c8dbc4acc0b6/hmp3/src/bitallo3.cpp#L1075 per-band threshold table] decides which bands receive adjustment.

			From code comments, it appears that the default value of 6 was determined to generally work best. It's unclear in what situations modifying this value is beneficial.

	== Reasonable Settings ==		== Reasonable Settings ==

Maikmerten: /* Reasonable Settings */ remove superfluous low-pass parameters

2024-03-26T06:46:56Z

Reasonable Settings: remove superfluous low-pass parameters

← Older revision		Revision as of 06:46, 26 March 2024
Line 90:		Line 90:
	\|-		\|-
	\|		\|
	~~-F24000~~ -HF2 -V150		-HF2 -V150
	\| ~ 256 kbps		\| ~ 256 kbps
	\| Maximum quality VBR encoding, with full audio spectrum. (ca. LAME -V 0)		\| Maximum quality VBR encoding, with full audio spectrum. (ca. LAME -V 0)
Line 108:		Line 108:
	\|-		\|-
	\|		\|
	~~-F16000~~ -V50		-V50
	\| ~ 128 kbps		\| ~ 128 kbps
	\|		\|

Maikmerten: /* Encoder switches */ update to -U switch

2024-03-25T17:47:00Z

Encoder switches: update to -U switch

← Older revision		Revision as of 17:47, 25 March 2024
Line 70:		Line 70:
	\| -U		\| -U
	\|		\|
	Select assembly optimizations. 0: ~~No assembly~~ optimizations, 1: unused (was supposed to be AMD's 3Dnow!) 2: use SSE assembly optimizations (Intel Pentium 3).		Select assembly optimizations. 0: Only generic optimizations, 1: unused (was supposed to be AMD's 3Dnow!) 2: use SSE assembly optimizations (Intel Pentium 3).

	This only has an effect if the encoder is compiled with Visual Studio (up to version 2015) for 32-bit Windows. No effect if the encoder is compiled for Linux, 64-bit Windows or, e.g., ARM processors.		This only has an effect if the encoder is compiled with Visual Studio (up to version 2015) for 32-bit Windows. No effect if the encoder is compiled for Linux, 64-bit Windows or, e.g., ARM processors.

Maikmerten: /* Assembly optimization */ minor typo

2024-03-25T17:43:04Z

Assembly optimization: minor typo

← Older revision		Revision as of 17:43, 25 March 2024
Line 160:		Line 160:


	=== Assembly ~~optimization~~ ===		=== Assembly optimizations ===

	In platform/win/i386, there are optimized assembly versions of speed-critical routines. These target 32-bit x86-CPUs in general (but usually optimized for the Pentium 2), with some routines also being available for SSE, targeting the Pentium 3 (assembly files starting with "x"). These SSE optimizations can be selected via the -U parameter.		In platform/win/i386, there are optimized assembly versions of speed-critical routines. These target 32-bit x86-CPUs in general (but usually optimized for the Pentium 2), with some routines also being available for SSE, targeting the Pentium 3 (assembly files starting with "x"). These SSE optimizations can be selected via the -U parameter.