45 replies to this topic

[Talisin]

Hi to everyone!

I come here to ask for your help. Well, I know actually what I'm gonna ask is something I must decide to, since is a personal thing, but maybe hearing (reading) your opinions will help me.

My concern is 'bout which atrac codec use. Maybe I have to explain a little more before ask my question: At last, after a loooooong time, I have all my music ready to be transfered to MD. I mean, I have lots of MDs with music, since is my daily format, but all in chaotic order, with albums repeated, etc. and now I have put in order all my cds, casettes, music files, etc to record MDs, print covers, etc.

I always thought to record in SP mode via optical output from my computer or from a DVD player, so I could hear my MDS in my net-MD units as in my Hi-MD units, and convert my favourite songs or albums to atrac3+ 256 in order to fill my Hi-MD discs for journeys and so. But reading these days older and not so older posts here I'm hesitating between proceding this way or choose another: all my music to atrac3+ 352 codec. I believe this codec is better for quality music, but can't use it in my net-md recorders.

So here comes my question: is really a3+ 352 so worthing over sp to take the second path? If so, what do you think I could do with my net-md units? Selling them is out of question, since they were my first units and have sentimental strenght to me, but maybe some of you could give me (and others) smart and innovative ideas for using it.

Or maybe I must stand in my first way and use sp as main format?. If so, do you think qhimd linux software will ever allow to transfer sp files from pc to md and back through usb (without a RH1)? I have read what this software can do now, and have read sp transfers from MD to PC are only analogue ones, but don't know if there is technical possibility to make this transfer digital someday.

I have to add that MD is my main (and almost only) way to hear music this days, since I live in a very small apartment, without space to a HI-Fi, and using my laptop to hear music is a mess (again, space problem), so, at home, I use my MD units joined to a couple of small but good quality speakers. And when I'm out, at leat for now, only use MD (I'm starting to hink to buy an ipod touch, not for music, but other concerns, but if I do it, maybe could use it outside so not having to carry two gadgets, since when I hear music outside, I'm not in quier rooms).

Well, I know these are very personal choices, and even have made blind tries between sp and 352, but still can't choose, and also my net-md units fate is making me hesitate. I hope, at least, my question will start an interesting conversation.

Thanks a lot in advance, and best wishes to all of you.

PS: as always, sorry for my english XD

[PhilippeC]

I suppose you will record your favorite songs from original LPs or CDs
a classic 80mn MD formatted in Hi-MD will allow you to record
- 1h35' in 352kbps = a better quality sound but only plus 15' of music
- 2h20' in 256kbps = a small lack in quality sound (I can't hear any difference from a SP recording) but a huge gain in total recording time
For sure, you will lose compatibilty with Net-MD units.
As it is to listen to music with your Hi-MD unit outside your appartement, why not.
It is more natural to use a 1Go Hi-MD disc with a Hi-MD unit, total recording time is important
- 5h30' in 352kbps
- 7h55' in 256kbps
If you don't have a MZ-RH1, no way to upload your music from your SP recordings. I don't know about Linux and any possibility to upload SP recordings with a Net-MD unit.

Where are you living ?

[Talisin]

Hi PhilippeC. First of all, thanks for your answer.

Actually, recording my favorite songs is equal (in a lot of albums) to record all songs from these albums (I choose albums sooooooo minuciosly ).

So you take the atrac3plus path, don't you? Any innovative advice about what to do with net-md units?

Do you think 256 is close enough to 352 in sound quality so bump in space is worth? I mean, I have made some blind tests, but really didn't get absolute results. My main concern is, since they were just 3 or 4 tests, with a more long term music hearing, I will notice differences. If answer to my question is yes, then, as you say, I will gain a lot of space, so I could record in each 80 minutes Md not only an album, but also (as I intended at first time) live versions and the same.

Oh, and I live in Malaga, Spain (I'm from Seville, also Spain). You're from France, aren't you?

[sfbp]

just to inject my $0.02 worth of advice, which you can easily disregard.

The most compatible format of all is LP2. This works on all NetMD players, and can be used on HiMD (you get 15 or 16 hours on a single HiMD 1GB disk). It is still 24-bit ATRAC so you get most of the benefits of SP with the additional enhancements of Type-S on all HiMD units (and some NetMD units) for playback, which IMHO improves the sound so much you really don't notice a lot of difference (yes there is SOME difference) from ordinary SP.

