Can minidisc (ATRAC SP) capture the essence of vinyl?

[netmduser]

Hello,

Can minidisc (ATRAC SP) capture the essence of vinyl? What do you think? Have you ever recorded directly from vinyl, I was wondering if minidisc could capture that vinyl sound so sought after? I have experienced a few vinyl albums recorded on minidisc and it definitely sounds different than when the same song is recorded from CDs.

Edited July 18, 2009 by netmduser

[sfbp]

FWIW my whole focus for MD was to convert vinyl to CD, via MiniDisc.

I think probably we end up doing a much better job than when the record companies did it in the 1990's.

Cassettes, too. I laugh when I have converted a cassette and I can hear the vinyl-characteristic pops where they (big companies) transferred to cassette. Of course, I can get rid of those, too.

Answer to your question: I dunno. I remember being very wary of CD's remastered initially. I think now it may be better, especially anything that was digitally recorded to start with, there should be no obvious problems.

[netmduser]

Okay, wow you're the vinyl expert, you will have to pass your master skills to us here.....

I really do enjoy hearing that occasional pop and crackle, it really does give that distinctive quality for mixed music. CDs often then not, give me that white room feeling.

Edited July 18, 2009 by netmduser

[sfbp]

Okay, wow you're the vinyl expert, you will have to pass your master skills to us here.....

I really do enjoy hearing that occasional pop and crackle, it really does give that distinctive quality for mixed music. CDs often then not, give me that white room feeling.

<blush> I think that's putting it a bit strong. I find from talking to Avrin and others that I have ended up doing the right thing mostly by trial and error, then figure out WHY it worked afterwards, lol.

Maybe I should write it up, but it's not top of my project list right now. I do think that a decent editor is essential. I tried wrestling with the one included in Nero packages (that I paid for, not the completely free ones!) for the sake of a friend I want to introduce to this, and got frustrated very quickly.

I was lucky enough to get CoolEdit96 for about $50 and it has served me well. I'm not sure if I would spring for its child, Adobe Audition, though. But you need something really decent.

The other prerequisite is, IMO, an inexpensive sound card with digital In and Out, and the ability to ignore SCMS.

One more thought - I only recently found out that ripping ***from*** CD is a most uncertain process. Could this be a clue to your listening experience?

[Sony_Fan]

I have a CD recorded deck that I can hook up to the outputs of a vinyl record player. I have not done such a recording, but I'm sure it can capture the vinyl sound without loss of sound quality. ATRAC SP, even as good as it sounds, is a lossy compression codec, so some vinyl sounds could be lost during the recording. If you have a Hi-MD unit, you can record using PCM from vinyl and without loss of sound quality. I own Michael Jackson's "Thriller" album on vinyl. Will experiment with this.

[bobt]

I have a CD recorded deck that I can hook up to the outputs of a vinyl record player. I have not done such a recording, but I'm sure it can capture the vinyl sound without loss of sound quality. ATRAC SP, even as good as it sounds, is a lossy compression codec, so some vinyl sounds could be lost during the recording. If you have a Hi-MD unit, you can record using PCM from vinyl and without loss of sound quality. I own Michael Jackson's "Thriller" album on vinyl. Will experiment with this.

Hi Chris,

You may need a preamp for your table, as most cartridges hav too little output to dive a recorder, as well the preamp adds the preemphasis to get the bass and treble right. Unless your CD recorder has a built in preamp I'm not sure it would work. OTOH one of the newer USB turntables would be interesting!

Try it and let us know how it works

Bob

[Tom in RI]

I have done a fair amount of transfer from vinyl to cd (and then onto md for some things). I've been happy with the results. But in terms of whether md can capture the essence of vinyl, its a non issue for me as I only listen to md through headphones and I listen to vinyl predominately through a couple of floor speakers. It wouldn't occur to me to set up the md as a source player on my home stereo.

Sfbp, I picked up Goldwave some time ago for $40 Canadian. I do not even scratch the surface of all the functionality it has but I do use the program to add and delete track marks and more importantly to filter pops and clicks. I find the track mark editing cumbersome but I love having the ability to edit pops and clicks, although I find that the best result comes from removing pops an d clicks one at a time which can be quite time consuming.

