Jump to content

Tweaking any MiniDisc recorder - FOR FREE (with spectacular sonic results)!

Rate this topic


MDietrich

Recommended Posts

Hi guys & gals!

I´ve written an article - about MD again - where I describe a tweak to improve sound quality of any MiniDisc unit. This tweak basically describes how to use an equalizer to improve the ATRAC encoding. I really think that the results are spectacular and that they should be shared with anyone else. Enjoy the article!

http://marlene-d.blogspot.de/2013/09/spectacular-sounding-minidisc-tweaking.html

Link to comment
Share on other sites

Actually what you describe is done automatically by many situations when either:

a. Sony decks or portables record; or

b. Sound Forge converts; or

c. SonicStage transfers (with conversion).

However it is possible as many people have discovered, to beat the "automatically" by choosing some slight screwy conversion or sequence to pass the sound through. This is why, IMHO, ATRAC failed the critical test comparisons back when it was actually competing against MP3. The organizers of the test tried to find a "fair" way to compare. Unfortunately there probably isn't one, and in doing so they triggered the effects you describe in your article. Thus the ear is the only way to be sure you are making a decent, "keeper" recording.

Yes, I can record good classical music off the radio at 66kbps (LP4) onto a deck. It comes out well. But the weirdest of things will mess it up. For example double mono announcer (attributable to the fact that LP4, unlike LP2, is joint stereo ie signal + difference). For example too much good data such as a 320kbps signal.

If you want to get good LP4 rips from a CD, for example, it may be advisable to rip to a format of intermediate bit rate (256 or 132 kbps) and then convert a second time from the intermediate form. This ensures that the proper equalization is one (after all Sony "knows" that you don't really want to convert CD to LP4, heheh). Forget all the nonsense about doubling the mythical "ATRAC generation losses" - it's all about achieving the transformation you describe.

And I swear I can tell stuff that's been RECORDED at MP3 immediately. Converted to MP3 isn't such a problem, as Sony and others learned to do similar tricks to what you are describing sometime after the fact.

Stephen

Link to comment
Share on other sites

Yes, many lossy codecs do this. Yet they all try to keep those high frequencies. They base their 'decisions' on how loud this content is. ATRAC however shouldn´t have been enabled to encode them in the first place... it never was the best codec. With ATRAC3 Sony was wiser since they remove this content anyway, especially LP4. But not ATRAC1 4.0 and higher.

Look at the samples I uploaded:

this is the standard ATRAC encoding. You can see that ATRAC tries to keep frequencies beyond 15.5 kHz. It does this based on their gain. Generally, the ATRAC codec isn´t good enough to keep those frequencies.

hqc4.jpg

My 15.5 kHz cutoff gets rid of them and enables the rest of the spectrum to be encoded in a superior way.

s9lx.jpg

Link to comment
Share on other sites

In other words, Sony ***knew*** this about the original (SP) Atrac, and took steps to quietly ditch it in favour of a more reliable, modified scheme (in A3 and A3+). The problem they were left with was (is) compatibility.

So SP has a special flavour but it's not necessarily the most accurate; it's what people got used to hearing. Makes perfect sense (to me) since for most, what came before SP was analogue. HiSP with a better encoding scheme and almost as many bits is probably better (well, more accurate).

Link to comment
Share on other sites

In other words, Sony ***knew*** this about the original (SP) Atrac, and took steps to quietly ditch it in favour of a more reliable, modified scheme (in A3 and A3+). The problem they were left with was (is) compatibility.

So SP has a special flavour but it's not necessarily the most accurate; it's what people got used to hearing. Makes perfect sense (to me) since for most, what came before SP was analogue.

This sums it up perfectly.

A few weeks ago I´ve read a comment on Tapeheads were somebody was talking about sonic differences between the MDS-JA 50 and the MDS-JA 555. The first with ATRAC 4.5, the latter with DSP Type-R of course. The JA 50 was described as sounding 'warmer', the more recent as too cool. Without knowing it, this person wasn´t talking about actual differences between the two units, he merely described ATRAC codec differences.

With my method they both will sound the same. Say, the sound quality of DSP Type-R turned into a non-existing, theoretical DSP Type-Z. Not everyone will like it though, since using my method creates a sound even closer to the pristine CD original, it simply sounds less 'warm'.