Except for LIVE recordings I would choose this format. For HiMD conversion from CD's I typically convert to AAL 256kbps) which can convert to anything with little degradation, but usually for simplicity I simply transfer to HiMD - without further bitrate converting. This leaves the :lossless: part behind (on the PC) and gives you ordinary 256kbps (Hi-SP) on the MD.

All the formats are 24-bit, except when you convert to WAV on the PC from one of the atrac formats. For SP, this conversion is OK. For the other formats, I feel it tends to be a bit noticeable in the final result.

Finally, Sony has somehow figured out MP3 conversion (with their MP3CONVERSION.EXE tool) that as long as you always choose 256kbps for the MP3, the results are quite decent. I just converted a bunch of files for my dad, and was pleasantly surprised how good the MP3 results were. Most MP3s however are not nearly that good. The key is to keep a high bitrate, even for 66 and 132 kbps ATRAC sources.

[PhilippeC]

Never thought to put LP2 Atrac3 132kbps on a Hi-MD formatted disk. Net-MD units cannot read the discs. Atrac3+ 192kbps is maybe a little bit better in sound quality than LP2 + type S ?

[Talisin]

HI Stephen. Dont worry, your 2cents always seem interesting to me

You have put me out with your opinion: I hadn't even considered lp2. When I buyed my first md unit, a N710, I tried it, but didn't satisfied me. With heavy metal songs, bass sounded with no strength, and with more sweet sounds, Loreena McKennitt albums, for example, it sounded...don't know how to say it...maybe restricted? As if something was missed from the sound. So my problem with lp2 wasn't artifacts everybody talk about, but something more...personal? Sorry if my english is not good enough to express me better. But maybe is something I was doing wrong if you talk about it being close to sp. Any advice to try it in the right way?

What disturb me is you reccommend to me using lp2 but you say you use aal256. Do you use both or just the second one?. I have read you in other threads on the aal theme, but something still isn't clear to me: if you rip a cd, for example, to aal256, what you get to transfer to a Hi-MD is the 256 part, so lossless section is just for play from pc, of is it also useful for reencoding to another bitrate?.

Oh, and on mp3...I think I have almost no one of them, since my music is on cassettes, CDs or flac files, but you've interested me: that mp3 conversion tool you talk about, is part of sonicstage? Is the service this software uses when you add to library a mp3 and convert it to atrac format? Or is a totally apart software?.

On 24 bits atrac...sorry, my knowledge about sound is low, so don't know what you're talking about :S

Anyway, thanks a lot for your help.

[sfbp]

my problem with lp2 wasn't artifacts everybody talk about, but something more...personal? Sorry if my english is not good enough to express me better. But maybe is something I was doing wrong if you talk about it being close to sp. Any advice to try it in the right way?

I would try Simple Burner, for a start. And remember to use Type-S playback I also initially rejected LP2 but that was because I went about it the wrong way. I ripped to WAV and then converted. This tends to produce garbage.

What disturb me is you reccommend to me using lp2 but you say you use aal256. Do you use both or just the second one?

Sorry I wasn't clear. I mean that lp2 is the most compatible playback format. But I rip to AAL on the hard disk, which if need be can be converted to LP2 or any other format.

. I have read you in other threads on the aal theme, but something still isn't clear to me: if you rip a cd, for example, to aal256, what you get to transfer to a Hi-MD is the 256 part, so lossless section is just for play from pc, of is it also useful for reencoding to another bitrate?

The PC never plays back the lossless part. The only point of it is for when you want to convert to some other format, or maybe edit it (I am not certain, but I assume Sound Forge WILL use the lossless part to rebuild the internal wave when you edit one of these files).

Oh, and on mp3...I think I have almost no one of them, since my music is on cassettes, CDs or flac files, but you've interested me: that mp3 conversion tool you talk about, is part of sonicstage? Is the service this software uses when you add to library a mp3 and convert it to atrac format? Or is a totally apart software?

It's a standalone program from Sony, though I have a strong suspicion SonicStage needs to be installed before you run it. Once it is run, of course the MP3 files are totally standard and can be used anywhere.

On 24 bits atrac...sorry, my knowledge about sound is low, so don't know what you're talking about :S

The real "secret" of ATRAC is that it is in many ways BETTER than standard 16bit 44.1Khz PCM (CD format) because it actually has 24-bit resolution on all samples, even at the lowest data rate. It does this by making the magnitude a floating (instead of fixed) point number. This will tend to avoid many of the weird artifacts in other formats (have you ever tried converting MP3 back to ATRAC?).

