Downloadable Mp3s Are They Killing Quality?


Status
Not open for further replies.
sam fisher said:
I'm bored of this now. I kindly request you just give up. Your failure to induce any form of negative reaction in me is pitiful. It provided me great humour but now just boredom remains.
Sam I suggest you go off and "emulate" to relieve some of those teenage hormones.
 
Last edited by a moderator:
Sam seems to have missed the point and WarmFluffy has missed where Sam missed the point. He seems to be talking about the losses you get when you convert an analog waveform into digital, and also the losses you get when you rip directly from a CD, do a lossy encode and then burn the encoded file, decoded. WarmFluffy is talking about ripping data directly and exactly from a CD, then putting the exact same data on another CD for a theoretical perfect copy.

Can we all be friends now?

(Also, don't give me any bullshit about planting the hypocrisy there for people to 'find'. Admit that you were wrong and that I criticised your argument in a way that I deem very clever indeed and am actually quite proud of.)

Also, to get back on topic, I can tell the difference between 128 and 192 kbps. You miss a lot, especially from the background, and especially stuff like triangles, cymbals and whatnot. It sounds more flat. Above a well-encoded 192 kbps CBR MP3, I can't tell much with most stuff. I rip to OGG Q5.
 
sam fisher apparently never heard of electronic music.

Also, yes MP3s are ruining quality. I download anything less than 192kbps very reluctantly. iTunes is probably the worst example. They are charging for use of these 128kbps files. I can begrudgingly accept 128k when I get it for free, but there is no excuse in the broadband era to not offer higher quality audio.

sam fisher said:
See you are wrong there, Vinyl records are good for one reason, they store an analogue representation of the original sound wave. Thus an identical copy, not a digital approximation, thus potentially they are of higher quality than any possible digital recording (unless you sample at an infinite sample rate, with an infinite number of quantum levels with voltage differences that were tending to zero, or infinitely small so to speak). Of course they have other drawbacks as well such as being easily damaged and degrading with use etc.
False for most modern music. Nearly all music is digitally processed and hence not 'the original sound wave', even when put onto vinyl. Analogue -> digital -> analogue has the same drawbacks as analogue -> digital -> digital.
 
Last edited by a moderator:
NIck, you are correct. However, we both posted that ages ago. The argument stems from WF not admitting she was wrong to call me wrong as she misinterpreted what magus meant.

On topic: I was always told OGG was better quality per bit rate compared to MP3 however I now find OGG's just don't sound "right", any idea why? They just don't sound real.
 
sam fisher said:
On topic: I was always told OGG was better quality per bit rate compared to MP3 however I now find OGG's just don't sound "right", any idea why? They just don't sound real.
Depends what quality you encode at. Q5 to me sounds indistinguishable from a CD.
 
Last edited by a moderator:
I used to use Q2 (eek). However the other day I ripped a CD song to multiple different OGG and MP3 bitrates. The MP3 192k one sound more "real" as if you were there in real life, than the OGG one at around the same bitrate. I also use WMP as my main player so I can't really use OGG's as it doesn't add them to its library etc.
 
Talking of players, I find that certain players do it better, here ae my fave:
deliplayer - Excellent playlist options
Foobar - Amazing quality, and better for the audio folk. It will also scan all your files and set an audio level so when you play them back, it will be all the same level!!!

My hated players are:
Winamp - It's too commercial these days
Media Player - Urgh, if you like adverts, this is for you
 
sam fisher said:
See you are wrong there, Vinyl records are good for one reason, they store an analogue representation of the original sound wave. Thus an identical copy, not a digital approximation, thus potentially they are of higher quality than any possible digital recording (unless you sample at an infinite sample rate, with an infinite number of quantum levels with voltage differences that were tending to zero, or infinitely small so to speak). Of course they have other drawbacks as well such as being easily damaged and degrading with use etc.
You're an idiot, and you talk out of your arse, because you obviously know nothing about signal reconstruction theory.

A digitised signal is not "only an approximation", if you knew about the Shannon-Nyquist theorem you would know that every frequency component under half the sampling rate of a signal is kept intact, in the case of CDs that's any sound from 0 Hz to 22 kHz (to put this into perspective, most people can't hear anything above 16 kHz, and few people can hear 22 kHz. Besides, while vinyls are analog, that doesn't mean they have an infinite bandwidth, if you knew anything about the frequency response of vinyls you'd know why an infinite sampling rate isn't needed.

