Avrin

A hundred or so MDs is probaly worth renting (or even buying) an RH1. It will save you a lot of time, transferring tracks at about 8x speed at their original quality, keeping all names and splits. Just don't forget to set SonicStage to import to PCM.

Currently recording Phil Spector's "Back to Mono" 4-CD box set to MONO on the RH10. As for codecs, don't forget that they consist of two independent parts - the encoder and the decoder. The manufacturer may remove the encoder, but still keep the decoder. That's exactly what is done in the RH10 - they removed the ATRAC3 encoder, but the unit is still able to decode ATRAC3. Similarly, they put an MP3 decoder in, so the unit plays MP3s (in whatever quality), but is unable to record (encode) to that format. And again, Hi-MD recorders contain only 64 kbit/s and 256 kbit/s encoders for ATRAC3plus, but many more (while still not all) decoders for the format. As for "Other setting" from the service manual, I’ve checked the entire group of addresses around 1113. The group (1110) is called ChkChg, and contains three addresses - 1111 Power, 1112 Servo, and 1113 SrvSys. The addresses actually contain much more capacity than others: In the NH600 address 1111 stores 41 two-byte hex values, address 1112 stores 13 two-byte hex values, and address 1113 stores 27 two-byte hex values. The values are mostly zeroes, though some are set. In the RH10 address 1111 stores 41 pairs of two-byte hex values, address 1112 stores 13 pairs of two-byte hex values, and address 1113 stores 27 pairs of two-byte hex values. Again, these are mostly zeroes, but some are set. Values specified in the service manual were not set in my unit (the respective pairs contained zeroes), setting them manually according to the manual changed nothing. In the RH1 each of these addresses stores 125 two-byte values, and most of them are actually programmed with non-zero values. All this doesn't look to me like power or charge calibration settings at all. I looks more like some microcode or patches written to the units. But to understand those, one needs to know the assembly language (machine code) used for such units. As for recording to MONO on the RH1 - the commercial version doesn't do it, and it is not specified in the manual. Only SP, LP2, and LP4 are available. But the prototype with firmware version 0.960 was actually able to record in MONO. Probably that's how MONO got to the specification page.

It just occured to me, why exactly SONY had to keep the ATRAC encoder, whatever the implementation (hardware or firmware). It is required for pseudo-SP downloads. The PC sends ATRAC3-encoded files, and they are decoded and encoded to ATRAC on the unit. Removing the ATRAC encoder would have made even pseudo-SP downloads impossible. And MONO obviously works since it uses the same encoder as SP, with two channels merged into one. But why SONY disabled MONO recording on the commercial version of the RH1 is beyond me.

More probably Distribution Flag Low and Distribution Flag High. This may be exactly the case. The non-volatile memory used to store the firmware is not changed, while the RAM used for data processing may be used differently. E.g., the first byte at address 0114 controls what is actually loaded - the MP3 decoder or MD recording routines. While the second byte at address 1821 simply controls the interface.

And made a horrible typo (already corrected). Sorry. This is easy. The following addresses are linked together, so setting a value at one of them also sets the same value at the other. Address 0113 = Address 1811 (DistFL) Address 0114 = Address 1812 (DistFH) --- Overall, I think the optimal configuration is as follows: 1811: 23 - no output signal compression, full titling options (you already have this value if your unit is manufactured for the Japanese market). 1812: 7F - one step down from the original value, to enable ATRAC recording at the cost of MP3 playback. 1821: 97 - one step down from the original value, to enable SP/MONO selection and normal recording start. 1822: 14 - the original value. --- And now I wonder what these "hardware flag" addresses (1831-1834) are for?

A quick analysis of common service mode setting patterns in the RH1 revealed nothing that is not known already.