[paperclip]

I would get Sonic Stage running on that laptop and just use LP2. You probably won't be able to tell any difference with the speaker setup you have. I find Sonic Stage burns very, very good on protables used that way and with headphones. In fact I had to play the LP2 dics on my large home stereo to notice any difference. I rip all my CD's into wav files, then I can trancode to anything from those. The only drawback being not having any tags. I do use AAL for some things and it's a great codec: tags; can be uncompressed to wav with no loss, for use by other programs, space savings.
Let me know what you decide on, I think if you take the time to tag the files in Sonic Stage (and use it for playback) it is an easy way to burn MiniDiscs. There are dynamic playlists that are quite powerful and the artist link feature is what keeps me using it daily.

[Talisin]

Answering one by one, that my mind is not too well .

Paperclip: I have already installed sonicstage in my laptop Tags aren't a problem, since I'm used to fix them everytime I convert an album, and also Dinamic playlist are in my weponry XD My concern is with quality. The way you say you rip and convert is the same I usually take (first to wav, then to atrac) but I have read, not only from sfbp, that converting tracks with this system doesn't give good quality songs.

Sfbp: Simpleburner works under windows 7 or is necessary to have a xp virtual machine?. Also, this'd imply to burn flac album to cd first and then rip with Simpleburner, wouldn't it?. About aal, thanks for clarification, I had understood it the opposite way. So you have all your songs in a hard disk and transfer or convert it when you need, aren't you? I was thinking to pass the hard disk files, just keeping my CDs and . And you have caught me with that mp3conversion tool: I always used hi-md renderer for doing this, but in W7 it gives me problems. I have to try it.

Oh, and thanks a lot for 24 bits explanation. Does that means that with a lower bitrate songs sounds better in atrac than with a higher bitrate in mo3 or other formats?.


I think my main concern now is to try again lp2 and then choose between aal256 and lp2/sp (depending of results I get). Thanks a lot.

[paperclip]

The myth that Sonic Stage cannot correctly convert wav files is simply that. A myth. I do it all the time and it works great. I don't really know how that myth got started.
I have my CD collection on a HDD and convert to whatever format I need. I use LAME 3.98 to make mp3 files at the same time as the CD is ripped, mostly for listening on computers that don't have all my favorite software. I use wav or some form of ATRAC for my playback. If you want quality, I suggest you use LAME instead of Sony's MP3 tool.

[sfbp]

The myth that Sonic Stage cannot correctly convert wav files is simply that. A myth. I do it all the time and it works great. I don't really know how that myth got started.

I take 100% responsibility. I (like, it seems many others here) tried LP2 and LP4 very early on in the game and found the results were awful. I did so by importing a WAV from a CD, and then downconverting it. In the end I concluded it must be the import step which was at fault, for whatever reason.

There's a good reason that I believe this may be a defective process. I think that CD playback, and also ripping direct to ATRAC actually manage to recover more than the nominal 16 bits on an audio CD. We have read for 25 years about oversampling which effectively gives you 18, 20 or even more bits of resolution with a good player, or an effective ripper.

I totally agree that if one has GOOD wav files (such as those generated by EAC, when properly configured) then they can be downconverted quite well. But I think you can see that a 16-bit file on disk might not downconvert to a low rate 24-bit format (ATRAC) as well as direct ripping to that same format.

YMMV.

[paperclip]

A ripping program doesn't even sample a signal, it just reads data. Your claim that a 16 bit file might not compress to a low bit rate codec defies all logic really. And (on a undamaged) CD the output can and usually is, bit for bit exact between EAC and Sonic Stage. What do you think Sonic Stage is using to creat the ATRAC file? It's reading 16 bit data. If you compare 16 & 24 bit wav files compressed with ATRAC, you will find their files sizes are exactly the same.
I've done the testing to proof the results. I don't know why you want to spread mis-information on a hobbyist website like this?

[sfbp]

Before we get to name-calling, please stop and consider why as early as 1985 CD player manufacturers were advertising 20-bit oversampling......... (well, actually 16x oversampling, i.e 16 + 4 = 20 bits)

My impression of the oversampling technique (and it may well apply to different techniques for ripping the music off a CD) is that you end up with proper dithering of those values which lie "between" two digitally-encodable values, and that the effect is to "anti-alias" the sounds, so there are no interfering harmonics. Musicians will tell you that all kinds of notes that are never played make up the timbre of an instrument. CD's start to sound natural (and, incidentally to compress well) when the anti-aliasing (which is incidentally used in graphics processing systems ubiquitously) is applied, and the false harmonics cancel out.