[sfbp]

I have done a fair amount of transfer from vinyl to cd (and then onto md for some things). I've been happy with the results. But in terms of whether md can capture the essence of vinyl, its a non issue for me as I only listen to md through headphones and I listen to vinyl predominately through a couple of floor speakers. It wouldn't occur to me to set up the md as a source player on my home stereo.

Sfbp, I picked up Goldwave some time ago for $40 Canadian. I do not even scratch the surface of all the functionality it has but I do use the program to add and delete track marks and more importantly to filter pops and clicks. I find the track mark editing cumbersome but I love having the ability to edit pops and clicks, although I find that the best result comes from removing pops an d clicks one at a time which can be quite time consuming.

Thanks, I'll check it out. My own approach to pops and clicks is very simple in concept: the ear cannot possibly miss 1-3 milliseconds of COMPLETELY DELETED sound.

As far as playback through my stereo goes, I had to be content with SP until I finally got a decent unit that had Type-S, since I cannot afford and don't wish to live with a million SP discs. This enables excellent playback of LP2 (which is plenty for portable listening) and, in some situations, LP4. OTOH the JE640 which has now become my workhorse for capturing internet streams I want to playback later (rather than littering the HD) lasted about 5 minutes as a playback machine. It just wasn't good enough at the MDLP modes.

Would be delighted to compare notes, particularly since I never even *tried* it your way. My philosophy was: if the recording to MD is so good (as claimed) why gamble or spend too much money on a computer sound card with essentially unknown specs if I can get the sound digitised by Sony first, then transmitted to the computer, later. But I'm open to discussion, for sure.

[Sony_Fan]

Hi Chris,

You may need a preamp for your table, as most cartridges hav too little output to dive a recorder, as well the preamp adds the preemphasis to get the bass and treble right. Unless your CD recorder has a built in preamp I'm not sure it would work. OTOH one of the newer USB turntables would be interesting!

Try it and let us know how it works

Bob

Hey Bob, yeah, I think I understand what you're saying. What about just using an RCA Y-cable and connect it to a Hi-MD unit and turn the recording level all they way up? I don't see why that wouldn't work.

[bobt]

QUOTE (Chris G @ Jul 18 2009, 12:31 PM) Hey Bob, yeah, I think I understand what you're saying. What about just using an RCA Y-cable and connect it to a Hi-MD unit and turn the recording level all they way up? I don't see why that wouldn't work.

You can try, but the other thing then levels that a preamp does is the preemphasis on frequencies, If i remember right the output from the cartridge is treble high, bass low, and the preamp corrects for that.

Try it, and let us know,

Bob

[Strungup]

The Preamp is s the RIAA EQ curvature , without it the record will sound like a cat caged in a Tin bucket on a Hot day

Some turntables have the preamps built in and have a Standard Line out . I have one of those , and Transfer a lot of Vinyl to Md ,

[darrencouch]

I've done plenty of copying vinyl to SP and LP2 with my JE-780 deck. I have an older decent Kenwood direct drive TT and a mid-70's panasonic/technics receiver with phono amp. Needle is a Shure M97xe. I have heard of all the "lossy" this and "lossy" that, but the results are great. I like the pops and hiss-on a good piece of vinyl I can set the mark trip at -34db or quieter and get accurate track separation. Of course, it takes a LOT of fiddling (which i happen to like). The turntable needs to be set up just right, cartridge alignment spot on, counterbalance set, etc. Then finding the loudest passage on the record to set recording volume without clipping. Record should be clean as possible. I plan on eventually getting a good record cleaner. Like anything, the results are heavily dependent on the quality of the source.

[Avrin]

The "vinyl sound" is just a myth. What actually makes LPs sound better than their CD counterparts is proper mastering. And there actually were properly mastered CDs, that sounded as good, if not better, than LPs. Excellent examples are the Toshiba "Black Triangle" issues of "Abbey Road" and "Dark Side Of The Moon". With pre-emphasis and all other technologies properly used.

