It all boils down to this:

You can't take out something that simply isn't there in the first place, period.

When the details and nuances of an original recording get lost in a conversion to MP3 (or any other format involving signal compression of the audio spectrum) you just won't get those back. They're gone. Poof. Forever. Thinking you can somehow "get them back" by some high-tech gimmick action is just hopelessly delusional and mere marketing BS.

Some months ago I had the chance to be at a "real" mastering studio (AcousTech) and listening straight to a real master tape through a real mastering system, man, I swear not even the best piece of analog playback I've auditioned comes even close to what I heard there. (a $100.000 DSD system comes close, though). Which makes Creative's claims even more ludicrous for all counts.

The horribly distorted crap that passes as music in 90% of all recordings that make the current Billboard's Hot 100 is another big topic on its own. MP3'ing these is just adding insult to injury. But for now, suffice it to point ya to an oldie but goodie link on the subject:

The Death Of Dynamic Range

Frankly I am suprised you spilled so many chars on this.

I made a filter that increases the artistic merit of any music played through it. It makes Pharrel sound like Jimmy Rushing. Christina Agguilera suddenly competes with Dame Joan Sutherland.

And what's funny is these people were the same ones that said that MP3 was CD quality.

When is your sample coming Danny?

I don't disagree about the sorry state of modern music mastering, but it is wrong to claim that it's a recent phenomenon. Since the beginning of radio, pop music has always been engineered with the limitation of the method of transmission in mind. For example, all the early Beach Boy's music was produced to sound best on mono, AM radio, as was most (if not all) of the chart singles of the time. (To this day, Brian Wilson insists he prefers mono, but that's a whole 'nother story).

Listen to any of Phil Spector's "wall of sound" pop recordings and you won't hear much dynamic range either.

Even if a record engineer doesn't compress the hell out of a track, the radio broadcaster sure will before they put it on air.

Pop music was and always will be made with the lowest common denominator in mind.

Good work Danny! I saw a product release for this technology I think almost a year and a half ago, and wondered how long it would take our community to give the technology a "once over".

As always, it is much easier to "market over" a problem than actually fix it. We as audiophiles take offense, but the sad reality is that the marketing hype will work on the vast majority of American consumers.

My problem is not with the big corporate audio companies per se, but rather it is with the low standards of these companies. Have some pride in what you do, and use your resources to produce superior products, as opposed to building mediocre products for estimated customer metrics, and letting the marketing team pick up the slack...

While I understand the urge and even perhaps need to write an article like this, I'm a little put off by all the Creative bashing that seems to be going on. Has Creative's marketing department gone overboard? Sure, but then this isn't new or even isolated. I think everyone (or almost everyone) who comes to this site knows that the X-mod won't be magically restoring all their music to the way the original artist intended to to be heard. Keeping this in mind, however, there is nothing that inherently suggests that the X-mod can't be a good thing.

I think there are plenty of audiophiles who swear by oversampling DAC's. I know there are plenty of home theater people who believe in the power of postprocessing and upscaling for video. Now these might not be the best comparisons, but doesn't the X-mod at least deserve a listen before it gets panned. I hope that someone actually listens to this thing and then has the cohones to admit whether they truly like how it sounds in their system.

Would it be great if all audio were perfectly mastered? Of course. Is it important to be aware of trends in the recording industry that ruin musicality? Sure. But just as we don't buy a piece of equipment because it looks great on paper, I don't think we should bash something before giving it a listen. If I hear the X-mod and it sucks I'll be the first one to say so, but I think we at least owe it to our ears to hear components first. Here's to hoping for some listening impressions when this thing hits the street.

I'm working on a device that will turn the lint in my pocket into a million dollar bill. Stay tuned for more info!

Hey Audio Newb,

On the Creative site (link now posted above, sorry!) there is a "before" and "after" demo of the software. Obviously, the "before" sounds much better and even the surround sound feature is fun.

I have no problems with EQ and other sound manipulation devices -- in fact, I believe high end EQs like the DEQX and TACT are the future -- but when the simple answer to the MP3 issue is to rip in lossless, then products like this are just ridiculous.

Also, the biggest problem with devices like this and generic EQ products with features like "concert" and "rock" and "easy listening" is that, obviously, they don't apply to all music. While the Xmod may sound good on White Zombie (which it probably would), it's going to ruin an track with good mastering. Does anyone ever switch back and forth?

And that's why upsampling DACs are not the standard. Some tracks it really helps, other tracks it murders. Who is going to catalogue every track that benefits from EQ or upsampling? George Louis, probably, but that's neither here nor there.

The best way to make music sound good is to keep it simple. Speaks, amp, source. Done. Products like this just confuse the hell out of normal people and put them $199 away from having some Paradigms.

I think that most audiophiles have a mental block regarding ultra high quality mp3 files... What do I mean by ultra high quality mp3 files? Lame 3.97 encoder, at 320k, highest quality setting. Every time I play these Lame encoded files, eyebrows are raised as most of the time THAT files is preferred to the original cd.

