dex Otaku

I'm using a lot of links to those threads, and incorporating most of the original FAQ, and the last pass will be linking terms to wikipedia entries as glossary material. It should take a few days. I'm attributing it as co-written among us, rather than using quotations. Hope that's acceptable.

Lossless compression is difficult to manage at higher ratios in realtime. Considering the increased power requirements for processing that intense, and the relatively low savings in space available with realtime lossless compression, I'd say we'll be waiting a while before this practical to implement. Lossless compression of 1.5 - 2:1 ratios can be done at between 2-4x realtime on an Athlon XP 2500+ [barton, what my system uses] using nearly 100% CPU. I sincerely doubt that an LSI running from a 1.5V power supply will be able to manage this in the very near future. Also, PCM cannot store "about 2 CDs .. on each HiMD". A CD holds 74-80 minutes, a HiMD holds just short of 95 minutes.

With everyone's permission, I hope, I am working on integrating all of these along with information from the original HiMD FAQ and many notes from Wikipedia. I will be editing extensively, and I hope this doesn't offend anyone. Likewise, I will post it back here and expect the rest of you to review it and edit as needed. Hopefully we can make a single document out of this that addresses as much as possible.

1 - Yes, you should. Keep readin for the cautionary words. 2 - Yes, you should back up your dB first. I can't give any information regarding Realplayer, though, as I have never used it with a netMD [since I own a HiMD, and RP does not support HiMD yet] 3 - No precautions that I can suggest for the conversion tool. As it does not install anything that changes core OpenMG or SS modules, I would not worry about installing it. Also, I have never experienced any problems with the converter, myself. 4 - The conversion tool is for converting tracks uploaded from HiMD recordings to WAV files, so no, this is not for updating your OpentMG tracks. There is a utility for this that is included with SonicStage, however. I cannot attest to its function or reliability as I have never had a need to use it.

No. The audio functions of HiMD require the use of SonicStage, which is Windows-only. To my knowledge, users who have tried it with emulators have met limited success at best.

There was a need for these, so please do read this before posting support requests. The first suggestions usually given for software problems are these: * Update your operating system. Unless you have a very good reason not to [no net connection, known software incompatibility], apply those service packs. * Update your SonicStage, Simple Burner, HiMDRenderer, and Wave Converter to the most recent versions [note that Wave Converter is no longer required as a separate download asof SonicStage v3.0] When making a support request, be as thorough with your information as possible. The more you tell us, the more likely a solution can be found. Include as much tech info as you can provide, including but not limited to: * Operating system version and service pack level * SonicStage and Simple Burner versions, if relevant * Whether codec packs have been installed on the system [especially FFDSHOW] * the model of the recorder/downloader involved * Basic computer hardware info, including as a minimum: CPU type/speed; amount of RAM installed; USB interface version [1.0, 1.1, or 2.0] * List of applications that were running when the problem occurred, especially USB helper utilities for devices such as mice or cameras, antivirus app and their version, and anything else that loads on boot and is always open in the background [or the windows system tray, down by the clock in the corner] Please do contribute to this. While some users won't be able to provide all the information asked, the more that is provided, the easier it is to find solutions - as well as the common elements between seemingly-unrelated problems. Thanks, D.

I usually do transfers in batches of about 4 tracks. With previous versions of SS, it seemed as though the likelihood of lost tracks was higher if you ran the upload in large batches. This is also a good reason to keep your recording split up into shorter tracks. This way if one or more are lost, you only need to pull small segments out of the backup you've hopefully made beforehand. I still haven't had any problems with SS 2.3 uploading, though.

There are a number of processing methods used for things like this. The best ones tend to use well-established HRTF [head-related transform functions if I'm not mistaken]. My favourite system for multichannel sound, though I have never even had the opportunity to hear it on a real system, is Ambisonics [which has been around since the late 60s/early 70s]. For the most part, for headphone listening, I prefer simple binaural recordings, though.