Link to comment
Share on other sites

Hello, I enjoyed reading your article. I downloaded the two wav files and listened to each wav file with the 3 samples on each one.

I was surprised they pretty much sounded the same.

Could you just clarify when you listen to them you are noticing a big improvement?

Yes, I indeed do. Occasionally, these differences are bigger as differences from, say, two different portable CD players.

Link to comment
Share on other sites

Count me as one who likes the "warm" analog sound, as from the MDS-JA20ES. My older ears don't care much why I like it better than the admittedly more-detailed MDS-JA333ES, they just do. I've read the somewhat famous JA555ES/JA50ES review/comparison several times now, and I would like to get hold of the JA50ES. Much easier said than done, though. IMO, an MD unit's DAC may be almost as important as its ATRAC version. DACs really do vary in sonic characteristics.

Link to comment
Share on other sites

Count me as one who likes the "warm" analog sound, as from the MDS-JA20ES. My older ears don't care much why I like it better than the admittedly more-detailed MDS-JA333ES, they just do. I've read the somewhat famous JA555ES/JA50ES review/comparison several times now, and I would like to get hold of the JA50ES. Much easier said than done, though. IMO, an MD unit's DAC may be almost as important as its ATRAC version. DACs really do vary in sonic characteristics.

I know what you mean... the JA50ES is a legend. I wanted to own it the moment it came out. But for used units the price is so high... I´m afraid owning it will forever be a dream. I myself would go for the JA555ES (or JA333ES). Although the reviewer makes a very convincing point by mentioning that the JA50ES with its warmer sound might make recent Loudness-War recordings bearable. On the other hand, I usually get rid of Loudness-War-related artifacts myself.

In case someone else wants to read what we were talking about: http://www.tapeheads.net/showthread.php?t=16445

Link to comment
Share on other sites

Fascinating articles although not sure my ears would notice significant difference. Having said that of the two MD decks I've owned, both Sony, D-D40 & JE480 I've always preferred the sound of the D-D40. Not sure why it sounds any different, and I only ever use SP so it's not down to the ATRAC type S of the 480. Was the output more tuned to the CD player combined in the D-D40?

I'd really love to hear what you do with "loudness war" recordings, they annoy the hell out of me. It'd be great to be able to take one of these obnoxious sounding CD's and copy a much more "mellow" version to MD.

Link to comment
Share on other sites

Fascinating articles although not sure my ears would notice significant difference. Having said that of the two MD decks I've owned, both Sony, D-D40 & JE480 I've always preferred the sound of the D-D40. Not sure why it sounds any different, and I only ever use SP so it's not down to the ATRAC type S of the 480. Was the output more tuned to the CD player combined in the D-D40?

I'd really love to hear what you do with "loudness war" recordings, they annoy the hell out of me. It'd be great to be able to take one of these obnoxious sounding CD's and copy a much more "mellow" version to MD.

MXD-D40: DSP Type-R, JE480: DSP Type-S. However, if I remember correctly, the MXD-D40 received very good reviews all those years ago. I´ve never heard it myself of course and I´ve also never read a review about the JE480. So maybe the MXD-D40 simply is the better sounding unit. The recordings themselves (from both recorders) should in theory sound the same.

Getting rid of the Loudness War? Try 'ReLife': http://www.terrywest.nl/utils.html

It´s a VST-plugin (you´d need a host for it). As a source, do you employ a digital output on your PC like me? If so, you´re in luck: just increase bit depth to 32 bit floating point, then 'treat' it with ReLife (volume is diminished by -4.21 dB to make room for 'new', interpolated dynamic peaks) and leave it in 32 Bit floating point. Play that back with 24 bits using a player like foobar and record to your MD units. You have to be aware though that ReLife introduces a slight (or severe, depends on how you look at it) low frequency phase error (which doesn´t hamper encoding quality but might introduce sonic errors).

In general, I treat 90% of everything I have on my HDD with ReLife and I love the result. Although the actual sonic difference is smaller than expected.

