Digital audio quality

Is there any big difference between burned audio CDs compared to audio CDs produced by record makers.

Would it sounds nice to burn 128kbps MP3s to Audio CDs. Would I enjoy quality audio ?

Some people say you can hear a difference between the original and a copy, but I have rarely heard anyone say this.

You can hear a difference between the original and a CD burnt from MP3, though this difference will be less when the MP3s are encoded using a good encoder and at a high bitrate (192 or up).

For discussions on quality you can find a lot by using the search...

I think there is no sound difference between an original CD and a copy, if the reading is error free.
The reading have been checked to be error free (SPDIF output data identical to ripped data in wav files) on standalone hifi players (Sony, Yamaha CDX 860). People pretending a difference is audible says so even on error free playback (Bob Katz, http://www.digido.com , jitter article).

My arguments :

-The digital audio data is strictly identical to the original when a copy is played
-As far as I know, there is no measurable difference in the analog output : no harmonic distortion added, no frequency lost, no wow and flutter... at least no burner manufacturer claiming to improve the sound of copies (Teac, Yamaha Audio Master...) ever spoke about it. Their only argument is jitter reduction.
-Pit/land jitter improved by Audio Master etc can't affect PCM jitter in the DAC output in theory. Bob Katz says that it may load the servo engine of the player that would in turn load the power supply, that would badly react and become unstable, that would pollute the crystal oscillator, and that would affect the DAC work, thus affecting the sound.
-I can't hear any difference on the old Yamaha CDX860, Arcam Diva A85 ampli, Senheiser HD600 headphones. Although I often though I could, blind tests convinced me that I couldn't hear any real difference.
-There is no positive ABX blind test showing that the difference is audible. A positive ABX blind test would be an objective result.

Anyway, even if such a difference exists, it is well inferior to the difference caused by encoding in 128 kbps MP3, so your burned CD will be faithful to the sound of your MP3 files, the MP3 encoding being the main, if not the only, source of distortion.

128 kbps?
That sounds quite awful. Maybe not at the beginning, but after some time you can hear a lot of artifacts that you can't take out of your mind once you start hearing them.
I think I can't hear the difference between original and 192 kbps.

you wont be able to on normal PCs. I extract to wav when i am making a compliation cd sometimes, and i keep all my cd on my hdd (images) - but i got a big HDD

What for ? Does it feel safer to keep your precious data this way ?

no, if i want to listen to a CD i cant be botherd looking for it - i just mount it instead.

If / when i run out of disk space i will just start deleting them

...or burning discs, maybe ?

yeah, but i wouldnt bother 'cos i dont bother to backup audio cds

OK, that sounds reasonable. Here's my next question - why do you keep their images on HDD then ?

i have all image in an image partition which is about 50GB in size ...why?

because hard drive space costs much more than blank cdrs

well the majority of my 80 gig would be sitting idle anyway

So, what's your point in this thread then ?

Whats urs? u started it - i'm only answering questions, but enough (from now)

I thought you were able to read between the lines.
I mean if the HDD's space's running short - either burn CDs out of those images or just delete the idle stuff.
Big deal !

Ok here we go Wave is superior than any compresion MP3 as DVD audio is Superior to CD audio thats how you get the compression Ratio elimination the part of the audio stream that the human ear cant process (hear) it dosent mater if you use CBR or VBR compression for Mp3 it will never be as good as CDA here is something to that fact it is missing the comparison with the MP3 but it gives you an Idea of what I am talking about


Here is something that might help you:

http://www.howstuffworks.com/mp31.htm

and the comparison:




Any way to finish my long boring speach you can convert Mp3 to wav but it will at best be Near CD quallity not actual CD quallity and NOTE THAT if you dload the Mp3 the wav will be only as good as the MP3 you start with that said im finished



hope that is what you was asking for hehe .....my english isnt that good

That's why I deal with WAVs only.

Btw, the given by you link is for some reason non-working
But here it is again -> How MP3 Files Work

Copying a CD is lossless, it's one thing.
Making MP3 files is lossy, it's another thing.