Good suggestion. I have one thing to add: This means using extra CDRs, but backup all the WAV files [or use lossless compression on them and then back them up] to CDR as well. Audio CDs are not a backup medium; they play on CD player, yes - and you can rip them again later if needs be. Rips will basically never be 100% accurate, though. Data CDs are more robust in terms of error-correction, and a truly random-access, meaning you can accurately pick off any file/track you wish to without worrying about dropped sample from your drives read offsets, &c.

This is what I was talking about when I said "living with restrictions" .. The uploading and downloading of audio to/from HiMD can only be done with Sony's SonicStage. It's PC-only, and Windows-only. While HiMD does work as a USB mass-storage device, using audio with it is not a matter of drag & drop. I bought my HiMD fully understanding that there would be restrictions on what software could be used with it, &c. Since I was very limited budget-wise, it was the only choice available that would allow PCM recording in the price range I could afford, and I went for it. If you require drag & drop use, or the ability to transfer tracks in the clear with linux or a mac, HiMD is clearly not for you. If you are willing to accept that you have to use SonicStage under Windows, and can adapt to the limitations imposed by this [mostly as relate to digital rights management] then I would consider it as an option at the very least. Poke around the HiMD forum a bit, and, especially, skim the HiMD info subforum .. there's some extra info there on how transfers are done. In any case, good luck with your venture. Cheers.

Why not just play the MDs as they are in the HiMD player?

It will make no difference whatsoever, other than wasting additional time on the task. The processing chain is thus: MP3 file -> MP3 to PCM decoder -> PCM to ATRAC/3/plus encoder -> OMG file What you're talking about doing is just making several of the steps manual, which serves no purpose whatsoever. People [including myself] generally go the CD Image route if they come across MP3s that SS refuses to convert correctly by itself. It is also often the easiest route to go when converting from formats that are no supported by SS at all, such as FLAC, OGG, MPC, and the like. This method has the added advantage of automatically tagging tracks if you are doing whole albums, since SS's CDDB support will identify the tracks on a CD image just as readily as it will a CD. Incidentally, if you want better-quality encodes [from any source], copying via the optical in on your portable [which uses the hardware codec rather than SS's software codec] will do a better job of things, regardless of what recording mode you use. Also - to really get the best sound quality, use an original CD as the source. Chances are, if you're complaining about the quality of the MP3-atrac/3/plus transcodes, it's because you have been downloading or encoding mp3s with poor-quality settings. It pays to be both informed and discerning in both cases, i.e. download or encode at the highest quality you can find. After all, the rule is GIGO [garbage in, garbage out].

More than that: even iRiver themselves state that while the function is there, sort of, it's not intended for serious use. I would not touch the iRivers for recording purposes. What you want is either a HiMD [if you can live with Sony's software] or a flash-based recorder. Flash recorders of equivalent or better quality are more expensive than HiMD recorders, and the media are still significantly more expensive than MD or HiMD blanks. As far as dollar vs. performance goes, HiMD is the best option currently on the market. If you prefer to work restriction-free and don't care about paying more, look for a flash-based recorder such as those made by Marantz [professional products].