Trouble is, at the beginning of the CD era, engineers were afraid to do flat transfers to CDs (and probably lacked proper equipment), and then loudness wars started, so the CD never had a chance to contain quality records it is designed to contain.

All this means that you can make really good sounding digital copies of LPs provided that you have quality equipment and apply a 20 kHz analog lowpass filter before digitizing (most, if not all, MD equipment doesn't have it, but it is required for a proper analog-to-digital conversion). And even a lossy copy of an LP may sound better than the respective CD.

[sfbp]

All this means that you can make really good sounding digital copies of LPs provided that you have quality equipment and apply a 20 kHz analog lowpass filter before digitizing (most, if not all, MD equipment doesn't have it, but it is required for a proper analog-to-digital conversion).

Are you sure that the MD decks don't have it (as opposed to portables)? I know little about such things but I always seem to get "perfect" digitization of LP's when using the analogue input to an MD deck. Either that or my line out on the Kenwood amplifier does it first. I looked at the block diagram of a couple of MD decks, and I didn't see any lowpass filters in the relevant signal path - but I wouldn't know one of these if it rose up and bit me

I don't even know how I would tell for sure if such a filter had been applied, but looking at the freq response of any WAV file that's come from Analog->MD recording, it seems like there is nothing at all above 18Khz. I always thought that was a side effect of the ATRAC encoding, and if so, your comment about lowpass filters means that one is not needed because ATRAC basically throws that away during signal processing.

[Avrin]

I am not sure about decks, since I never had one. But the problem with installing a lowpass filter on an analog input (be it a portable or a deck) is that its performance will seriously depend on the output parameters of the equipment feeding the signal. Thus it is impossible to create a "universal" filter suitable for everything.

On the other hand, the Nyquist theorem requires that the maximum frequency present in the signal be less than half the sampling frequency. Otherwise it will not be possible to digitize and then restore the signal unambiguously.

And a similar filter is actually needed after the DAC in the playback device. Only then the restored signal will be a copy the original analog one (to the extent the digital resolution allows). But if one or both filters are absent, horrible things happen.

Applying a digital lowpass filter after the signal is digitized does not improve things, since the signal is already incorrect.

Edited October 22, 2009 by Avrin

[sfbp]

There's no horrible things happening..... so I guess I'm ok

What do you mean "in the playback device"? This is an analogue signal.

I could filter wave files once they get to the PC, but I don't think I need to since the output has gone to effectively zero well below Fs.

[Avrin]

An analog output lowpass 20 kHz filter is absolutely required. Otherwise the output signal won't even come close to the original. Even if you have everything above 20 kHz removed digitally, you'll get tons of higher frequencies in the analog signal after the DAC, and the filter is needed to remove them.

A simple example is a pure 20 kHz sinewave. After digitizing it, the signal will contain only two values per period. So, after the DAC you won't get a sinewave. You'll get a square wave. But if you put that square wave through a 20 kHz lowpass filter, you'll get the original sinewave back.

That's the way digital audio must work to satisfy all mathematical requirements it is based upon: analog source->analog lowpass->ADC->media->DAC->analog lowpass->amplifier.

[sfbp]

you'll get tons of higher frequencies in the analog signal after the DAC.

What DAC? I have an analogue amplifier.

[Avrin]

Any digital player, including MD and CD, has a DAC. The exact configuration depends on the architecture. If the unit has an analog amp, it contains a 1-bit delta-sigma DAC, and then the amplifier (and hopefully a lowpass filter between them), and if the unit has a digital amp, the DAC (also 1-bit delta-sigma) is the amp itself.

[sfbp]

But I am not doing anything of the sort. I have an analogue amplifier. I am not attempting to record digital playback of anything.

I have connected to it: a tuner, a reel-to-reel deck, a cassette deck, a turntable (Phono preamp in the receiver), and several video components (which video components have only been used about once in this way).

Which one of these has a DAC?

[bobt]

But I am not doing anything of the sort. I have an analogue amplifier. I am not attempting to record digital playback of anything.

