Inferior quality of files created with SonicStage

[cochra1]

Hello fellow MD users.

Has anyone else out there found that the quality of music tracks converted into ATRAC by SonicStage are inferior to those recorded in real time via a digital output from a CD player?

Just try an experiment. Get your favourite CD track and use SonicStage to import the song and get it onto your Hi-MD player.

Then record the same song from your CD player onto the same minidisc, but this time record in real time using the optical input. Make sure you stick to the same recording standard, ie. Hi-MD best quality.

Now do some A-B testing.

The SonicStage version will be duller in the high frequencies, and harsher, less smooth than the real-time version. The difference is subtle, but strong enough to merit not using SoundStage to capture your tunes.

I would be really interested if any bods out there know why this is so, and if there is any other conversion software out there which can match real-time recording for quality.

[GQ Smooth]

Hello fellow MD users.

Has anyone else out there found that the quality of music tracks converted into ATRAC by SonicStage are inferior to those recorded in real time via a digital output from a CD player?

Just try an experiment. Get your favourite CD track and use SonicStage to import the song and get it onto your Hi-MD player.

Then record the same song from your CD player onto the same minidisc, but this time record in real time using the optical input. Make sure you stick to the same recording standard, ie. Hi-MD best quality.

Now do some A-B testing.

The SonicStage version will be duller in the high frequencies, and harsher, less smooth than the real-time version. The difference is subtle, but strong enough to merit not using SoundStage to capture your tunes.

I would be really interested if any bods out there know why this is so, and if there is any other conversion software out there which can match real-time recording for quality.

I haven't tried this, but I wouldn't be surprised. I have also noticed with SS that sometimes a track I have downloaded onto a disc, will sometimes have another track recored over it. I might just go back to using my CD player. More time consuming, but worth it.

[Richard]

The SonicStage version will be duller in the high frequencies, and harsher, less smooth than the real-time version. The difference is subtle, but strong enough to merit not using SoundStage to capture your tunes.

Have you got any hard evidence to back this (frequency analysis for example) or is this purely conjecture? You also don't mention the SS options you are using to rip..

[mdmania]

how about using simple burner 2.0? have you consider using that instead?

[cochra1]

how about using simple burner 2.0? have you consider using that instead?

That one is even worse! It saps the lower frequencies out of a recording and again sounds very harsh. Sorry if I sound fussy, but I'm a musician and like to hear things in reasonable quality (which minidisc is very capable of) but can't put up with the quality of ATRACK files created in Simple Burner or Sonic Stage (the better of the two).

I'm sure there will be others out there who have noticed the improvement with real-time recordings via optical cable. Try the comparison yourself my friend and see if you agree!

Thanks for replying and kind regards

Numbskull Andy

Have you got any hard evidence to back this (frequency analysis for example) or is this purely conjecture? You also don't mention the SS options you are using to rip..

Hi Richyhu

No, I haven't done any frequency analysis, I never felt the need as the audible difference is so blatant! I would be very interested of course to see the results of an analysis (and would be surprised if no one yet has done this) but for me the most important thing is trusting my ears. The very first time I used the Simple Burner software to copy an album from CD I was instantly aware of a weakness in the bottom end and subtle distortion. The latest version of Sonic Stage is, admitedly, much improved, but still doesn't match real-time recording through the optical lead.

Have you got any hard evidence to back this (frequency analysis for example) or is this purely conjecture? You also don't mention the SS options you are using to rip..

See my original reply. I should also mention, I did a quick experiment with Sonic Stage as follows:

I took a recording of a song that had been made in real time.

I transfered in into my Sonic Stage library.

I then transfered it back to my minidisc.

Again, I did an A-B test between the original recording and the one that had been transfered back-and-forth, and the latter definitely had an audible reduction in the upper frequencies and sounded less transparant, less smooth, more 'digital'. This convinces me that the moved file went through some form of processing during the move, and was not simply transferred from one location to another in its original form.

Not sure what you meant when you asked about which 'SS options' I used, but if you meant which quality setting, I always use the best quality mode in Hi-MD (not PCM but the top quality compression mode).

I haven't tried this, but I wouldn't be surprised. I have also noticed with SS that sometimes a track I have downloaded onto a disc, will sometimes have another track recored over it. I might just go back to using my CD player. More time consuming, but worth it.

Hi GQ

Have to say I never experienced that problem, but I'm glad you're considering the option of using real time recording through the optical lead as I know from personal experience that this pays dividends in a subtly - but definitely - improved quality.