Wrote an article about this: http://marlene-d.blogspot.de/2012/09/need-to-panic-disease-loudness-war-is.html

Link to comment
Share on other sites

There's a subtle difference, at least from what I can tell. I'm not any kind of audiophile or someone who understands the specifics of encoding music, I have trouble distinguishing the difference between 192kbps MP3 and 320kbps, but then I don't listen for the differences, I just listen to the music haha. As long as it sounds good to me I'm happy.

Link to comment
Share on other sites

Try 'ReLife': http://www.terrywest.nl/utils.html

It´s a VST-plugin (you´d need a host for it). As a source, do you employ a digital output on your PC like me? If so, you´re in luck: just increase bit depth to 32 bit floating point, then 'treat' it with ReLife (volume is diminished by -4.21 dB to make room for 'new', interpolated dynamic peaks) and leave it in 32 Bit floating point. Play that back with 24 bits using a player like foobar and record to your MD units. You have to be aware though that ReLife introduces a slight (or severe, depends on how you look at it) low frequency phase error (which doesn´t hamper encoding quality but might introduce sonic errors).

In general, I treat 90% of everything I have on my HDD with ReLife and I love the result. Although the actual sonic difference is smaller than expected.

Wrote an article about this: http://marlene-d.blogspot.de/2012/09/need-to-panic-disease-loudness-war-is.html

Try DFX audio enhancer... http://www.fxsound.com/dfx/ , not too much with lossless files.

Link to comment
Share on other sites

Try DFX audio enhancer... http://www.fxsound.com/dfx/ , not too much with lossless files.

I wouldn´t use that for one minute. It´s a cheap thing aimed to people who don´t have any idea whatsoever about audio. It´s for people wanting bass, volume and pseudo-surround. The advertising looks like 'Dude hits some knobs and gets more bass.' This plugin makes everything worse and it sounds awful, it really does (listen to it more than 10 minutes and you´ll know what I mean: it is never subtle and it makes lots of errors).

It´s the opposite of getting rid of the Loudness War. And this tool lies blatantly... like all those 'all-in-a-box'-tools. It promises that 'remastering' is as simple as a click of the button - which it isn´t.

Link to comment
Share on other sites

imho the sb x-fi crystalizer enhances the sound a bit.Ok you need the x-fi card for that but i don´t think this is any worse to the sound.

Marlene can you say something about it ? i used to use this on a x-fi xtreme gamer proi version that i don´t use atm.

sadly the card in my htpc is an audigy 4 in whose drivers respective hardware that can´t be enabled.

I thought the only good way way to get better sound through hardware, isn´t it ?

;)

Link to comment
Share on other sites

Using the X-Fi crystallizer AND the DFX audio enhancer at he same time is a good source of distorsion, I agree with that, Specially with lossless files. But il is also a question of original recording level (that is why I always use now the Replaygain plugin). On my desktop (before it recently failed) I used only the X-Fi crystallizer plus optical out / foobar2000 / Ks or Wasapi audio outputs / Replaygain.

DFX audio enhancer - fixed once for all at 5.0.0.2.2 - is still usefull (and most of the time with no distortion...) with my laptop on board audio card (no X-Fi...). I promise to forget it completely when I will get one day for example an M2tech interface and a DAC Naim audio :imsohappy:.

But as I have started to use (slowly) the "DAW" Samplitude pro, I hope to make some progress one day in my recording quality using some of his features :heat: , banishing forever DFX !

Link to comment
Share on other sites

imho the sb x-fi crystalizer enhances the sound a bit.Ok you need the x-fi card for that but i don´t think this is any worse to the sound.

Marlene can you say something about it ? i used to use this on a x-fi xtreme gamer proi version that i don´t use atm.

sadly the card in my htpc is an audigy 4 in whose drivers respective hardware that can´t be enabled.

I thought the only good way way to get better sound through hardware, isn´t it ?

;)

If one keeps the amount of influence of the Crystalizer small enough it might work for some material.

One could do that himself... add bass, treble and presence area, a bit of peak limiting, resample all of it to 96 kHz with a so-so resampler and make it 3 dB louder.

I had a X-Fi once. Never used the Crystalizer... this should be telling enough. My X-Fi HD USB uses the next version (without resampling I think) and I´ve never used that either.

But all of this is the opposite of getting rid of the loudness war.