I have connected to it: a tuner, a reel-to-reel deck, a cassette deck, a turntable (Phono preamp in the receiver), and several video components (which video components have only been used about once in this way).

Which one of these has a DAC?

AFAIK the DAC is in the input stage of the MD unit, I seem to remember a while back of someone using a MD unit as a DAC to go to another unit, recollection is vague. MD will not work without a DAC in it

Bob

[sfbp]

I think the only DAC is in an output stage? Anyway I don't care about output at all, since once I get the sound to the MD, it is digital and stays digital until burned or converted to analogue by being played.

So I can definitely master a CD without ever doing D-to-A.

[Avrin]

If you are only using analogue equipment (no CD or MD players), then there is no DAC, and no filters are needed.

[sfbp]

Remember my whole purpose was to turn analogue media into digital sound via MD.

Sure, I have a Sony CD player hooked up to the amp, but in an analogue way and Sony presumably may put such a circuit on the analogue output, just as they do on every MD deck I can find (take a look for yourself, I am sure you understand the circuits much better than I do).

I used CD optical out (from the changer) to go straight to MD deck a couple of times, and the results were exactly (AFAIK) the same as using the intra-deck transfer on the MXD-D400.

So what I am puzzled by is: what device could I possibly have that doesn't have such a filter? Unfortunately I don't have a service manual for that deck. But most of the time I do like anyone else, and use Simple Burner (or Sonic Stage) - it's faster.

[Avrin]

Many modern devices (especially the ones designed for several sampling frequencies, e.g., DVD players) don't have proper output filters. However, this doesn't affect direct digital transfers to MDs.

[sfbp]

Many modern devices (especially the ones designed for several sampling frequencies, e.g., DVD players) don't have proper output filters. However, this doesn't affect direct digital transfers to MDs.

Does this apply to the analogue (line out) output from a DVD player in the event I record sound from it? I notice that only 2 of my 3 DVD players will actually record to MD from the optical out, the other is presumably producing SCMS signal, or is restricted in some subtle way I don't understand.

[Avrin]

Any analogue output of any digital device should have a lowpass filter preceding it.

As for the DVD player that doesn't allow recording form its optical out, check whether any sound processing or upsampling is enabled in it. E.g., Panasonic's Double Remaster and Multi Remaster modes are known to increase the sampling frequency to the point no MD unit understands it.

But some commercial DVD releases really don't allow making digital copies. However, conventional CDs should pose no problem when creating a first-generation MD copy.

[sfbp]

Any analogue output of any digital device should have a lowpass filter preceding it.

Back to the original comment, then. I am at a loss understanding in what circumstances might I observe a problem with this effect? Certainly the MD decks have it (a filter, not a problem). And anyone recording Vinyl from a turntable will have an analogue circuit not a digital one, as the output.

[Avrin]

This stuff is really simple - if there is no input lowpass filter before the ADC, the digitized signal will not accurately represent the original analogue one. And if there is no output lowpass filter after the DAC, the output analogue signal will not represent the original analogue one, even if that was digitized perfectly. Whether you'll hear the actual difference, is another question.

Both filters are required by the mathematics the digital recording technology is based upon. Otherwise the end result (what you hear) will not be correct.

In other words, the digitization/restoration process has its mathematical limitations. Thus the input filter is required to make the signal suitable for digitizing at a particular sampling frequency. And the output filter is required to fully restore the unique original signal.

[sfbp]

No argument with anything you are saying. But I cannot think of a single example (that might affect me, at least!) where either of these problems might occur.

I do recall getting a big mess because I recorded something off internet radio that had a huge spike in it from the FM carrier frequency (but that's below 20Khz). But I gave up listening to CBC radio shortly after that! IOW this is something no ordinary mortal should ever have reason to worry about, except with other folks' incompetence.

[Avrin]

As I've already said, if there is no output lowpass filter, higher frequencies on any CD will be played as square waves, and not as sine waves.

[sfbp]

All this means that you can make really good sounding digital copies of LPs provided that you have quality equipment and apply a 20 kHz analog lowpass filter before digitizing (most, if not all, MD equipment doesn't have it, but it is required for a proper analog-to-digital conversion). And even a lossy copy of an LP may sound better than the respective CD.