[greenmachine]

What codec/bitrate exactly did you convert to/record in? ATRAC can stand for SP, which is a transcoded LP2 version if you transfer via software. For this particular codec, SQ will indeed be a lot better when recorded in realtime. This is a long known issue.

If you used a different codec (for example Hi-SP), i'd be interested if you did the test blind? Your brain can play tricks with you if you know the source (placebo effect). See here:

http://www.hydrogenaudio.org/forums/index....showtopic=16295

Edited February 5, 2007 by greenmachine

[timonoj]

What codec/bitrate exactly did you convert to/record in? ATRAC can stand for SP, which is a transcoded LP2 version if you transfer via software. For this particular codec, SQ will indeed be a lot better when recorded in realtime. This is a long known issue.

If you used a different codec (for example Hi-SP), i'd be interested if you did the test blind? Your brain can play tricks with you if you know the source (placebo effect). See here:

http://www.hydrogenaudio.org/forums/index....showtopic=16295

Wich version of SS are you using? It has been great improvement since the nasty v1.5 I started from. Also in coding quality.

[cochra1]

Wich version of SS are you using? It has been great improvement since the nasty v1.5 I started from. Also in coding quality.

Absolutely right, it has improved, I agree. As I said, I always use the maximum quality bitrate for HiMD recordings, whether using SS or reat-time optical recordings. But while the bottom end (bass end) is now sorted (it used to roll off considerably) the top end (high treble frequencies) are still just slightly duller, and the whole timbre is slightly less smooth when using SS.

I have a very good, trained ear - I'm a musician and sound engineer and have experience working with DAT masters and all sorts of wonderful recording formats, so I trust my ears!

What codec/bitrate exactly did you convert to/record in? ATRAC can stand for SP, which is a transcoded LP2 version if you transfer via software. For this particular codec, SQ will indeed be a lot better when recorded in realtime. This is a long known issue.

If you used a different codec (for example Hi-SP), i'd be interested if you did the test blind? Your brain can play tricks with you if you know the source (placebo effect). See here:

http://www.hydrogenaudio.org/forums/index....showtopic=16295

No, I only use the top bitrates in HiMD mode whether copying tracks in SS or recording realtime from an optical lead. I trust my ears - I'm a musician and sound engineer, and very experienced in recording in lots of formats (I used to use DAT for mastering years ago), and have a very acute awareness of sound and timbre and frequencies. To me the difference is so stark (even though subtle) there's no question of a 'placebo' effect. I've had a few replies from scepticle people all questioning me further, but so far no one seems to have actually tried an A-B test (blind or otherwise) themselves...

Regards

Andy

Edited February 6, 2007 by KanakoAndTheNumbSkulls

[timonoj]

Absolutely right, it has improved, I agree. As I said, I always use the maximum quality bitrate for HiMD recordings, whether using SS or reat-time optical recordings. But while the bottom end (bass end) is now sorted (it used to roll off considerably) the top end (high treble frequencies) are still just slightly duller, and the whole timbre is slightly less smooth when using SS.

I have a very good, trained ear - I'm a musician and sound engineer and have experience working with DAT masters and all sorts of wonderful recording formats, so I trust my ears!

No, I only use the top bitrates in HiMD mode whether copying tracks in SS or recording realtime from an optical lead. I trust my ears - I'm a musician and sound engineer, and very experienced in recording in lots of formats (I used to use DAT for mastering years ago), and have a very acute awareness of sound and timbre and frequencies. To me the difference is so stark (even though subtle) there's no question of a 'placebo' effect. I've had a few replies from scepticle people all questioning me further, but so far no one seems to have actually tried an A-B test (blind or otherwise) themselves...

Regards

Andy

You said the source was a CD? Does it pass through any mp3 conversion? Maybe ffdshow has an incorrect settings...Do you see at any time the ffdshow logo just left to the clock when encoding? I don't find any difference at all from taking directly the signal from a CD-player output or compressing with SS...

Edited February 6, 2007 by timonoj

[greenmachine]

Does this mysterious maximum quality bitrate have a name?

[cochra1]

Does this mysterious maximum quality bitrate have a name?

I must admit, I'm getting more than a little disillusioned with reading all these comments, but here goes!

The max bitrate to store songs on your portable minidisc is Hi-SP mode (besides PCM of course). I'm advised this is set to 256kbps.