Yep, you heard correctly...

(me: waiting for the onslaught of fan mail...) :-)

Danny,

I will issue you fan mail 99.9% of people who listen to music would not be able to tell the difference between a 192k to 320k sampled digital audio file. Most of our equipment is just not that good to resolve that much information. Anything lower then 192 you can start to hear a difference on most systems. I've tried this out on a $1000 system and a $20000 system. I listened to a $100,000+ system and I would almost bet that anyone could tell the difference, with a cd to cd test, at least with the 192 probably not on the 320.

I will, however, poo poo you for slamming a product which you have yet to listen to. I know it is an opinion piece but don't base it on heresay. I would value your opinion more if you gave it a true listen. It only cost $80 buy one, try it out, and bring it back if it sucks, which it probably is going to, but at least you tried it. Or just write it off as a business expence and use it a cool party trick. Oh yeah and how can a LAME encoded file sound better then the original CD? How can it add something that isn't there originally? If that is so then maybe the X-Mod might work.

BTW Creative actually has some of the better decoding chips in their Digital Music players. They were one of the few sites which state their S-N ratios and Harmonic Responce on some of their products. As an audiophile you should respect that. iPod Nano vs. Zen Nano with the same headphones the Zen is better. Better compatibility also.

Josh,

"The best way to make music sound good is to keep it simple. Speaks, amp, source." True that. K.I.S.S principle everytime.

@ Fred,

w00t! Fanmail! :)

the piece I wrote regarding the X-Fi was more of a "first look" then a review - obviously, I don't have the unit on hand to experiment with it; rather, I opined on the prospects and the shear marketing BS these types of releases usually bring in toe (FYI, I said similar things about Apple's boombox, err... Hi-Fi lest you thought I was a covert Apple operative).

Having been around computers for a long, long, long, long, long time, (did I say long time?), I know most of what's important to me regarding Creative. To date, most of their peecee sound cards were accompanied by a plethora of bloatware that would install itself even though I never asked for it - in some cases spectacular incompatibilities arose to which Creative's support only offered to turn the other cheek, as in blame someone else's sw or hw for the issues. As a result, I stopped using Creative products and switched to M-Audio's Revelation 7.1 which at the time (and still today) is a fine, if inexpensive audio card for your pc or mac. No bloatware, no office suite, no window dressing to be concerned with - all works as it should. Add to that some pretty significant controversy a few years back when Creative released sound cards claiming to offer 24bit resolution, when in reality it was nothing more than good ol' 16bits. Only after a great many enthusiasts websites offered proof to the contrary, did Creative actually reverse coarse though not much more.

Alas, you can imagine my enthusiasm and expectation when Creative claims to have found the holy grail to those evil sounding CD and mp3 files...

As to why listeners and experienced audiophiles would prefer a Lame encoded 320k max setting mp3 file to the original cd, I don't know - to me the Lame encoder offers a more present, focused and less muddled image then the original, though it depends on the source material; sometimes the improvement is rather dramatic, other times its about equal. That is of course if they can remain "open minded" (oh boy, my favorite 2 words!) about the process involved and most importantly, the result. I am working on an article with one of the Lame developers and I hope to provide more answers to these and other questions...

PS: offering S/N ratio and harmonic distortion figures to consumers who haven't got the faintest clue what these numbers mean, ie. without the proper context, is really somehow meaningless imho...

;-)

Don't get me wrong I hate the man, i.e. big companies with too much money to spend on marketing, as much as the next guy. But even a blind squirrel finds a nut once in a while.

I have been using computers for a while also, do you remeber when you got a new audio card and had to physically set the irq setting on the card with jumpers and then use Norton to try to see if you have a conflict, because you didn't get it you would have to try again. Setting up a modem was like trying to . . . well you know. Back in the day it was exciting to get an actual Sound Blaster card and actually have it work with your games. Because for the most part the "knockoffs" would always have issues. There was nothing worse then playing caveman "ug"lympics with no sound, it was bad enough in CGA graphics and it had to load off of 5 1/4. The alternative sound cards have gotten much better since those days though.

Oh well I look forward to the LAME article, har har. If for nothing else you are doing a good job in getting a reaction from the masses. :^>

Danny,
It doesn't suprise me that casually listening, someone might prefer 320 mp3 to lossless. The MP3 format did a masterful job of preserving fidelity compared to how many bits are disposed of. With 320, you are getting closer to lossless in terms of bit retention, and you still get all those tricky psycho-acoustic effects that are built into the format to compensate.

While someone might like the effect, it is not strictly accurate compared to the source. Like any other kind of filtering though, over time MP3 effects get boring as it is applied to every cut in the same way, robbing music of a bit of it's uniqueness.

I took a listen to the samples. The before samples of each of these sounded terrible. Worse than most MP3s by far. So it's not suprise that the after sounded better. Also the levels were raised about 50%.

