HOLY sh*t (by Bill Evans)

[crisnee]

Flecko, Thanks, that's great.

How about if I send you a couple of tracks from the "Debby," album--the original cd transfer, for you to compare to all the others? Tell me which tracks and if you think 320 mp3 would  be good enough for this comparison (I can also send flac, there just much bigger of course).

-Chris

[PeterSt]

Careful Adrian, any answer can be the wrong one here. Haha.

[crisnee]

Careful Adrian, any answer can be the wrong one here. Haha.

Hey, what's that supposed to mean? I just want you to know I'm not laughing!

Ok, now I am, but I don't know why.

-Chris

[PeterSt]

Oh, I just wanted to prevent (either) people to be embarressed about MP3 stuff. And somehow it slipped through my mind to offer those HiRes tracks I have here, as MP3. Yes, they are smaller ...

But let me tell you that you won't find any post or anything anywhere from my hand that tells about MP3 being "nothing". It's all about other things ...

Peter

[crisnee]

Oh, I just wanted to prevent (either) people to be embarressed about MP3 stuff. And somehow it slipped through my mind to offer those HiRes tracks I have here, as MP3. Yes, they are smaller ...

But let me tell you that you won't find any post or anything anywhere from my hand that tells about MP3 being "nothing". It's all about other things ...

Peter

Thanks Peter, that's down right decent of you (just not very clear, although I did cast a thought in that direction).

By the way, re mp3, I don't see why audiophiles have so much against the poor little darlings. And the same people rave about the historical performances of someone like Klemperer and orchestra, and other classical artists who's best work predated the fifties. For some reason quality of recording makes no difference in those circumstances "because it's all about the performance."

Personally I'll take a high bitrate mp3 over any other kind of common copy of music from the past. And I'll be very happy with mp3s of any music that I can't get or afford in a redbook or better format. I have some mp3s that sound so wonderful that I would even consider converting all my less good sounding music to mp3 if they too could sound as wonderful, including those Klemperers.

I've put up my steel umbrella to shield me from flying tomatos and eggs.

-Chris

[PeterSt]

I am borrowing one of your unbrellas now Chris ...

First of all I am fairly sure that nearly everybody wouldn't allow himself to express openly about MP3 not being bad at all. So, even if you'd think it is okay, it is not a good thing to openly say so. I don't do that either, although quite some years back I did, having joined a well setup blind test with pieces of Three Wishes from Amused to Death (Roger Waters). There was no way I could discern uncompressed from 256 MP3, as was nobody else (hundreds of competitors). I must say that by now I will be able to, just because I know the character of the sound, plus having a way better system than back then. And that character is not for the worse ...

Litterally I have been saying in quite some occasions that the stage of MP3 is wider, as an argument to "how wide the stage must be ?", obviously implying that wider not necessarily is better, referring to MP3 doing just that.
But does it mean that MP3 is worse ?

And so we have once again one of these very controversial subjects which pass from my hand. It doesn't happen often lately, but this one is worth it I think. It is brewing for so many years now, that it is time to do something with it. I'm sure it justifies a new topic, but I hardly ever do that. Too explicit or something.

I won't make it too long for now, but Chris, think about "Teresa's deseases" and what her solution was to play "lowres" uncompressed WAVs (or AIFs) ...
MP3 is listenable ...

Of course everybody starts shouting that someone is crazy (and full of deseases), but I am not so sure about this. I have (and always had) reasons to believe that MP3 might be better than uncompressed. And now, in the era of "HighRes is nothing", it kind of fits a few technical theories.

I may have referred to it one time only (somewhere), but depending on the DAC used it may have occurred to "you" that the music going along with a movie may sound better than the normal Redbook CD material. This isn't something anyone would notice much, and then I mean the opportunity for it; You'd need your audiophile DAC to play the movie music through, you'd need two channels only - you'd need the proper movie playback software - and I can tell you that this is not for everybody. This time I for once am NOT referring to the NOS1, because it already works with a 16/44.1 (48 of course for movies) NOS DAC (better put : having the extra sample rate and bit depth doesn't help here).