I have experimented in SonicStage in terms of importing into library from CD, choosing both ATRAC Advanced Lossless (recording Quality set to High, using 352kbps) and WAV (then converting to 352kpbs at the transfer-to-minidisc stage).

I couldn't perceive any difference what-so-ever between either of the above methods upon listening to the results on my Hi-MD walkman in an A-B test. Both were adequate enough.

HOWEVER

As soon as I listened to the same track taken from an optical real-time recording (in Hi-SP mode) the difference was immediately recognisable. More weight in the bottom end (bass drum) and a clearer, more transparant top end (most noticable on the hi-hats and cymbals). The whole track sounded more alive.

I then went back to the SS-created track. Again, less bass weight - the bass drum doesn't have the same power - and with the top end the hats and cymbals have lost clarity and sparkle. The mid frequencies appear the same to my ears.

If it makes any difference, I can tell you I had a colleague (fellow studio engineer/musician) listen to the different versions of the track today briefly to see if he could hear any difference (I did this cos I thought I was going mad and was maybe imagining it after reading some comments) - and I didn't prejudice him by giving any information. I simply said 'tell me which, if any, is the better sound'. He picked out the optical one, for the reasons I outlined earlier regarding the improved top and bottom ends.

So maybe we both have very fine-tuned ears. But that is not the point.

The real point is that there IS a difference. But at the end of the day, if your ears cannot perceive this difference then it's not a problem for you.

I would still be interested if anyone has any analysis data on this though! I will continue to trawl the web to look for any.

Regards

Andy

P.S. the reference track I used was 'Abacab' by Genesis

Edited February 8, 2007 by KanakoAndTheNumbSkulls

[greenmachine]

Hi-SP is 256kbps, which is also the maximum compressed bitrate for real-time recording. 352kbps has been introduced not so long ago and is available via software only. So you compared a 352kbps software encoded track against a 256kbps optical-in recorded track and preferred the latter? Interesting.

Since real-time recording requires you to adjust the recording levels manually (you didn't use AGC, did you?) even when recording from the optical-in, it might be possible that this real-time recorded track might be a bit louder or quieter than the other, which might contribute to the perceived difference.

Edited February 8, 2007 by greenmachine

[cochra1]

Hi-SP is 256kbps, which is also the maximum compressed bitrate for realtime recording. 352kbps has been introduced not so long ago and is available via software only. So you compared a 352kbps software encoded track against a 256kbps optical-in recorded track and preferred the latter? Interesting.

You're right - Hi-SP is 256kbps, I was wrong about that.

I'll clarify what I compared.

I compared a song transfered to my NH600 in Hi-SP mode via SonicStage (both from a WAV source and from a 352kbps source)

WITH

the same song copied in realtime via an optical cable (CD source).

But yes, absolutely, the realtime optically-captured sound is superior to transfering within SonicStage (either from a WAV source or from a 352kbps source). No doubt. I shall continue my trawl of the web to try and find analytical data on this.

Update

I have just tried a SonicStage transfer from an import captured in just ATRAC (not advanced lossless) at 352kpbs, and transfered to my minidisc in the original format (no further conversion). Again I used highest quality (rather than fastest).

Alas, this was still not as nice as the realtime copy, but was the best SonicStage version yet (in terms of the end product on my NH600 minidisc).

My conclusion based on my observations (and one colleague's) alone

The NH600 (I can only speak for this model) captures and converts music at a higher quality in realtime than using the SonicStage software to capture. (But I am still searching the web for analytical data to back this up for the rest of you!)

Edited February 8, 2007 by KanakoAndTheNumbSkulls

[timonoj]

You're right - Hi-SP is 256kbps, I was wrong about that.

I'll clarify what I compared.

I compared a song transfered to my NH600 in Hi-SP mode via SonicStage (both from a WAV source and from a 352kbps source)

WITH

the same song copied in realtime via an optical cable (CD source).

But yes, absolutely, the realtime optically-captured sound is superior to transfering within SonicStage (either from a WAV source or from a 352kbps source). No doubt. I shall continue my trawl of the web to try and find analytical data on this.

Update

I have just tried a SonicStage transfer from an import captured in just ATRAC (not advanced lossless) at 352kpbs, and transfered to my minidisc in the original format (no further conversion). Again I used highest quality (rather than fastest).

Alas, this was still not as nice as the realtime copy, but was the best SonicStage version yet (in terms of the end product on my NH600 minidisc).