So, the more or less useless hacking results are here. Many thanks for inspuration to TC- here and VovchikRH10 at www.player.ru/talk , who started all that. A few comments for a start. SONY disabled some functions on the RH10 menu for a reason. The unit's hardware/firmware does not provide full or any support for them. Legacy MD recording works only for SP and MONO (LP2 and LP4 are recorded as silence), V-Surround does not surround, but just increases the sound level, and SpeedControl has no effect at all. Still, the meaning of addresses 1821 and 1822: Address 1821: First byte: 2, 3, 6, 7, A, B, E, F: The V-Surround item appears in the Sound menu. 0, 1, 4, 5, 8, 9, C, D: The V-Surround item is not shown in the Sound menu. Second byte: 2, 3, 6, 7, A, B, E, F: If the first byte at address 0114 (or 1812) is set to 0-7: the REC Mode selector in the REC Settings menu works for MD mode. MD mode recording can be started with the unit on. Only SP and MONO are recorded properly. If the first byte at address 0114 (or 1812) is set to 8-F: no effect. 0, 1, 4, 5, 8, 9, C, D: The REC Mode selector in the REC Settings menu doesn't work for MD mode. If the first byte at address 0114 (or 1812) is set to 0-7: MD mode recording can only be started with the unit off. Recording is done in SP or the mode previously selected. Only SP and MONO are recorded properly. If the first byte at address 0114 (or 1812) is set to 8-F: no effect. Address 1822: First byte: Unknown effect. Second byte: 1, 3, 5, 7: The SpeedControl item is shown in the menu, but has no effect on playback. 0, 2, 4, 6: The SpeedControl item is not shown. And now to torturing the RH1.

Work in progress: Seems like no "magic combinations" exist. Each byte at each address is responsible for certain functions.

To hell with the manual. This topic is dedicated to disobeying it. The unit records perfectly in SP and MONO. The entire problem seems to be caused by the limited amount of memory for the firmware. SONY needed to put MP3 playback code in, so something had to be sacrificed. And they chose to sacrifice MDLP encoding to make room for [crippled] MP3 decoding. While the ATRAC SP codec was probably hardware-based, so it remained. But to avoid confusion, they completely disabled all legacy recording modes through service mode settings. Later, in the RH1, they simply added more memory, so it has all you may ever want. But the reason for disabling the MONO recording ability (present in the prototype) needs to be investigated.

Work in progress: Addresses 1821 and 1822 are the key. Copying values from the RH1 enables lotsa stuff, but: Looks like there is no ATRAC3 encoding circuitry in the RH10. Recorded LP2 and LP4 tracks play as silence. MONO is OK. V-Surround changes the sound, but whether it does it properly - remains to be seen. SpeedControl doesn't work.

An interesting point with the RH10 is paragraph 3-5-3 of its service manual, quite vaguely called "Other setting". It describes a procedure of programming address 1113. The procedure is very different from the usual value changing. Usually, an address contains a single value (e.g., 23 (Japan) or A0 (Europe) at address 0113). But address 1113 is successively programmed by entering six (!) different values in groups of three. And no explanation whatsoever is given about the meaning of these values. No name for the address itself is displayed even with the remote connected. The address itself seems to belong to the ChkChg group (1110). But the Charge Function Check is fully described in paragraph 3-3, and contains no references to address 1113 (addresses from group 2000 POWER are used for charge check and calibration).

Recording level is actually saved without any hacks. But the unit switches to Auto Level mode as soon as you press the STOP button. If you switch back to Manual, the previously set level will be used, even after switching the unit off, removing discs, etc. I mostly use the RH10 to record to SP via optical, so I have the level set to 23 (pass through). But each time I record a disc, I need to go through the DISABLE SYNC REC > ENABLE MANUAL LEVEL > ENABLE SYNC REC routine. If you just record from a mic or line in, and don't need SYNC REC, the entire procedure is much simpler. And I don't think it's possible to prevent the unit from switching to Auto Level without modifying the firmware itself (not just by changing service mode settings). Same may apply to enabling full legacy MD functionality.

Well, the VPT trick was not my idea. It's from a Russian forum. Easily discovered by looking at the service manual for the NH700/NHF800.

Well, I'd better stop messing around with this service mode stuff. Just made one of my NH600s able to title in Katakana on the unit and show Japanese menus on the RM-MC40ELK. Nice for an absolutely non-Japanese unit, heh? Just by setting address 0113 to 23. And it already had VPT enabled by resoldering.