Two weeks or so back I wanted to play my first King Crimson, and I only had it in 256 MP3. Well, it sounded as gorgious as that other stuff from that era can sound (what is it ? 1970 ?) and this time I set myself to make something of it. To work it out. To make it "science". Ok, let's say it lasted until this post that I'm really doing something about it, but man, I wish I had more time.

Let's say that it won't last long until XXHighEnd will contain a downconversion from whatever uncompressed to MP3. Maybe 256 is even better than 328. I don't know yet. But it is about some very different means of filtering ... something really nobody thought about. But then it already takes some steps to believe in real good NOS in the first place, and now this will be about an even more strange combination : first use MP3 as a good means of filtering, to next upsample that sky high. My theories are now that this could be the best of all worlds.

Before I really implement this I'd have to measure it, because if it doens't measure well it won't go anyway. But I have the hunch this will outperform everything and all ...

So, this is the real story behind my earlier post. Call me crazy.
Peter

[GerardA]

Wow, that's a nice new topic!
I listen to MP3's a lot because sometimes it's all I have, but don't really know what the signature of MP3 is.
A lot of times I think with music from movies, why is this not happening with CD?
So what is this thing making them better?
Is it the focussing on important details and skipping on the low-information part like Philips talked about in their DCC-days?

Well anyway XXHE does a very good job on MP3's!

[Flecko]

Flecko, Thanks, that's great.

How about if I send you a couple of tracks from the "Debby," album--the original cd transfer, for you to compare to all the others? Tell me which tracks and if you think 320 mp3 would  be good enough for this comparison (I can also send flac, there just much bigger of course).

-Chris

I compared the first two tracks from your original transfer to the 2010 remaster and the 1992 Analogue Production Remaster. I like the original transfer the most. It is in tonality closer to the 2010 remaster, the AP Remaster seems to add some tubesound (distortions) to the mix. The original transfer is the most clean of all and due to its high DR, very natural in sound. AP remaster is slighly more exciting but that is a kind of coloration. The original transfer is it! Fantastic music and sound. Thanks again chris.

I listen to MP3's a lot because sometimes it's all I have, but don't really know what the signature of MP3 is.

I would say MP3 lose details, especially in the highs. They sound more dull than an original cd or flac. They do not sound bad, if they are made well. Just not as good as the original.

[GerardA]

I would say MP3 lose details, especially in the highs. They sound more dull than an original cd or flac. They do not sound bad, if they are made well. Just not as good as the original.

Well then Peter must be able to put the sparkle back in!

[crisnee]

Re mp3, just in case folks aren't aware of it, one of the major compression areas in compressed sound is right after the attack or loudest part of the signal, because the ear can't adjust quickly enough to hear the lower level. (fodder for the scientists among us (Peter) to use  for the betterment of all XX users).

-Chris

[crisnee]

Re nothing whatsoever. Well actually it has to do with sending files to Adrian (Flecko). Bear with me.

Dropbox is a free program. It's main purpose is online storage--2gb free, and keeping all you computer's data synched. It becomes a folder on your desktop with various subfolders anything you save in it will appear on your other pcs (and Macs I think).

So what's that got to do with Adrian. Well Dropbox has several other great functions, one is that it contains a Public Folder. Anything you place in it can easily be sent to anyone--Adrian (actually a link is sent via email and they click on it to download the corresponding file). Pretty useful for forums (tests and listening--I'm not a fan of rapidshare and the like) such as this one.

But it's even more useful for people who have Dropbox installed because they can share folders. In other words say a few of us here wanted to test some files, we could place them in the shared folder(s) and each can test/add them to the folder anytime. Also good for you Peter when dealing with XX problems for examining files.

You can also access your files from anywhere via a browser such as Firefox or IE.