My conclusion based on my observations (and one colleague's) alone

The NH600 (I can only speak for this model) captures and converts music at a higher quality in realtime than using the SonicStage software to capture. (But I am still searching the web for analytical data to back this up for the rest of you!)

Another point: Did you check your cd player? I mean, does it have the "bass button" enabled? I agree and think it has something to do with the CD playback... I didn't notice any difference.

[cochra1]

Another point: Did you check your cd player? I mean, does it have the "bass button" enabled? I agree and think it has something to do with the CD playback... I didn't notice any difference.

No man, it's a really nice high-end seperates CD player, very minimalist, can't even program it! Certainly no 'bass boosts' or anything like that. The optical out is just a pure digital signal of the CD. I also have an optical connection from my DVD player, and recordings to my minidisc from that are equally good.

Let me emphasise, the improvement to most people would be very subtle, not like a bass boost. Many on this forum have said they can tell no difference, although I KNOW this is down to their limited listening abilities, and I don't say that to offend them.

The improvement is in a more 'weighty' bottom end (in other words bass drums, for example, pack more of a punch and sound more dynamic), and also the top end (ie. treble) has more clarity and sparkle, and the whole sound is smoother and less coloured or 'broken-up' sounding. But again, to your average bod who wouldn't know hi-fi if it slapped him about the head with a wet fish, the difference wouldn't be recognised.

The only way to understand my point is to make 2 copies of the same song yourself (preferably one with a good solid beat), one optically and one via software. Put the tracks on the same disc. Listen through earphones, close your eyes, experience the sound in your head, the movements, the dynamics, the timbre. Then switch to the other version. Pay attention to the bottom end. Does it pack the same punch? Listen to the top end. Does it shine in the same way?

For me the answer is unmistakably NO.

And no, it's not placebo, or anything else, it's just fact.

And to anyone who has any other bright suggestions about why I might be wrong, eg bass boosts on my CD player, using low bit rates on SS, placebo effect, conjecture (that was a good one), etc: stop being so lazy and just try the test for yourself (if you can be bothered, or if not, fine). If you can't hear any difference then cool, continue to enjoy SS.

Hi-SP is 256kbps, which is also the maximum compressed bitrate for real-time recording. 352kbps has been introduced not so long ago and is available via software only. So you compared a 352kbps software encoded track against a 256kbps optical-in recorded track and preferred the latter? Interesting.

Since real-time recording requires you to adjust the recording levels manually (you didn't use AGC, did you?) even when recording from the optical-in, it might be possible that this real-time recorded track might be a bit louder or quieter than the other, which might contribute to the perceived difference.

No, Greenmachine, it's a timbral thing, not an issue of volume. I'm talking about the pure tonality and dynamics and warmth (or lack of warmth) of the recordings. Realtime recordings do not REQUIRE you to adjust the volume, they just give the option.

Look, this is ridiculous everyone. Why are you all so damn eager to argue without actually trying it and listening (closely) for yourselves? It's like some of you are scared to find out SS (in terms of sound quality) has a problem.

Actually, hears a simple way to explain it.

You know when you import a song from a CD into your SS library? You can choose between importing it quickly, or using a slower capture but reaping the benefits of a higher quality sound.

Well, real time recording is just another step up the quality ladder.

Edited February 8, 2007 by KanakoAndTheNumbSkulls

[kino170878]

What you are saying is, that the highest quality possible, besides having pcm (WAV) recording, is recording in Hi-SP at 256kps via the optical connection in real-time? Is this statement correct?

This would then imply, logically, that there would be little if any point in uploading the pcm file, downconverting to 352kps, and finally transferring the file to the minidisc player with our lovely sonicstage software....

If so, I am recording real-time from now on.

Edited February 8, 2007 by kino170878

[roem]

Well,

my first thought by reading this thread was, 'this is rubbish, there shouldn't be a difference by copying digitally either by optical in or with SS'.

But just for fun I took a CD from Kean and recorded that with my MZ-RH1 in HI-SP mode via optical in. The source was an old Panasonic CD Player with an optical out (Panasonic SL-CT570, definitely no high end device ...).

After that I used SS (Version 4.2) to create the Files to compare with. Atrac, 256kBs and with the option high quality enabled.

Then I copied them to the same HI-MD and into the already existing folder with the digital-in recordings.

I checked that the volume for the songs is about the same which seems to be the case.