Some freshly discovered data on some of the effects of address 0113 and 0114 settings in the RH10 for those who are interested: Address 0113: First byte: 0 to 7: Custom equalizers enabled; [presumably] full output power. 8 to F: Custom equalizers disabled; [presumably] limited output power. Second byte: 0, 1, 8, 9: Menu languages: English, French, German, Italian, Spanish; MD & Hi-MD titling: English. 2, 3, A, B: Menu languages: Japanese, English; MD & Hi-MD titling: Katakana, English. 4, 5, C, D: Menu languages: Japanese, English, French, German, Italian, Spanish, Simplified Chinese, Traditional Chinese; MD & Hi-MD titling: English. 6, 7, E, F: Menu languages: Japanese, English, French, German, Italian, Spanish; MD titling: English; Hi-MD titling: CANNOT EDIT. Address 0114: First byte: 0 to 7: No MP3 playback or download; SP is displayed when the unit is in MD mode; SP recording can be started when the unit is off. 8 to F: MP3 playback and download; SP not displayed; no SP recording. Second byte: Unknown effect. This is by no means exhaustive.

Surely it won't work 100%, but it allows you to get as close to gapless as possible with the existing hardware/firmware.

I afraid that's hardly possible.

At least, you may be interested in the fact that such files also provide "spliced tape" playback on the RH1 and the unhacked RH10.

So, it looks like ATRAC heavyweights here are not happy with modern SONY players because of the lack of proper gapless playback. Pink Floyd fans are the unhappiest of them all. Let's see what can be done to amend the situation while SONY sleeps. Let's look back at ATRAC and what makes it gapless. The main factors are as follows: 1. Constant bitrate, which facilitates synchronization between tracks. 2. Continuous encoding - successive tracks are parts of a single gapless stream. 3. [Possibly] Use of internal timing tags. All this is easily achievable with MP3 using a command-line version of LAME. While SONY players do not use any timing tags for MP3, continuous encoding at a constant bitrate will certainly facilitate playback, and, while not fully eliminating gaps, will make them as short as possible. Depending on material, most of them will be unnoticeable, while others may sound like spliced tape, without affecting the rhythm. At least that's how they sound on my new SONY NWZ-B143F. So, the "magic" command for LAME is as follows: LAME -b 256 --cbr -q 0 -m s --nogap "Track 01.wav" "Track 02.wav" "Track 03.wav" "Track 04.wav" "Track 05.wav" "Track 06.wav" ... "Track nn.wav"[/CODE] What do all these switches do? Let's see: [b]-b 256[/b] instructs the encoder to use a constant bitrate of 256 kbit/s. You may specify any numeric value supported by your player. [b]--cbr[/b] enforces strictly constant bitrate. [b]-q 0[/b] sets the highest possible encoding quality. [b]-m s[/b] instructs the encoder to use true (not joint) stereo. [b]--nogap[/b] instructs the encoder to create a single stream split by files. After the switches, you need to specify all files to be encoded one after another, separated by spaces (using quotes for names containing spaces). Obviously, encoding files separately will not help us achieve our goal. The end result may depend on the player model. Some models may still introduce larger gaps. But at least it's worth a try. At least this makes it possible to [i]enjoy[/i] Pink Floyd's "The Wall" and "The Final Cut" on an NWZ-B143F.

The BASS feature is only useful with the supplied mediocre-quality 'phones. If you connect good 'phones, you won't need it.

It doesn't look like there is a way to skip albums. Reminds me of the MZ-RH1 without a remote. Enabling SHUFFLE may help to get to the desired album sooner than by simply pressing the FORWARD button lotsa times. After a song from the album starts, disable SHUFFLE, and press BACK several time to skip to the beginning of the album.

Excellent!

A nice way is to use LAME.EXE with the "--nogap" option to encode music. In this case you need to specify all files in a single line. E.g.: LAME --preset insane -q 2 --nogap "Track 01.wav" "Track 02.wav" "Track 03.wav" "Track 04.wav" "Track 05.wav" "Track 06.wav" "Track 07.wav"[/CODE] This will not give you perfectly gapless playback on SONY devices, but gaps will be almost unnoticeable. On my NWZ-B143F they are about 0.01 secs long when tested using stereo white noise (the hardest possible material to encode/decode).

I'm terribly sorry to confuse you. Mea culpa. It's not 0 to 127 for the above addresses (although this range may apply to some other ones, in hex representation). It's 00 to FF hex. So, 256 possible values for each, and 65,536 combinations for the two addresses.

I'm not good at chess at all.