I've been using Dropbox for years, actually I'm an original adopter, but I'd never used the sharing function until I sent the files to Adrian--which made me think how useful it could be in a forum like this. It's really simple/intuitive because it's part of your file system; I save all my docs to it, plus my password vault and it's a great safe (encrypted) backup too.

So if anybody's interested, google Dropbox, or if you feel like doing me a favor, pm me with a request for an invitation (include your real email address) and I'll get some extra free storage space for the effort.

-Chris

[crisnee]

Well Peter you may need more than an umbrella, how does bunker sound--the mp3 version anyway? (That was a dumb joke and should be compressed).

"Teresa's deseases"? What are you referring to with that?

Wow that's some interesting stuff. It never occurred to me that compressed could sound better, but it makes a certain kind of sense (intuitive). Would it have something to do the psychoacoustics (I think that's the term I mean)? And the technicalities of hearing (maybe compression could prevent a kind of hearing overload? your ear has less info to respond to and therefore responds to what is there better? Or the better the recording (the higher the res) the more extraneous info is included that somehow messes with the essential musicality or sound we hear when we hear live instruments--because mics and ears are not the same. I'm obviously just speculating. I'll stop now.

I'm not sure you should be volunteering these kind of thoughts being in the business you're in. I wouldn't be offended if you deleted your mp3 post, and mine.

-Chris

[PeterSt]

Hi Chris,

Teresa is that woman over at CA who ditched all her CD rips and can stand hires only. And MP3's ...

And no, no way the MP3 thing is to be subjective. If it were in my view, I indeed wouldn't have dare to post about it like I did.
It is just a technical thing. That a (brickwall etc.) filter is needed for Redbook. You know, that filter wich always compromises. Arc Prediction is one which works totally different from all of the others. This "MP3" based filter would be another new one.

Hey, I must have something to do here !
Peter

[crisnee]

And no, no way the MP3 thing is to be subjective.

I agree that it's not subjective, or if it is, you're in trouble. You didn't think I meant it was, did you?

The question though is, why is less, more. What could the cause be? That's what I was speculating about, and wonder if you have, or anyone else has, who either thinks that mp3's are the equal of or superior to uncompressed material (in terms of quality of music reproduction to a human's ears).

Do you know what Teresa's complaints were about standard rips, or what she liked better about hi-res and mp3's?

-Chris

[PeterSt]

Well, let me tell first that it's all a bit tough to explain it from the start, also referring to the fact that in our forum here I never talked about this. So, some things must be taken for granted I guess;

Let me paraphrase the thinking of many that they want to perceive the higher frequencies beyond the official audioband, just because a. the frequencies are in there for HighRes and b. because it may be so that we can perceive that indeed (not hear, but perceive by unknown/various means).

Ad a.
There is hardly any HiRes to be found where the higher frequencies are not plain false because it is distortion. This by itself caused by dozens (?) of errors made at creating the HiRes. At least all the DVD-A and SACD are so, and if not I have yet the first to see which is okay.
To keep in mind : I have been saying for over two years that HiRes doesn't sound good at all, and since some 10 months or so I started measuring it. By now everybody can and does, and it has become a sheer fact that it is all BS. Total BS.
Not a hoax per se, but errorneous production. The mastering people didn't know.
Notice that the current new productions should be excluded from this, but there too sufficient examples exist it fails.

Ad b.
Since a. is the situation for a fact, we can wonder what people "perceive" for the better when they like HiRes over Redbook. They do ? fine. But they like distortion apparently.
Mind you (Chris), this is nothing much different from a whole world screaming about failed HiRes because they now can look at some stupid graphs while at the same time those same people keep on telling it sounds better to their ears, while when they are proposed to come up with fine examples (as per your CA question) there are no answers.
People just shout around (not in our forum !!).

Am I drifting off ?
No.