I actually like the idea of using the x-fi for games as a 3-d generator for headphones. I also wonder how it would be for a USB dac overall?

I wouldn't recommend the virtual surround sound algorithm on the X-Fi because I have tried it with my headphone system. It adds distortion to the sound, and doesn't really image better than a pair of headphones that images well to begin with.

@ John,

I would like to *stress* that my improvement results have been achieved exclusivley through the use of the LAME mp3 encoder, NOT the more typical Fraunhofer or Xing encoders. The Lame encoder is simply superior in every way to the Fraunhofer codec and by about a galaxy over the Xing.
I can only report on what I hear and my experience with the Lame encoder which is again as stated an overal improvement to the original sound on the disc, the FLAC file or any other such lossless codec. Sometimes the improvements are quite dramatic, other times its about a draw. As to the "accuracy" compared to the source, well, in that regard and by definition, neither is 100% accurate to the source, ie. the mic feed into the mixer/recorder. I presume you mean more the "effect" that the conversion to mp3 generates - as stated, I am reporting an improvement in sound quality - take it for what it's worth.
Most likley the highest point and source of contention is the typical dismissal of any such claims by the audiophile community, which more often then not issues blanket statements about this and that. You won't be surprised to hear that a few "audiophiles" did the unthinkable and changed their opinion of the seemingly superior sound quality of said mp3 file *after* I told them that they were listening to a compressed mp3 file to begin with.

tsk... tsk... tsk...

;-)

If you are going to do MP3, Lame is the way to go. I thought that Lame was ubiquitous for encoding MP3.

I personally avoid any additional layer of processing I can in audio. Thats what many years in digital has lead me to. I also prefer lossless codecs because they are archival. I know I can convert them to future formats perfectly and never have to re-rip my CDs.

@ John,

understandable - computer processing and encoding algorithms will no doubt only improve further, it's deffinitley smart to have a "master" copy somewhere on your HD!

@Danny

What do you use as software to encode your MP3s into LAME?

Did some reading and I am interested.

@ Fred,

my preferred program of late is www.poikosoft.com CD-DA Extractor... it is the most comprehensive converter I am aware of - some of the features like being able to download specific cover art, etc, is outstanding - it also features extensive error correction and data retrieval ala' EAC, but is much more user friendly and comprehensive.

Most importantly, it includes LAME 3.97 (as well as a host of other encoders, FLAC, etc.) - you have to make sure to select the LAME encoder NOT some of the other mp3 encoders included... you then have the option of tuning LAME settings, etc.

Best,

Danny

Sometimes the quality of streaming mp3 is quite good. Ryan Adams has many of his songs up on his site and the quality is very good.

http://www.ryan-adams.com/cardinalRadio.html

@ John,

there are some really really good stations I access through my SB3 at 128 and 192k... even at 128 some of them sound really good!

Hi,

I just have to say that it is theoretically possible to restore what has been lost (not exactly, but a very very close match). It takes a tremendous amount of processing power; much more than we've got as consumers--but it does exist. It might take days and teraflops of crunching time, but it *can* be done. Example.
Suppose you see a chess board, take a snapshot-then take another snapshot later on, and another. If you are a skilled enough player, you can recreate the steps inbetween. Computers can do it too, but only with brute force.

Music MP3's are the same in that respect. What I assume this device is doing is filling in the likely sound with predicted values that are very close or the same to the original, and then it keeps going, eliminating the jagged waveforms of cds. This is also done to create super audio cd's from old digital recordings where the master analog tapes were lost. I've heard them, and they sound roughly 90% as good as records.
I also assume that this device could identify artefacts from poor mp3's and eliminate them-warbling, pre-echo, harshness. A 64kbs might take petaflops of time to reprocess it back to it's original state. (There will be statistical differences, mind you, but inaudible). If one really wanted to get fancy, a device could (theoretically) contain a list of songs, and the sound module samples used, and halls (for live) to contain the acoustic properties and calculations to recreate the proper sound presence and use cleaner sound samples. This is also done in remastering (talking about it anyway). It'll be a while before they squeeze it into a cell phone size device.

Everyone knows that matter cannot be created or destroyed only change form, but there's a little known corollary that goes along with it-information can neither be created nor destroyed, only exist in a more or less accessible state.

@ Joe,

excellent post! Thank you!

Another good example is turning a totally out of focus picture into a sharp, clear and pretty one. As you say, it's all mathematically possible and has been done before. Like you said however, the problem is that the processing power required to do this is still several orders of magnitude higher then what even the latest quad core cpu's are capable of.

I remember Steve Jobs saying that one day you'd be able to go into iPhoto and bring back all those auto of focuse pictures with a mouse click... no doubt, he was being his usual self, however, it's all moving closer to reality.

Alas, I somehow doubt this little $79 Creative device is capable of restoring content back to mp3 files...

Great post!