Then I played the songs in shuffle mode and tried to guess what the source was. Well, actually I didn't expect to guess more than 50% right. But now after beeing right with my guess with 9 of 10 tries (20 trials in all) I'm a bit confused. It's not that the SS recordings sound worse but they sound like they have something added. The optical in recordings seem to sound a bit 'cleaner' to me. Ok, sounds strange now but there seems really to be a difference.

Maybe someone with more experience in ABX-testing should give that a try too.

best regards

Roman

PS: I didn't try if Simple Burner behaves the same way.

EDIT: After rereading what I wrote, I think I have to explain how I actually tried to figure out which song came from which source. It's not that I compared two songs with each other and picked the one that sounded better. I was listening in shuffle mode and when I thought that this song sounds different than it used to cause I was using SS (and sometimes also SB) to fill my HI-MD with music before, than I knew it has to be a song which was recorded via optical in. I was listening the HI-MD with Kean on it lately rather ofen, so I took that one for that test.

I'm fully aware that this 'test' can't be compared with a classic ABX-Test but I couldn't figure out another way to compare songs on the player/recorder without knowing in prior what I'm listening to.

Edited February 9, 2007 by roem

[roem]

I do have both *.omg files (the one created with SS and the one which was recorded via optical in) on my hd now and am wondering if they would run on another computer as well.

If so, I would offer to send those two files (around 7MB each) to someone who would like to do an ABX-test. My soundcard is just not decent enough to do that kind of testing.

But anyway, I'm not going to send those files just to anyone because they like to get a song for free (copyright issues are also the reason I don't attach the files to that post). So please provide first a link to an ABX test you did.

Roman

[cochra1]

What you are saying is, that the highest quality possible, besides having pcm (WAV) recording, is recording in Hi-SP at 256kps via the optical connection in real-time? Is this statement correct?

This would then imply, logically, that there would be little if any point in uploading the pcm file, downconverting to 352kps, and finally transferring the file to the minidisc player with our lovely sonicstage software....

If so, I am recording real-time from now on.

Absitively posolutely. I'm saying that if you use SonicStage to capture your recordings to minidisc it degrades the sound to a subtle extent, whereas realtime/optical capture results in the cleaner, smoother, more pleasing sound. I don't know why this is, it certainly isn't logical, but it happens.

[cochra1]

I do have both *.omg files (the one created with SS and the one which was recorded via optical in) on my hd now and am wondering if they would run on another computer as well.

If so, I would offer to send those two files (around 7MB each) to someone who would like to do an ABX-test. My soundcard is just not decent enough to do that kind of testing.

But anyway, I'm not going to send those files just to anyone because they like to get a song for free (copyright issues are also the reason I don't attach the files to that post). So please provide first a link to an ABX test you did.

Roman

Thanks for your contribution, Roman. Problem is, I'm guessing that even the optically-captured track was transfered to your hard drive via SonicStage. I believe that the degradation occurs at the point of capture to the SonicStage library.

I base this on an experiment I carried out whereby I took my optically-captured song, transferred it to SonicStage, and then transfered it BACK to my minidisc. The sound had degraded simply by transfering.

I believe the loss happened on it's way IN to SonicStage rather than the journey back out because SS has a tickbox for selecting to transfer out in the original form with no further conversion (which I selected). It doesn't have a likewise tickbox for the transfer IN, so some sort of processing always happens when a song is captured IN to SonicStage. It is this processing that causes the degradation (I theorise).

The best references to use for a test (besides having access to the HiMD disc itself) would be to transfer each version of the song via the headphone output (with the EQ off and the auto limiter hacked) to your PC via its soundcard, using whatever realtime recording software you have. I take on board what you say about your poor quality soundcard.

I have a nice soundcard (pro quality for music production) so I could produce the samples, and I could zoom closely on the wave graphics to compare for differences - but I don't really have any analytical software. But who would I send them to...?

Edited February 9, 2007 by KanakoAndTheNumbSkulls

[Sparda]

Just throwing it in there and guessing, perhaps the one made by SonicStage sounds different is because of your soundcard? Doesn't matter if it is a good one or not, the point is maybe there is a setting that is affecting the sound? And if Optical In sounds better, then by all means... just use it!

[1kyle]

Here's an interesting test

I plug the Optical Output of a CD player (quality Marantz Gear) into the optical input on my jbs980. I record in SP (legacy) on to a MD in the MD deck.

I also plug the Optical OUT of the deck into a HI-MD RH1 and record at HI-SP 256, 362 and PCM. The Deck is just acting as an "Optical Pass through" although I accept there might be an extra D/A step (or even 2). The MD deck is top notch quality so I wouldn't expect any signifiant degradation in the extra D/A step(s).