What I proposed (by a sort of indirect math) not so long ago, is that there CAN'T be a way we perceive good quality high frequencies, but IF we do perceive high frequencies, they are the wrong ones. This is just proven fact because the wrongness is proven. It only needs the combination of what people "state" and how the facts are from the other angle. So, what I say is that we better do explicitly NOT perceive those distortions, which we can't officially hear anyway. Or in other words : I set the limit to 16KHz, which is suffuciently good for everybody who can afford a fine system (meaning, when you're old enough to afford that, your hearing will have degraded "sufficiently").
Notice that the math which is the base for quite a few more things is left out now, but what this came down to is that this 16KHz (actually close to 17KHz) is reproduceable without stepping distortion. Above that it goes wrong.

Let me add that all is derived from the noticeable stepping distortion which starts to emerge somewhere in the 4KHz range, unfiltered (!). So, under there there is no problem, but above that it becomes noticeable. Let's say in brief that 4 (times upsampling) x 4 = that 16 Khz, and the 16KHz is still "doable". With this we must keep in mind that 16KHz is close to unrestorable (for the stepping distortion), with reasons which would make this story too long, but in any case I am always talking about "unfiltered" when it came down to the usual filtering always needed, and which creates the dreaded ringing (or see my graphs elsewhere).

Where Arc Prediction theoretically fails, is that images will remain beyond the audio band. This is because in normal circumstances the captured frequency is higher than the bandwidth of Redbook allows for (22.05KHz). Thus, spit out the higher frequencies from your whatever instrument, and it is doomed to go wrong later, during playback.
Here too, it shouldn't be a big deal, because those now false frequencies (remember, they are not filtered out by Arc Prediction) are way down soon. This is (not coincidentally) the limit of around 16KHz of original frequency where all is so low already that it at least shouldn't disturb any amplifier. But, if you'd look at a, say, 21.05KHz tone, the image occurring at 23.05KHz is only a few dB down.

The remainder (or base) of the story is that by no means we'd want this ringing. It is totally devistating (but sadly one can only compare with a DAC which first does nothing -> NOS). There is no (NO) way, decent filtering (not leaving images beyond the audioband) is able to compensate for that,, *if* it can be perceived at all as a positive. Notice though that I'm always assuming 4x upsampling first, and which I explicitly tried on the Mac the other day (because no Arc Prediction is available there). So, NOS alone is not enough; it must be able to do 192 at least (per my math), while that immediately needs the higher bit depth in order to not have "stepping" distortion in the amplitude domain.

And so, my story should be clear;
Prevent the frequencies above 16KHz to be there in the first place, and we should be good.
And yes, 128 MP3 limits all to 16KHz ...

I am not exactly saying that we should use 128 MP3, but merely that by some means without ringing (and I suspect MP3 "filtering" does not do that at all) we should limit the frequencies to 16KHz. All it takes now is measuring MP3 (at various bit rates) to see what comes from it, while I already suspect the reason why MP3 -not subjectively- sounds better.

Let's keep in mind that this is MP3 in the (my) NOS environment, and including Arc Prediction (8x or 16x in my case). This is always uncompareable to either 44KHz NOS or any OS DAC.
All 'n all it won't be MP3 I am hunting for, but merely the 16KHz limit, with MP3 being more or less the proof that it is going to work. When that really directly works, it's done. It would be a filtering means that does not ring.

It is all rather complicated, but some by now available pieces of the (yet quite unkown) puzzle help a lot;
Think about Debby et al. Why does it sound so crazily good ? Do I need to say more ? think about it.
It is without doubt that everybody owning an NOS1  has the same experience. Right, we could say that thus the NOS1 flaws to begin with, but now I really have to point out that the first who isn't raving about the NOS1 has yet to come. Plus I have some ears myself.
I just think it can be better again, by applying "tweaks" like this. Strange tweaks maybe, but why not.

Why not ? well, if something like MP3 really is going to be used, while it tears down on the  dynamics or something (as was told earlier), no, then it won't be a good thing I'm sure. But in that case it is a matter of rewriting the MP3 codec concerned, so that it doesn't cut down on *that*.

Peter