Using the X-Fi crystallizer AND the DFX audio enhancer at he same time is a good source of distorsion, I agree with that, Specially with lossless files. But il is also a question of original recording level (that is why I always use now the Replaygain plugin). On my desktop (before it recently failed) I used only the X-Fi crystallizer plus optical out / foobar2000 / Ks or Wasapi audio outputs / Replaygain.

DFX audio enhancer - fixed once for all at 5.0.0.2.2 - is still usefull (and most of the time with no distortion...) with my laptop on board audio card (no X-Fi...). I promise to forget it completely when I will get one day for example an M2tech interface and a DAC Naim audio :imsohappy:.

But as I have started to use (slowly) the "DAW" Samplitude pro, I hope to make some progress one day in my recording quality using some of his features :heat: , banishing forever DFX !

Yes, you have much more control with a 'DAW'. The thing with DFX audio enhancer is that it isn´t subtle, it cannot be configured to work with precise markings. So the only thing it produces is hyper compressed, distorted music.

And the Crystalizer and DFX? Gosh, your poor ears.

Link to comment
Share on other sites

  • 2 weeks later...

Hi guys & gals!

I´ve written an article - about MD again - where I describe a tweak to improve sound quality of any MiniDisc unit. This tweak basically describes how to use an equalizer to improve the ATRAC encoding. I really think that the results are spectacular and that they should be shared with anyone else. Enjoy the article!

http://marlene-d.blogspot.de/2013/09/spectacular-sounding-minidisc-tweaking.html

I'm just curious, when doing CD -> MD optical recordings, which will produce better MD recordings: portable MD Type-R recorder or MD home deck Type-R recorder? I've heard the using MD decks can produce better recordings because of the available power for computing/encoding/compressing. If a bit is a bit (0 or 1), wouldn't the sound quality be the same no mattery which Type-R recorder you use? Thanks.

  • Like 1
Link to comment
Share on other sites

I'm just curious, when doing CD -> MD optical recordings, which will produce better MD recordings: portable MD Type-R recorder or MD home deck Type-R recorder? I've heard the using MD decks can produce better recordings because of the available power for computing/encoding/compressing. If a bit is a bit (0 or 1), wouldn't the sound quality be the same no mattery which Type-R recorder you use? Thanks.

Oh if only life was that simple. This whole subject of "aren't 0's & 1's all the same" has been brought up in countless forums in the past. If this was true all MD's would sound the same, all cd players would sound the same, all DVD's, etc, etc. And we all know they don't.

Even if we have two MD's that are both type R they may well have different DAC's plus there are a multitude of other things that can influence the sound before it reaches your ears. Whether or not these differences are always noticeable is sometimes debateable.

Someone more knowledgeable than myself can probably expand on this topic far better than I can.

Link to comment
Share on other sites

I'm just curious, when doing CD -> MD optical recordings, which will produce better MD recordings: portable MD Type-R recorder or MD home deck Type-R recorder? I've heard the using MD decks can produce better recordings because of the available power for computing/encoding/compressing. If a bit is a bit (0 or 1), wouldn't the sound quality be the same no mattery which Type-R recorder you use? Thanks.

That is indeed a very good question. I´ve asked this several times now (not necessarily here at this forum) but no one has yet answered me. I´d prefer some measurements but people generally don´t seem to care.

I´d love to know too. There are many guys with a DSP Type-R home deck around, I could send one of them a testfile which would need to be played back in 24 Bit by the optical output of a PC and recorded by the optical input of the Type-R-equipped MD home recorder. The disc with the testfile then needs to be played back by the MD home recorder and recorded digitally by the PC - again in 24/44.1. The result would need to be sent to me for analyzing.

BTW, I do NOT refer to the quality of DACs, MDs or anything else. This is just about possible encoding differences between portable and stationary recorders.

Reason: several Sony manuals seem to suggest that stationary MD recorders have a superior encoding quality.

Link to comment
Share on other sites

HI..

Just send me the file and i´ll be glad to record it with a JB930QS. The Soundcard iam currently uisng in my HTPC is an Audigy4.. so 24 bit is no problem at all.

I already recorded a file to test the DAC´s between the JB930 and the JB980. Same way as you described..It was a 24/48 file from Yello.Oh yeah :D

Is that a good choice to test with ?