But we're talking about LP's not CD's. And you now (last few posts) admit that the only place one would see this problem is from a CD.

Are you perhaps suggesting it would be possible to "fix" badly mastered CD's? If not I still don't see the importance of your comment. Maybe I'm just being dumb...

[Avrin]

The problem is related to any digital media, including CD, MD, DAT, DVD, LaserDisc/video tape with digital soundtracks, Flash, HDD, etc. And to transferring analogue media to thee digital domain, both in the studio and at home. It has two effects:

1. No input lowpass filter = incorrect digital copy (most modern CDs are made this way, but things were different in the 1980s, even outside Japan).

2. No output lowppass filter = incorrect playback (even if the digital copy is perfect).

And it is not possible to fix improperly mastered media.

[Avrin]

Just a couple of illustrations.

A perfectly mastered (and fantastically sounding) CD from 1983:

A 2001 "Special Edition" CD, that will never sound good on any equipment (and I'm not even talking about compression and clipping distortion):

The track is the same - "Wanna Be Startin' Somethin'" by Michael Jackson.

Edited October 25, 2009 by Avrin

[sfbp]

So we're actually talking a situation where MD (SP, compressed) beats uncompressed because there's absolutely zippo (0) above 18Khz on anything that is recorded in ATRAC. I would never have guessed. However the recordings that come from the route I used (LP->MD, then MD->PC, then PC->CD) so closely resemble your first diagram, it's amazing. (Except that the roll-off is slightly lower, of course - but I'm sure you'll agree with me that this is basically unimportant from a listening perspective.).

I don't like MJ now, and I certainly didn't in 1983 But that just makes me a square.

[Avrin]

Well, I could have posted Pink Floyd originals and remasters, if MJ CDs were not on a closer shelf.

But I'll say it again, the SP (or any lossy format) high frequency cutoff will not improve a digital signal, if it is already incorrect. Removing higher frequencies after digitizing, further destroys the music. E.g., the "Special Edition" still sounds like crap in SP.

[sfbp]

Here's the loudest 5 seconds I could think of off a disk I processed this way. So it correctly stops at 20kHz, well done Sony. My comment about 18Khz probably arises from the fact that on average classical music is rarely this loud for more than transient passages.

I think that this suggests that MJ must be this loud all the time - no wonder I don't listen to Pop of any sort, I would have been deaf by 1983, instead of only just going so now.....

[Avrin]

The above picture nicely ilustrates an example of distorted signal with a high frequency cutoff. This can be easily seen from a non-flat curve beyond 20 kHz - all these low-level high frequencies are nothing but distortion. When higher frequencies are removed correctly using a proper input analogue filter, the curve is perfectly flat beyond the filter band. And again, even if you remove these frequencies digitally, you will only get additional distortion of the signal at lower frequencies.

[sfbp]

Why would the ear even hear the difference between -120 and -121 dB? As I told you above, there was NO digital-to-analogue step here. And the difference between the small residual and the portion of the curve you show above 20kHz on your "remastered" example is night and day. One is within the range of hearing, the other is not. You show significant sounds at -60dB, the comparable place on mine is at -120dB, which is 10 (logarithmic) times softer.

Distortion that's inaudible because it's too high freq is one thing; distortion that is so small that it cannot be heard on grounds of amplitude is another. Most likely that tiny variation at high freq and low amplitude is a side result of FT noise removal on the whole wave, I should think....? In other words the digital equivalent of noise, but not enough to interfere with hearing.

On your original point, I did check the output of (the only) music DVD that I recorded to minidisc, and it was horribly clipped at some point during its life, judging from the wave forms. Sounds pretty good though - I don't think this was the result of failure to cut off at 20kHz either.

[Avrin]

Clipping is only too common on all media now. The second version of the MJ track above is horribly clipped (as is the entire set of his 2001 releases, including "Invincible" and Special Editions of his earlier albums). And I've seen clipping even on DVD-Audio. And on CDs with classical music released in Japan. Even some vinyl releases now have traces of digital clipping.