Doing it like this allows me to concurrently record to a MD IN the deck AND record "Optically" to another device at the same time so I don't have to do 2 "Real Time" recordings for testing.

Now playing the SP recorded disk into High End gear with speakers using the optical output of the jb980 deck into the optical in of my amplifier gear sounds FIRST CLASS -- Maybe if you were hard pushed to it you *might* be able to tell the difference between the CD and the MD. It's difficult for me to tell as I know where the recording came from.

I can only imagine what an optical HI-SP output must be like -- What a miss here SONY.

It's difficult to get true optical output into a high end piece of gear without using a computer as I can't think of any HI-MD units that have an optical output. So the quality here will depend not only on SS but your HARDWARE as well. Often Computer gear has relatively cheap components installed like bottom of the range CD writers and HIDEOUS sound cards and systems that make the old fashioned "Ghetto Blasters" sound like a high quality classical Symphony Orchestra like the LSO (London Symphony Orchestra).

Finally you have to plug the ANALOG output of the RH1 into an amplifier to test it --whilst VERY VERY good indeed at HI-SP this is still not equivalent to an Optical output going in to the Optical input of a serious quality amplifier.

As I don't have any Optical OUT on the MD players (RH1) I have to use the analog out for playing through the high end audio deck.

All 3 codecs sounded first rate (PCM, HI-SP 352, HI-SP 256). Remember these disks were recorded OPTICALLY.

Making a CD again from the PCM file uploaded to the computer sounded identical to the original CD (as it should with decent gear).

I came however to the suprising conclusion that SP recorded Optically was actually the better than any of the HI-MD codecs including PCM on playback but that could be due to the fact that the play back quality of the jb980 with optical ouput is far better than the standard analog output on the RH1 good though that is.

However analog output into High end gear with HI-MD recorded at 352 and even 256 won't disappoint. It's still surprising to me that legacy SP via optical out and in still sounds the best -- maybe I'll keep my SP disks around for a while yet.

In a studio where you can get the PCM into other gear then I'd suspect this would be the best and probably easy enough to hear in that environment.

However on the road you are going to have to rely on using headphones or plug in to analog gear and in these cases where you don't want to upload to a computer HI-SP @ 256 is your best bet unless you have legacy gear where good old "SP" is still fine (60/74/80 min discs).

Nothing I've wriiten here disappoints me about the RH1 except lack of optical output. For recording on the move or even as a playback device the RH1 more than satisfies.

Cheers

-K

Edited February 10, 2007 by 1kyle

[The Low Volta]

@1kyle: if you're going to compare formats and the way you put the music on MD (SS or optical) at least use the same setup for listening to them... use the RH1 and the best possible headphone you have in stead of a deck with optical out and a portable player with analog out

[roem]

Thanks for your contribution, Roman. Problem is, I'm guessing that even the optically-captured track was transfered to your hard drive via SonicStage. I believe that the degradation occurs at the point of capture to the SonicStage library.

I base this on an experiment I carried out whereby I took my optically-captured song, transferred it to SonicStage, and then transfered it BACK to my minidisc. The sound had degraded simply by transfering.

I believe the loss happened on it's way IN to SonicStage rather than the journey back out because SS has a tickbox for selecting to transfer out in the original form with no further conversion (which I selected). It doesn't have a likewise tickbox for the transfer IN, so some sort of processing always happens when a song is captured IN to SonicStage. It is this processing that causes the degradation (I theorise).

The best references to use for a test (besides having access to the HiMD disc itself) would be to transfer each version of the song via the headphone output (with the EQ off and the auto limiter hacked) to your PC via its soundcard, using whatever realtime recording software you have. I take on board what you say about your poor quality soundcard.

I have a nice soundcard (pro quality for music production) so I could produce the samples, and I could zoom closely on the wave graphics to compare for differences - but I don't really have any analytical software. But who would I send them to...?