I send you a PM with Email if youo like ;)

Link to comment
Share on other sites

That is indeed a very good question. I´ve asked this several times now (not necessarily here at this forum) but no one has yet answered me. I´d prefer some measurements but people generally don´t seem to care.

I´d love to know too. There are many guys with a DSP Type-R home deck around, I could send one of them a testfile which would need to be played back in 24 Bit by the optical output of a PC and recorded by the optical input of the Type-R-equipped MD home recorder. The disc with the testfile then needs to be played back by the MD home recorder and recorded digitally by the PC - again in 24/44.1. The result would need to be sent to me for analyzing.

BTW, I do NOT refer to the quality of DACs, MDs or anything else. This is just about possible encoding differences between portable and stationary recorders.

Reason: several Sony manuals seem to suggest that stationary MD recorders have a superior encoding quality.

How about just using a regular CD deck with optical output and connected to the portable Type-R recorder and then to the MD deck Type-R recorder. I do all my optical recordings from a CD recorder deck.

Link to comment
Share on other sites

How about just using a regular CD deck with optical output and connected to the portable Type-R recorder and then to the MD deck Type-R recorder. I do all my optical recordings from a CD recorder deck.

You cannot encode a 24 bit signal using that method. CD only accepts 16 bit. And any possible improvement of stationary decks will be eaten up by the quantization noise caused by those 16 bits.

Link to comment
Share on other sites

HI..

Just send me the file and i´ll be glad to record it with a JB930QS. The Soundcard iam currently uisng in my HTPC is an Audigy4.. so 24 bit is no problem at all.

I already recorded a file to test the DAC´s between the JB930 and the JB980. Same way as you described..It was a 24/48 file from Yello.Oh yeah :D

Is that a good choice to test with ?

I send you a PM with Email if youo like ;)

Can I say it? I love you! ;)

Just kidding. But I AM extremely thankful that you have offered this. And on the JB930QS of all things - gorgeous.

BTW, a 24/48 file is not the best thing to use for your MD recorders. Feeding them 24/44.1 is the better idea. Reason: they need to resample these 48 kHz files to 44.1 before it can be recorded to the disc (the MD cannot work with anything else). While the resamplers used for the MD recorders were very good all those years ago they aren´t so good in 2013. Most software resampler people can use for the PC offer a quality way higher (right now thinking about the free SoX plugin for foobar2000).

Link to comment
Share on other sites

You cannot encode a 24 bit signal using that method. CD only accepts 16 bit. And any possible improvement of stationary decks will be eaten up by the quantization noise caused by those 16 bits.

I'm not sure I understand. I'm reallly not interested in doing 24-bit recordings. Just a regular CD->MD recording using optical cable. Which will sound better: using a Type-R record deck or Type-R portable recorder?

Link to comment
Share on other sites

I'm not sure I understand. I'm reallly not interested in doing 24-bit recordings. Just a regular CD->MD recording using optical cable. Which will sound better: using a Type-R record deck or Type-R portable recorder?

Ah, now I understand. Yes, I´d like to know that too.

In order to know I need to perform some measurements. I lack a Type-R deck so punkrockaddict was so nice and offered to help me out. Thing is, I need to do this with a 24 bit signal, any other signal will simply hide any improvements, a possible advantage will never reveal itself with 16 bit.

So, to find out the best machine for 16 bit or 24 bit I need to measure a 24 bit recording.

But 24 or 16 bit... you do not need to concern yourself with that difference. The results will be posted here and I´ll draw a conclusion that should be helpful to everyone.

Link to comment
Share on other sites

iam not sure wether we will or won´t be surprised by the result- I could guess a small advantage by the Deck. Anybody else thinks so ?

But if the were on the same performance line.. wouldn´t that be fantastic too.?

I send the Test recording out so in a couple of days I`ll probably know more ;-)

While we wait for those results, how about another comparison: pre-recorded MiniDisc album vs CD->MD optical recording of the same album. Which would have better sound quality?

Link to comment
Share on other sites

While we wait for those results, how about another comparison: pre-recorded MiniDisc album vs CD->MD optical recording of the same album. Which would have better sound quality?

Oh, that´s actually pretty easy: pre-recorded.