As for voltage levels, the dynamic range of a signal sampled at 16-bits per sample is about -96 dB. That means that if you play your music as loud as 100 dB, which is pretty fucking loud (listening to anything at that level for a prolonged period of time leads to permanent hearing loss), the artifacts due to this "approximation" are heard in the form of a 4 dB white noise. 4 dB is about the dimmest sound you could hear in a perfectly silent room, so if you're listening to something so loud it might make you half deaf, you definitely won't hear that.

Besides, the iPod/stereo or whatever you're listening to adds a noise a few orders of magnitude louder than that, and a vinyl has at best a dynamic range around -60 dB, that means waaay more noise than a CD.
 
Last edited by a moderator:
A_SN said:
sam fisher said:
See you are wrong there, Vinyl records are good for one reason, they store an analogue representation of the original sound wave. Thus an identical copy, not a digital approximation, thus potentially they are of higher quality than any possible digital recording (unless you sample at an infinite sample rate, with an infinite number of quantum levels with voltage differences that were tending to zero, or infinitely small so to speak). Of course they have other drawbacks as well such as being easily damaged and degrading with use etc.
You're an idiot, and you talk out of your arse, because you obviously know nothing about signal reconstruction theory.

A digitised signal is not "only an approximation", if you knew about the Shannon-Nyquist theorem you would know that every frequency component under half the sampling rate of a signal is kept intact, in the case of CDs that's any sound from 0 Hz to 22 kHz (to put this into perspective, most people can't hear anything above 16 kHz, and nobody can hear 22 kHz, if they say they do they only heard non-linearities). Besides, while vinyls are analog, that doesn't mean they have an infinite bandwidth, if you knew anything about the frequency response of vinyls you'd know why an infinite sampling rate isn't needed.

As for voltage levels, the dynamic range of a signal sampled at 16-bits per sample is about -96 dB. That means that if you play your music as loud as 100 dB, which is pretty fucking loud (listening to anything at that level for a prolonged period of time leads to permanent hearing loss), the artifacts due to this "approximation" are heard in the form of a 4 dB white noise. 4 dB is about the dimmest sound you could hear in a perfectly silent room, so if you're listening to something so loud it might make you half deaf, you definitely won't hear that.

Besides, the iPod/stereo or whatever you're listening to adds a noise a few orders of magnitude louder than that, and a vinyl has at best a dynamic range around -60 dB, that means waaay more noise than a CD.

I agree, and as I mentioned earlier in this thread, Vinyl cannot be exact because of the bass effect on the grooves. If you are looking for 100% representation of an audio recording, the digital representation would actually be more accurate, as it's it impossible for vinyl to do that.
 
Last edited by a moderator:
A_SN said:
You're an idiot, and you talk out of your arse, because you obviously know nothing about signal reconstruction theory.

A digitised signal is not "only an approximation", if you knew about the Shannon-Nyquist theorem you would know that every frequency component under half the sampling rate of a signal is kept intact.
Yes, that is to prevent frequency aliasing. I am aware of this. HOWEVER, it is not kept intact, it is however an accurate enough copy to prevent aliasing and sound correct to the human ear. I was on about compression artifacts. Then recompressing a waveform already containing artifacts. This would create more artifacts.

Unlike most things I say, in this case I am absolutely certain of what I am saying and this whole argument is about failure to interpret what someone else is saying correctly.
 
Last edited by a moderator:
WarmFluffyUK said:
Talking of players, I find that certain players do it better, here ae my fave:
deliplayer - Excellent playlist options
Foobar - Amazing quality, and better for the audio folk. It will also scan all your files and set an audio level so when you play them back, it will be all the same level!!!

My hated players are:
Winamp - It's too commercial these days
Media Player - Urgh, if you like adverts, this is for you
I completely agree about Winamp... I stopped using it for that very reason!
Foobar2000 rocks! I love the how much you can customise the player itself. I set it up so that I have Translation tags for my non-english music so that when not playing it would display the native titles but when it was playing would display the translated tags. :)

Unfortunately Linux does not have foobar and the best player I can find is Amarok... certainly pretty and good for managing music, but doesn't have a nice plugin system as foobar, meaning I can't use Amarok to play all those funky formats that Oldplay supports!
 
Last edited by a moderator:
daclassicgamingmaster said:
Fluffy, you aren't really gonna give up your mod position, are you?
Yes I already requested it, And am amazed I am not a De-MODed already. Squidge would be better than me.
 
Last edited by a moderator:
sam fisher said:
HOWEVER, it is not kept intact, it is however an accurate enough copy to prevent aliasing and sound correct to the human ear.

But we don't give a shit about whatever's above 22 kHz. You talked about quality, the ultra-sounds you might find on a vinyl (heavily dipped in noise) have nothing to do with the quality of what we hear.