I took your point up that SS might do some conversion to the file that is transfered to the computer. The idea behind my test is that the sound should get distorted when the conversion is done over and over again on the same track. Here is what I did:

Initial Step:

-Made a copy of a song with the MZ-RH1 via optical in

Step 1 (done with the help of SS):

-Transfer the song with the help of SS to the computer (conversion)

-delete the recorded track on the HI-MD and replaced it with the one already transfered/converted track again using SS

Step 2 (using the file explorer/recorder functions):

-copy the *.HMA track from the HI-MD disc

-format the HI-MD

-make another copy of the same song via optical in

-replace the new *.HMA track on of the recording on the HI-MD with the one previously taken from the HI-MD

-go back to Step 1

I did that 4 times. So the same song was converted 4 times. Maybe this are not enough conversion processes to show distortion cause I couldn't spot any. What I observed was that the file size increased (had about 400kb more in the end) and that the song gained additional time (4s). This is probably due the toc manipulation though and not because of the conversion process.

Roman

PS: if someone wants to retry that test you will have to do addtional steps in order to prevent SS from crashing. When the file size grows you have to copy additional tracks to the HI-MD before you can copy the track back to the computer or SS will crash and the HI-MD needs to be formatted before it can be read/written to again.

[cochra1]

I took your point up that SS might do some conversion to the file that is transfered to the computer. The idea behind my test is that the sound should get distorted when the conversion is done over and over again on the same track. Here is what I did:

Initial Step:

-Made a copy of a song with the MZ-RH1 via optical in

Step 1 (done with the help of SS):

-Transfer the song with the help of SS to the computer (conversion)

-delete the recorded track on the HI-MD and replaced it with the one already transfered/converted track again using SS

Step 2 (using the file explorer/recorder functions):

-copy the *.HMA track from the HI-MD disc

-format the HI-MD

-make another copy of the same song via optical in

-replace the new *.HMA track on of the recording on the HI-MD with the one previously taken from the HI-MD

-go back to Step 1

I did that 4 times. So the same song was converted 4 times. Maybe this are not enough conversion processes to show distortion cause I couldn't spot any. What I observed was that the file size increased (had about 400kb more in the end) and that the song gained additional time (4s). This is probably due the toc manipulation though and not because of the conversion process.

Roman

PS: if someone wants to retry that test you will have to do addtional steps in order to prevent SS from crashing. When the file size grows you have to copy additional tracks to the HI-MD before you can copy the track back to the computer or SS will crash and the HI-MD needs to be formatted before it can be read/written to again.

Cool. Four times should be more than enough though (and it may even be the case that the degredation only happens the first time a track which has come from an outside source is imported into the SS library; multiple transfers back-and-forth may make no difference, but I don't know).

If one can tell no difference between a realtime recording and the same song grabbed in SS (after just one conversion) then that means SS is fine for that person (as it is for most people).

The test I (and a colleague) did gave blatant results to us (with experienced ears), but I'm sure if the average person gave them a listen they would hear no difference. I mean, to illustrate the point, I know people with mp3 players who use quite low bitrates to get as much crammed onto their players as possible, who say they can't tell the difference in quality. Mosty people just aren't that perseptive!

Anyway, the bottom line is the difference is there (less bass weight, fractured top-end), but the only thing that matters is if you can hear it.

Watch this post - I will be attempting to produce graphical representations of the effect of SS as compared to realtime captures.

Edited February 10, 2007 by KanakoAndTheNumbSkulls

[roem]

Cool. Four times should be more than enough though (and it may even be the case that the degredation only happens the first time a track which has come from an outside source is imported into the SS library; multiple transfers back-and-forth may make no difference, but I don't know).

If one can tell no difference between a realtime recording and the same song grabbed in SS (after just one conversion) then that means SS is fine for that person (as it is for most people).

The test I (and a colleague) did gave blatant results to us (with experienced ears), but I'm sure if the average person gave them a listen they would hear no difference. I mean, to illustrate the point, I know people with mp3 players who use quite low bitrates to get as much crammed onto their players as possible, who say they can't tell the difference in quality. Mosty people just aren't that perseptive!

Anyway, the bottom line is the difference is there (less bass weight, fractured top-end), but the only thing that matters is if you can hear it.

Watch this post - I will be attempting to produce graphical representations of the effect of SS as compared to realtime captures.

The copying and replacing sheme I did was necessary to make SS think that the file to transfer is actually an optical-in-recording although it wasn't. So the same file was converted (degraded?) 4 times. I couldn't hear a difference to the first optical-in recording after those 4 conversion processes though. Whatever SS does to the files it doesn't seem to degrade the quality of the sound imo.

The difference in the optical-in and SS-recordings we claim to hear could be due different atrac versions used in the hardware of the MZ-RH1 and in SS. Maybe there are some tiny differences we might be able to spot. Maybe there is just some side-effect I'm not aware of ...