Except very old pre-recorded MDs. Those have a frequency response up to exactly 15.5 kHz, beyond that point there´s nothing (except quantization noise).

Then there are mastering differences. Early albums have been mastered differently. One example: 'Love Deluxe' by Sade. The CD has way more bass and treble, the MD sounds boring in comparison. But this isn´t the fault of the pre-recorded MD, it´s just a different master (the difference between pre-recorded MD and CD suspiciously looks like a simple EQ). Another album would be 'Emotions' by Mariah Carey. The MD sounds - again - boring while the CD sounds more crisp.

But the more recent pre-recorded MDs (sometimes around 1995/1996) the better they sound. I have several albums from 1997, 1998, 1999 and 2000 that sound exactly like their CD counterparts. This cannot be topped by recording the CDs to MD.

Link to comment
Share on other sites

Oh, that´s actually pretty easy: pre-recorded.

Except very old pre-recorded MDs. Those have a frequency response up to exactly 15.5 kHz, beyond that point there´s nothing (except quantization noise).

Then there are mastering differences. Early albums have been mastered differently. One example: 'Love Deluxe' by Sade. The CD has way more bass and treble, the MD sounds boring in comparison. But this isn´t the fault of the pre-recorded MD, it´s just a different master (the difference between pre-recorded MD and CD suspiciously looks like a simple EQ). Another album would be 'Emotions' by Mariah Carey. The MD sounds - again - boring while the CD sounds more crisp.

But the more recent pre-recorded MDs (sometimes around 1995/1996) the better they sound. I have several albums from 1997, 1998, 1999 and 2000 that sound exactly like their CD counterparts. This cannot be topped by recording the CDs to MD.

Oh, that's very interesting! So pre-recorded MDs from 1995 and after sound better than the CD->MD optical copy? I remember a moderator on this forum claimed that there was no difference in sound quality. Have any tests been done? I always figured that the master tapes converted to MiniDisc would sound better.

Link to comment
Share on other sites

Oh, that's very interesting! So pre-recorded MDs from 1995 and after sound better than the CD->MD optical copy? I remember a moderator on this forum claimed that there was no difference in sound quality. Have any tests been done? I always figured that the master tapes converted to MiniDisc would sound better.

Well, I assume that pre-recorded MDs prior to 1995/1996 sound worse than their CD counterparts.

After 1996, well I haven´t heard a difference between pre-recorded MD and CD. Which means: the ATRAC encoder used for pre-recorded MDs is way better than the one used in recorders anyone was able to buy. One indication is that later pre-recorded have a frequency response of (max.) 19 kHz. This never changes, it always stays at 19 kHz. With MDs recorded on stationary/portable recorders this limit fluctuates depending on the music. Sometimes it ends at 15.5 kHz, sometimes at 17 kHz, sometimes at 20 kHz, and after that it´s at for example 15.5 kHz again. This goes on and on until the music ends.

That´s why this thread exists in the first place: those frequencies pose some problems for our recorders, so I came up with the idea to remove them.

Link to comment
Share on other sites

Dear. MDietrich

I want to hear the sound difference, however sound cloud didn't find your files. Can you send me? My e-mail address is mze10@naver.com

I'll try this method asap. It seems like amazing.

Huh? I can still see them.

Here´s the whole set: https://soundcloud.com/marlenes-musings/sets/atrac-tweaks

Here are the individual tracks:

Original ATRAC: https://soundcloud.com/marlenes-musings/atrac-4-0-encoded-standard?in=marlenes-musings/sets/atrac-tweaks

ATRAC 15.5 kHz cutoff: https://soundcloud.com/marlenes-musings/atrac-4-0-encoded-encoded-with?in=marlenes-musings/sets/atrac-tweaks

Link to comment
Share on other sites

What is exactly is being compared here? Stationary MD recorder vs portable MD recorder?

Link to comment
Share on other sites

What is exactly is being compared here? Stationary MD recorder vs portable MD recorder?

No, the things compared are what´s stated in the OP. Meaning: I compare a standard ATRAC encoding (made with ATRAC 4.0) to an ATRAC encoding where the to-be-encoded files had my 15.5 kHz cutoff (also made using ATRAC 4.0).

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Loading...
×
×
  • Create New...