The best reasons to consider HiMD are the ability to upload your recordings digitally [no generation loss], and the ability to record directly in uncompressed linear PCM [meaning, again, no generation loss]. It sounds to me as though you'd be most interested in using a format like MD or HiMD exactly as I usually do - as an intermediary format only. I do not -store- anything on these discs, I simply record, upload, and erase. Oh, and back up and edit, of course. If you just need a starter rig, I'd consider a MZ-NHF800 or an MZ-NH900. I have the NH700 myself, which is basically the same as the NHF800 without a radio. Take a look on http://www.minidisc.org and stroll around the equipment browser for an idea of what the various models are. You'll be wanting to make sure that whatever you pick up has the following features, whether it's MD/HiMD or not: * a microphone input with plug-in power [powers condenser mics] * a line input [just in case you need it] * manual record-level controls [an absolute requirement for serious recording] * a decent AGC [auto-levelling] for when it's appropriate to use * the ability to mark tracks manually [it is more useful than it sounds] * a clear and readable record-level metre * uses commonly available batteries that you can find literally anywhere for cheap * control over mic sensitivity Things to look for that are great to have but not an absolute necessity for everyone: * PCM recording * the ability to either upload digitally or copy digitally by optical or coax * cheap removable media that are common and easy to find in most areas There are a number of things I could add to both lists, but these are the most important ones coming to mind at the moment. As far as a mic goes, for foley [sound effects recording] work I would suggest using a M/S stereo microphone. This is a compromise solution that I'd say is a good place to start. Sound reasoning behind this is: [no pun intended] * you can record stereo ambience quite well with them * they can mix down to a true mono signal * the "M" microphone element is directional, so a mono mixdown will usually give you a fairly "tight" recording that rejects sounds other than what you want to record Basically, M/S stereo mics tend to be the most versatile, IMO, in terms of being able to make both mono and stereo recordings with the same mic. Sony makes one called the ECM-MS907 which I have used for sound effects recording, and it works well. There are other mics out there that are of higher quality that can be found for the same amount of money, though; http://www.soundprofessionals.com has some models you might be interested in, including a 'copy' of the MS907 that outperforms it, though to be honest I'm not certain whether it's actually an M/S stereo mic or not [they don't actually specify]. I would avoid using "binaural" microphones for sound effects recording. They work great for ambience, great for bootlegging, and they can even be used as two clip-ons to record, say, two people during an interview; they are almost all omnidirectional though, which makes them [again, IMO] ill-suited for recording sound effects outside of a tightly-controlled environment [they simply pick up too much background sound to be really useful for this purpose]. Cheers.

Just a niggling fact here. That's DIGITAL transfer using SS, not analogue. The difference between digital transfer using SS and Total Recorder should be naught, technically. TR leaves merging tracks &c. up to the user, though. And lastly - all things considered, having the HiMD make track marks automatically actually works to your advantage even when you don't want it to do so. If your upload is completely successful, you can quite easily merge the tracks within SS, making this a total non-issue if you only use the HiMD as your recording medium. If your upload fails in parts, you lose less of the complete recording. If it were recording in all one large track, you would lose the entire thing from beginning to end. If you do digital backups first, you can use the successful SS tracks and merge them with the backup copy in the missing parts, or simply use the backup as your editing copy. Back up your backup first, of course. Thank you for the heads-up.

If you wish others to learn about or from it, you might consider including the reasons why [at least one of which I posted in your original thread]. This is theory which I haven't had the time to verify yet, but I strongly believe it may be correct: FFDSHOW contains filters for many audio and video formats. By default, the mp3 filter it uses [on my machine this is libmad] puts out >16-bit audio to the sound card's driver, which will then convert it to 16-bit if necessary. SS appears to have difficulty dealing with the output of the libmad filter, probably because it is sending out >16-bit audio. The advantage to FFDSHOW is that it allows near-complete control over what filters decode what streams for both audio and video within directshow applications. My current proposed solution to those users who wish to stick with FFDSHOW is to go into its audio config dialogue [in the start menu], scroll down to "Input/Output" and under "Supported output sample formats" deselect all but 16-bit integer. A second possible solution is to, again, open the FFDSHOW audio config and right at the top of the list, select "Codecs", then look for "MP3" and select "disabled" from the dropdown box. This will stop FFDSHOW's included filters from decoding MP3 at all. FFDSHOW is currently installed on my machine and does not interfere with SS in any way, shape, or form.

Huzzah. I forget zee miracles of daisy chaining.

You don't think step 6 is clear enough?

I disagree now.. having figured out what the problem possibly is, you were right to blow it away. If you want to learn how to mess with codecs/filters, reinstall it and mess around a bit. FFDSHOW is forgiving in that you can go in and reset things to the way they were. On the other hand, there are a crapload of controls there to play with, so if you're unsure of yourself - just leave things as they are. That old addage, "if it ain't broke, don't fix it," comes to mind. re: warbling - no idea. Does it play the same way through SS as on a minidisc? Perhaps [if SS is playing the mp3 from the library as it normally does with imported files] the mp3 codec you have installed is also already broken in some way. Weirdness: I'll try the FFDSHOW mp3 codec with it limited to 16 bit tomorrow. In the end, what caused your first problem might end up fixing the one you have now.