By the way, the picture posted is completely misleading, and plays with word in order to make people believe that DVDA sound superior to CD.
The graphs drawed are NOT the CD and DVD output, but a picture of the samples values.
The CD and DVD output are both EXACTLY identical to the black line, in the analog CINCH plugs, because we have here a 10 kHz sine.
Any signal with no frequency component above 22kHz will, on a perfect CD player, be reconstructed exactly as the original smooth analog wave. Only accidents with frequency components above 22 kHz will be lost. If they fit between 22 kHz and 96 kHz, DVD-A will be able to record and play them, because it can use sample rates up to 192 kHz (twice the highest reproductible frequency).

Cool Edit Pro gives a good idea of how a digital wave is smoothed by the analog output of a CD player.
Here is a wav file, whose samples values are showed as rectangles :



here is a part of the same wav file, but the wave is recostructed between the samples, that are showed as green little squares instead of rectangles :



It's just a sine, like the 10 kHz one posted above. From those steppy rectangle, any CD Player is capable, as well as Cool Edit, to reconstruct the original sinewave. A capacitor ensures that the output is completely continuous, with no steps.

Mr. Pio, what is your opinion of the new digital audio formats SACD & dvd-a? -- I try to read everything I can on webpages about this subject. -- I've read that: although CD's have a greater dynamic range than vinyl LP's, they still sound "flat & digital" and can't reproduce "warmth" and "ambience" well like vinyl LP's can.

Websites I've seen about SACD & dvd-a claim that the new digital formats produce sound thats more true to the original analog recording. -- Like producing more "timbre" and "depth".

Is this stuff true, or just hype?

I Agree With Dillweed LP to me represents the Analog Signal you see and the closser the Digital format be it CDA or DVDA can imitate the audio stream the better the Quallity of the sound cause in the end digital signal is converted to Analog signal via de DAC (Digital to Analog Converter) and as you can see if you beleave the image DVDA has less of a gap between the 1's and trhe 0's that form the digital signal at least that is what I know to be a fact but I am not bull headed if you have something to back up what you are posting PIO2001 I will be more than happpy read it

one always endevers to learn more and you learn more when you make mistakes

This is a common misconception propagated by people who love the hollow, hissy sound of their LP's scratching along.
The fact is, CD audio records more information than the human ear can process. The sample rate and frequency range is beyond the capacity of the human ear's range. That means that any sound, noise, or music can be perfectly duplicated (as we hear it) by CDA.

Vinyl recordings are analog, and introduce artifacts and other noise. If you don't like how flat your CD sounds, it's because a) your speakers aren't reproducing the sound like the ones in the mixing studio or b) the record was mixed poorly or c) your turntable is hooked up to good speakers and the ones on your CD player blow. CDs play back exactly what's recorded onto them, so faults in the audio are usually from the studio, the listener, or the crappy musician.

The fact that we can compress out 80% of the signal into a 256kbps mp3 and not notice any decline in quality is a statement of how much overkill CD audio is. For stereo recordings, it has no flaws.

DVD and SACD audio can have more than two channels, which gives them an advantage over CD. The higher sampling rates, etc. are useless, since humans can't even process 44kHz.

If anyone would like to disagree that's great, I still have a bit to learn about this.

basically, you can perfectly reconstruct any signal, so long as the sampling rate is at least twice the period. CD's use sampling rates that are about 10% greater than they absolutely need to be, in order to simplify reconstruction.

i think thats nyquist's theorum

About higher audio definition, in theory, they only add inaudible frequencies (above 22 kHz). In practice, the direct comparison is difficult because the music released on SACD and DVDA benefits from a high quality mastering, that can be of higher qualty that the one used for CD. Meaning that if the same care would have been provided for the CD, it might have sounded as good as the DVDA, or the SACD.
Hybrid SACD should allow to compare the same recording on CD and SACD format. I've read such comparisons... some claim that the SACD is superior, some claims that it is inferior (link at the bottom).
96 kHz soundcards, like mine allow to compare 96 kHz 24 bits recordings with 44.1 kHz 16 bits. I recorded a vinyl in both modes.
I chose an old vinyl with a heavily analog sound : Sandra - 10/10, Maria Magdalena.
Technics SL-3100 turntable with Stanton Trackmaster EL, Cyrus One integrated pre amp. Marian Marc 2 soundcard recording, in CoolEdit pro. Listening : AKG K-400 headphones in the soundcard.