I'm still offering the two atrac files for ABX-testing (one recorded via optical-in the other recorded with SS).

Roman

[cochra1]

The copying and replacing sheme I did was necessary to make SS think that the file to transfer is actually an optical-in-recording although it wasn't. So the same file was converted (degraded?) 4 times. I couldn't hear a difference to the first optical-in recording after those 4 conversion processes though. Whatever SS does to the files it doesn't seem to degrade the quality of the sound imo.

The difference in the optical-in and SS-recordings we claim to hear could be due different atrac versions used in the hardware of the MZ-RH1 and in SS. Maybe there are some tiny differences we might be able to spot. Maybe there is just some side-effect I'm not aware of ...

I'm still offering the two atrac files for ABX-testing (one recorded via optical-in the other recorded with SS).

Roman

You could be right mate (about different ATRAC versions).

By the way, my own testing has failed at the first hurdle(!) as my software can't analyse accurately enough. I did zoom in close on the waves and saw a very tiny (negligible really) delay between the left and right channels on the SS version in comparison to the optical version. But this is not really usable data because I didn't repeat the test to see if the results would be duplicated (no point as it is really negligible anyway).

I did try some frequency analysis (of literally a split second of a song which contained a hard bass drum beat followed by a hihat - perfect for testing). But unfortunately the software I have is not up to the task as it wasn't really designed for this purpose. The resulting graphs and data were at too low a resolution.

I shall leave this thread now, but I really appreciate everyone's suggestions, thanks, and see you on another thread!

Edited February 10, 2007 by KanakoAndTheNumbSkulls

[raintheory]

I'm not disagreeing with anyone (because I think the issue of sound quality is inherently subjective) but I would love to see a frequency analysis of the same file both ways overlayed...

Edited February 10, 2007 by raintheory

[roem]

I'm not disagreeing with anyone (because I think the issue of sound quality is inherently subjective) but I would love to see a frequency analysis of the same file both ways overlayed...

If I would know how to do a frequency analysis I would have done that already. I had a look at the files with audacity just to see if I can spot some differences.

On top is the SS conversion and below the optical-in recording.

There are differences which can be seen by eye but I have no idea if they can also be observed by ear. If someone wants the the two files to make this kind of frequency analysis, let me know.

Roman

[Sparda]

These graphs really don't say much. You could transfer the same song twice via optical in and it would should up differently.

[cochra1]

These graphs really don't say much. You could transfer the same song twice via optical in and it would should up differently.

You're right. What I was looking for would not be affected by these slight differences though. I was looking for more obvious peaks and troughs in certain frequency ranges (eg. for bass around the 60-200hz region, and for treble around the 8khz+ region) which would be dupilcated on multiple 'takes'. But you are quite right, sampling from optical (or even direct from a CD drive) many times would produce many subtly different results.

The only exception would be using software which reads and duplicates EXACTLY the data on the audio CD (which the likes of SS, Nero and Windows does not do). Even then the conversion process into ATRAC would produce differing results each time.

But the trick of a good test is to duplicate it and look for common results in each trial.

My graphs were much like the ones submitted on this thread - not really specific enough (but thanks anyway to Roman).

I must admit also, that when I carried out the analytical test myself, I could not hear any differences when comparing the different samples on my PC (like Roman said also). I am so tempted to say (as was suggested earlier) that the whole thing has been a placebo - but how come my colleague agreed with my observations (see earlier post)..? And why would there even be a placebo effect given that I had no prejudices about SS being inferior to optical? I'm confused now by my own contradictory observations and am going to get a stiff drink.

Edited February 11, 2007 by KanakoAndTheNumbSkulls

[Sparky191]

...

The only exception would be using software which reads and duplicates EXACTLY the data on the audio CD (which the likes of SS, Nero and Windows does not do). ...

I'm only seeing this thread now. I'm confused. Why parallel SS (Audio App) Nero (CD/DVD bruning) and Windows (an operating system). They are completely different types of software. Do you mean you encode/rip using each of these?

Edited February 4, 2008 by Sparky191

[netmduser]

I occasionally have found the same problem with lp2, not with all SS transfers as well, which is puzzling. do you have other mp3 decoders installed ie. like ffdshow? I will have to agree in general with your statement.

I removed ffdshow and found the sound to be better, but I haven't been able to spend much time figuring this out. Anyone have suggestions? With lp2 recording on the recorder itself I found the quality to be much better.

Anyone else have experience with lp2?