As I said, my $.02 - I can't prove most of this, and I don't have a proper background in sound engineering. But I definitely know it's possible to get a perfectly good-sounding wav file which absolutely refuses to compress properly, using SonicStage as the initial ripper. I presume this is what Sony is referring to when they talk about "fast" or "high quality" ripping speeds for their other codecs. Interestingly they don't offer that option (analogous to the oversampling CD players) for Sonic Stage's WAV extraction.

Stephen

[paperclip]

I don't know what you mean by name-calling unless you are setting up an ad hominem argument?
Ripping just doesn't oversample anything, it just reads the data on the disc.
Dithering only applies when reducing word lenght which makes no sense in your example of increasing word lenght.

Unsupported claims do nothing to further the hobby. Post the results, methods and repeatability like I've done.

[sfbp]

Ripping just doesn't oversample anything, it just reads the data on the disc.

I'm not sure you are right. Otherwise why would Sony specify different speeds of ripping ("fast" vs "high-quality") for some of its codecs? That word "just" is always a red rag to this particular bull.

And why should one not think that oversampling on a CD player/reader is not matched by the same thing in the digital domain?

[paperclip]

What data do you think the codec uses? If it's from a CD, it's a 16 bit wav file. It's that simple. A CD player putting out an analog signal is a wholly different animal. Oversampling while ripping would defeat the purpose, how could it compare the data it's reading for accuracy if it changes it from what is on the disc?
I'm wholly unaware of Sonic Stage having any speed or quality options in converting a wav file to ATRAC (or importing it directly in wav format from a disc) and that is what the OP is trying to do. Why do you insist on making it so hard when there is a simple solution? Again, post your results. Maybe we can figure out why you have a problem and I don't.

[sfbp]

Take a look at the options for ripping CD's in SonicStage - perhaps you will see what I mean. I've explained my logic, I don't think there's anything else I can say to convince you.

[paperclip]

Yep, other than post real examples and a screen shot you are correct.

[eriktous]

My guess is the different options for ripping speed in SonicStage may have to do with error correction.

I would have to search around for the specifics, but from memory this is how I understand it. The red book standard for audio cd's has less inherent error correction information than e.g. data cd's. That's why a ripped cd above a certain length can not be burned back to a cdr as a data file (.wav). The cd data format uses more room for error correction (digital data has to be read back bit perfect, obviously), so the usable size is less.

A cd player during playback will usually misread some samples, but a single wrong sample (or even a few close together) will hardly be audible once converted to the analog signal. (The error rate during reading may be one of the differences between high and low quality cd players, although laser technology is now so evolved that even cheap players should do pretty good.)

When ripping a cd on the computer, you can do this high speed, i.e. read just once and don't care about the misread samples, or you can do this more accurate, e.g. by reading twice and comparing the two data streams, or computing a checksum and comparing this to a database (AccurateRip).

[sfbp]

Erik, if you're saying that Sony does (or did) the first when doing a normal Rip of CD->WAV with SonicStage (read fast and don't care about the error correction), I'd be inclined to believe you.

If, OTOH paperclip is saying that SS now does the second way, and can prove that the rip is correct and matches EAC, then I will believe him.

Could it be that when I originally got turned off to LP2 that the ripper in SS was defective and they later fixed it?

If so, I'm willing to eat humble pie and publish a retraction. But I saw what I saw, and there has to be an explanation, even if we haven't (yet) found it. And others have declared themselves similarly fed up with MDLP.

It's even possible, isn't it. that this was the reason that Sony "lost" the MP3/Ipud comparison wars 8-10 years back - because someone with left-brain thinking decided (just as I did, to be fair) that the "level playing field" way was to start by ripping the CD directly to WAV file and then rip to different rates of ATRAC compression. Of course that explanation only works if the testers used SS ripper to rip for the sound files being used to test to convert to ATRAC, but some other ripper for the files being converted to MP3.

What I *do* recall was that originally the ripping speed was really fast. I *was* subsequently puzzled by slower ripping on a later version of SS.

So it is indeed possible that this artifact only existed in 4.0 and/or 4.2 SonicStage (or versions even earlier than that), and is now gone. I think I recall one of those two, but I don't remember which.

Stephen

Attached Files

•  quality.jpg   46.5K   3 downloads