The 96 kHz recording sounded immediately much closer to the vinyl than the 44.1 kHz one. Then I tried to ABX them in a blind test.
Then I couln't make anymore difference between them. I was hitting A and B in the ABX program, and couldn't hear the difference. The worst is that I couldn't even decide myself about if they both sounded analog, or both digital ! Returning to CoolEdit, I had to admit that the analog sound of the 96 kHz file was just a false feeling : it actally sounds exactly the same as the 44.1 kHz one to my ears. Placebo is a thing you can only believe in after having tried blind ABX testing

The only other similar test I found was this one : http://jn.physiology.org/cgi/content/full/83/6/3548
It shows that unaudible frequencies added to music can affect the brain activity, though the listeners say that they don't hear a difference.

But without some positive ABX blind tests between 44.1 and 96 or 192 kHz sample rates, there is no reason to think that a difference is audible. A possible effect is added distortion, because amplis and speakers can't handle properly such high frequencies, they intermodulate and may produce added distortion under 20 kHz.

The image is not wrong, it is misleading. This one is easy to dismiss because it takes a 10 kHz sine as an example.
But a cd player output is lowpassed at 22 kHz in order to avoid aliases. Thus the steps visible on the picture, that represents the values stored into the CD are smoothed. In reality, the DAC first performs an oversampling, i.e. a digital smoothing, before a capacitor finishes the work in the analog side.

This picture is more difficult to comment, because it may represent the real analog output of digital devices, if they are not lowpassed :


48 kHz is near CD, 192 kHz is DVDA, DSD is SACD.

Two important remarks to begin with : "analog" stands for the pulse generator directly coupled to the oscilloscope. No analog tape recorder can capture such a short pulse (3 µs) , and vinyl isn't even worth mentioning.
Now, if the pulse is 3 µs wide, it means that there should be one sample every 1.5 µs in order to sample it. Thus, its frequency limit being half the sample rate, it is
1/(1.5*10^-6)/2=333 kHz
that's why no analog recorder could record it.

Now, look at what a pulse becomes when lowpassed. Let's draw a perfect pulse in Soundforge :



Then apply a lowpass



Here's the result (same scaling as above) :



It looks exactly like the lower quality versions of the picture above.
The maths behind this are that the frequency spectrum of an instant pulse is a flat line : all frequencies are represented at an equal level. Lowpassing a pulse is nullifying the frequency content above a given limit. Thus drawing a rectangle as spectrum. Then, applying an inverse Fourier Transform to the rectangle, the result is a
f(t)=Sin(t)/t
function, it's what we get on the pictures.

Conclusion : all three copies of the pulses are perfectly recorded without any distortion, exept that each of them is affected by a different lowpass : 24 kHz for the 48 kHz one, 48 kHz for the 96 kHz one, 96 kHz for the 192 kHz one, and something I don't exactly know for the DSD one.
Since human hearing is limited to 20 kHz at most, all copies, whatever they look like, must sound perfect. The ones looking closer to the original on the picture have just inaudible frequencies in addition, that sharpens the graph.

By the way, the first picture also shows that the steps are smoothed out in the analog output by the lowpass filter.

Forum discussions about digital audio at higher definiton than CD :

Help! Sacd Good Or Bad?, Does SACD ought to sound like crud? @ HydrogenAudio

DSD vs. PCM: Progress, Or Just More Junk Science? @ George Massenbourg recording forum (DSD is SACD and PCM is CD or DVDA)

George, Watch this!!!....(96k) @George Massenburg, abstract in page 33