I digress. I installed the most recent version of FFDSHOW last night and it didn't install or alter any audio codecs at all. It does install filters for several formats, but as far as I can tell these should only be active when it's decoding video. In any case, my default FHG codec is what gets used for MP3s most of the time; the players I use tend to have their own decoders such as libmad built-in. I don't understand where the conflict comes in, but I do understand that I'm not experiencing it, and neither was I experiencing it before [with FFDSHOW installed then as well]. Mystery.. edit: a-HA! I'm a bloddy liar am I, I am! Removed FFDSHOW and reinstalled. It does have filters for AC3, DTS, mpeg audio, AAC, &c. - and I did let it do what it wanted with all of them. And I still don't have problems because of it. This doesn't get us any closer to the reasons why, but it does prove that I'm occasionally completely full of sh*t. edit #2: After spending an hour or two messing with codecs/filters.. FFDSHOW, by default, will override the decoding of most audio and video formats thrown at directshow players. [Ostensibly, the reason I dl'd it was to enable using mpeg4 video files in video editing applications that otherwise wouldn't recognise their 4cc codes, but that at least still doesn't work].. The interesting thing about FFDSHOW is that it lets you override anything you want with it. It puts icons for its video and audio control panels in the start menu. If you go into the audio filters and tell FFDSHOW to be 'disabled' for a particular format, it will simply pass it to the next filter that directshow would otherwise use. For instance: I have PowerDVD installed, and with it the Cyberlink AC3 filter [which is nice in that it has software Prologic II decoding, which means I can properly set the balance in my rear speakers, since the speakers/amp I have don't permit this.] After installing FFDSHOW it overrides the Cyberlink filter. However, by going in and selecting "disabled" beside a52/ac3 in its options, then closing/re-opening the player, it's back to using Cyberlink again. The same goes for mp3 decoding - disable it in the FFDSHOW dialogue and it goes to the next filter that the system would otherwise use, in my case, the FhG filter. The problem with FFDSHOW's mp3 filter? It's also using libmad, and is putting out 32- or 24-bit audio by default. This is probably what is screwing up SS. It's possible that by limiting its output [by going into the FFDSHOW audio config, scrolling down to "input/output" and unchecking all options but 16-bit integer] the whole issue with SS losing audio from mp3s might be fixed. I'll try this out tomorrow if someone reminds me. In the end I've left FFDSHOW on my machine for the simple reason that it lets me control what codecs get used, rather than letting the system decide. Another example here: I have the Koepi 1.1.0-beta1 XviD codec installed on my system, and it's told to decode XviD video. However - The Intervideo MPEG-4 filter installed by WinDVD overrides this, and there's no way to shut it off without uninstalling WinDVD. FFDSHOW, on the other hand, overrides the Intervideo filter, and also lets me manually choose the XviD 1.x filter I have installed instead. [Lesson learned: the intervideo filters are EEEEEEEEVIL, since they don't give you any options, don't let you override them, and don't show up as system codecs in any of the normal places] Yes, a great tangle. In the end - FFDSHOW gives you control over what codecs get used. The thing is, you have to want to have that control, and to have some idea of what to do with it..

Ah. That explains the sudden unexpected explosion of colour.

You can email them to me, and I can host them until kurisu can get them on the MDCF server.

I prefer to know the source of what I post, and to know it is legal. I have found my own photography on such sites, and do not approve of it.

I always have SYNC off, and my NH700 adds track marks during silent passages.

If anyone has a scanner or digicam and some audio equipment handy, the following pictures would be helpful to go along with this: * RCA outputs on the back of an a/v component * a 3.5mm stereo to dual-RCA cable * a 3.5mm stereo to 3.5mm stereo cable * a 3.5mm mono plug for comparison * 1/4" TS and TRS plugs * an XLR connector [female] * the back of a pro audio component with 1/4" and XLR jacks * a DIN connector and the DIN in/out on the back of a component Yep.