QUOTE
I was on about compression artifacts.
What I quoted you on had nothing to do with compression artifacts, it was all about analog vs. digital sound. You're just trying to make yourself sound less like an idiot anyways, and fortunately for you, it's working, I don't think you sound as ignorant as I first thought.
 
Last edited by a moderator:
I was on about analogue vs digital as I was going off on a tangent to try and explain it. Go back to my original post the argument stemmed from.

Edit: Furthermore, in an even further back post I mentioned that although we cannot hear waves above 20khz directly. They CAN superpose waves that are audible to us. This is why live music almost always sounds better.
 
sam fisher said:
They CAN superpose waves that are audible to us. This is why live music almost always sounds better.
It seems that you're erroneously trying to refer to non-linearities (the only way you could *detect* (and not directly hear as they are) ultra-sounds). How does it make live music sound better when the result of it hardly matches anything desirable? They don't, and you lose some more of my respect for daring thinking that a concert sounds better than music in your iPod because of non-linearities from ultra-sounds. If anything live music sounds better because it's louder, you have no additional linear systems in your way (microphones and loud-speakers or headphones have their own frequency response and they're rarely close to being flat, that makes them like an equaliser) and you have a better spatialisation.
 
Last edited by a moderator:
The waves > 20khz can superpose and change the < 20khz waves, thus if you don't bother recording them in your digital copy you are not getting the EXACT wave you have if you were actually at the concert. This is just something my physics teacher mentioned to us while covering that topic.

There are multiple meanings for superposition. I mean it in the terms of combining more than one sound wave into a resultant wave as shown simply in this diagram:

twowavescombine.jpg


Edit: But I do agree that non-audible wave superposition is not the ONLY reason that live music sounds better.
 
sam fisher said:
The waves > 20khz can superpose and change the < 20khz waves
No, utter bullshit, where'd you got that shit from? And you're a retard for even disagreeing with me considered the link in my signature. If you're talking about beat frequencies then you've got it wrong, if you've got say a 30 kHz wave + a 31 kHz wave, you'll effectively get a 1 kHz beat frequency out of it, but you won't hear it, because the beat frequency is the frequency of the envelope, not of the signal. Well like I said due to non-linearities (distortions) you could hear something but is it what you're talking about?

As for the rest of what you talked about, and as I struggled to explain to Orkie and vimacs, what you're talking about has strictly nothing to do with ultra-sounds. It's good if you remember what your physics teacher says but it's of little use if you use it on something unrelated.
 
Last edited by a moderator:
Could everyone please stop calling other people idiots? I was following this topic with high interest as I'm sure there's something new to learn for everybody, but personal attacks have made it somewhat unpleasant to read. C'mon guys, we're all adults after all! :)

Back on the subject, music encoded from audio CD's to mp3 files with bitrates lower than 192kb/s sounds worse than the original, period. Even 320kb/s mp3's suffer quality loss greatly at times, but are a lot more audible than the lower bitrates. Over the past years that I've been collecting music, I've come across some really awful sounding mp3's, both from private / public sources as well as commercial ones. From time to time I update my audio collection and get rid of very poorly encoded music.

About half a year ago, I bought myself a "Philips SBC HP830 closed shell headphone" after my uncle advised me to grab one while I can, as they were discontinued and replaced by a newer model. I got it NEW from Curry's for 39.95 UK pounds. And tell you what .. I was shocked by all the things I could hear that my old Sony headphone didn't reveal (and I thought that was a good headphone back then) In fact, I had to go through my mp3 collection AGAIN to filter out everything that didn't sound right, and there's not much left below 192kb/s now!

I'm not sure whether people can't or don't want to hear artifacts in low encoded audio files, but I believe that a lot of people who say they can't hear a difference simply use the wrong audio equipment. A good set of headphones does magic to your ears, even on some of the cheaper audio players available. Some of you may have a lot better ears than you think! :)

Being a music artist myself, I am planning to use .flac as the future format for my online downloads as .mp3 doesn't seem optimal for this task. I'm using 80kb/s for preview audiostreams, so they do sound a lot worse than the original media but that's not without a reason - it saves space on the webserver and motivates people to buy the high quality song ;)

"SunSpire"
 
SunSpire said:
Could everyone please stop calling other people idiots?
Not really, sorry. The only way I can stop calling people idiots is if the said people quit having the nerve to disagree with me, mainly on a subject I know about like 100 times better than them.
 
Last edited by a moderator:
Status
Not open for further replies.
Back
Top