The importance of Components?

[olblueyez]

Quote:

Originally Posted by Real Man of Genius /img/forum/go_quote.gif Being a serious oenophile myself I can tell you with absolute certainty that the quality of gasoline used in the vehicle to deliver the wine has a significant impact on the taste of the vintage. We do not need to prove this. We can taste the difference and that's all that matters. Paying exorbitant amounts of money for wine that was transported by vehicles using this specially made petrol is therefore completely reasonable. Any fool can see the logic here (and don't even get me started on the importance of the tires!) The fact that no one has ever once been able to remotely PROVE any difference is meaningless. I trust my taste buds.

Another tool who knows everything about nothing.

 

[Real Man of Genius]

Quote:

Originally Posted by olblueyez /img/forum/go_quote.gif Another tool who knows everything about nothing.

Your inane response, along with the quasi-arrogant quote in your signature is priceless irony.
Thank you.
You actually have a link to the Wiki on sarcasm in your sig. Wow.
Settle down Beavis, it was meant for amusement.

Amusement - Wikipedia, the free encyclopedia

 

[PhilS]

Quote:

Originally Posted by Real Man of Genius /img/forum/go_quote.gif Being a serious oenophile myself I can tell you with absolute certainty that the quality of gasoline used in the vehicle to deliver the wine has a significant impact on the taste of the vintage. We do not need to prove this. We can taste the difference and that's all that matters. Paying exorbitant amounts of money for wine that was transported by vehicles using this specially made petrol is therefore completely reasonable. Any fool can see the logic here (and don't even get me started on the importance of the tires!) The fact that no one has ever once been able to remotely PROVE any difference is meaningless. I trust my taste buds.

I can't tell which viewpoint you're mocking. On the one hand, you could be suggesting that subjectivists are so foolish that they think it matters what type of gasoline is used to deliver the wine. On the other hand, you could be suggesting that objectivists are so arrogant and condescending that they would argue that subjectivists think the quality of gasoline used in the truck to deliver the wine is important.

Maybe it's just a joke, though, and you have to have a particular type of sense of humor to get it.

 

[Real Man of Genius]

Quote:

Originally Posted by PhilS /img/forum/go_quote.gif I can't tell which viewpoint you're mocking. On the one hand, you could be suggesting that subjectivists are so foolish that they think it matters what type of gasoline is used to deliver the wine. On the other hand, you could be suggesting that objectivists are so arrogant and condescending that they would argue that subjectivists think the quality of gasoline used in the truck to deliver the wine is important. Maybe it's just a joke, though, and you have to have a particular type of sense of humor to get it.

Well done sir. You are pretty bright for someone who believes in cables

 

[olblueyez]

Quote:

Originally Posted by Real Man of Genius /img/forum/go_quote.gif Well done sir. You are pretty bright for someone who believes in cables

I have never seen a more appropriate avatar in my entire life. Does your mother lay out your cloths in the morning and help you get dressed? Not exactly the brightest bulb on the tree are you? Another self proclaimed Thomas Edison. On to the ignore list for you as you are truly useless.

 

[PhilS]

Quote:

Originally Posted by olblueyez /img/forum/go_quote.gif I have never seen a more appropriate avatar in my entire life. Does your mother lay out your cloths in the morning and help you get dressed? Not exactly the brightest bulb on the tree are you? Another self proclaimed Thomas Edison. On to the ignore list for you as you are truly useless.

Wait a second. He said I was "pretty bright." Whose side are you on?

Real Man of Genius, I also don't believe in aliens or that the earth is flat.

 

[olblueyez]

Quote:

Originally Posted by PhilS /img/forum/go_quote.gif Wait a second. He said I was "pretty bright." Whose side are you on? Real Man of Genius, I also don't believe in aliens or that the earth is flat.

Compared to some of these characters you are the most intelligent man on Earth. I don't think I have ever seen a more rag-tag group of trollers in my life. Pathetic, truly Pathetic.

 

[JaZZ]

I truly hope this thread doesn't get closed.


Quote:

Originally Posted by b0dhi /img/forum/go_quote.gif I'm really curious to hear the rest of the filter conversation.

As to the amplitude modulation and its interpretation as aliasing artifact:

14 kHz

18 kHz

20 kHz

21 kHz

22 kHz

.

From Mother of Tone - The CD Format:

«In fact, we always get that beat frequency with any sampled wave that is not exactly a whole division of the sample rate. So we expect no beats with 22.05 kHz, 14.7 kHz, 11.025 kHz, 8820 Hz, 7350 Hz, 6300 Hz, and so on.

Let's give that a try and take a look at 14.7 kHz:

Yes. There's no beat to see. All exactly the same amplitude.»

Another excerpt:
«Funny enough, it is said to be a consequence of the sampling theorem that a signal cannot be band-limited and time-limited. If a signal is 100% band-limited, it cannot be time-limited anymore. If it is time-limited, it cannot be band-limited.»

This scenario is very descriptive. A 100% band-limited signal will ring infinitely. Thus it can't carry any time information (hence transients) anymore.

Wikipedia:
«Low-pass filter for prevention of signal interferences.
Frequencies higher than half the sampling frequency have to be removed from the signal, otherwise artifacts occur. The removing of these components leads to an alteration of the signal and should only be applied if this alteration is insignificant or an increase of the sampling frequency isn't an option.»

Since band-limitation of the signal in question is a primary precondition for digital sampling, the sampling format dictates shape and degree of the signal corruption beforehand. Now this pre-corrupted signal may indeed be able to be passably or even perfectly «reconstructed» by a digitization following the Nyquist-Shannon premises.


What I also find fascinating with this topic is the dualistic aspect of the amplitude modulation. Primarily it's a mere amplitude modulation caused by interference. Secondarily it's also a foreign frequency (~24.2 kHz) seemingly occurring out of nowhere. So aliasing works in both directions. In our case we have mirror frequencies projected into the ultrasonic range.

When it comes to filtering the «two» unwanted aliasing products out, again one and the same measure does the trick: a sharp brick-wall filter at ~21 kHz. It suppresses the 24.1 kHz frequency, and at the same time it makes for the time-smearing eliminating the amplitude modulation... two different aspects of the same thing. Additionally the acoustic impression of such a beat may vary -- depending on frequencies, intensity relations and context -- between amplitude modulation (or roughness, resp.), two tones and both effects together.

Mother of Tone:
«The beats are reduced the more, the sharper the filter works and the longer it rings... It is actually the ringing that bridges the beats.»
.

 

[nick_charles]

Quote:

Originally Posted by JaZZ /img/forum/go_quote.gif Interesting technical stuff snipped to save bandwidth

You forgot to quote this bit

Quote:

What about Oversampling ? We can have higher fidelity with 16 bits amplitude resolution and 10 or more sample-points per wave-period, as we can recover the original signal with a very small amount of beat products (less than1%). And then we can achieve better sound quality in reality, not just on the paper

But more to the point.

The Beat phenomenon is most marked on simple waves Sine and more so with square waves, on my recordings of square waves I found a lot of this effect and when you compare the generated pure square wave with the recorded 44.1 khz samples of the same there is a marked effect.

However this does not tell the whole story. When you sample a complex waveform such as a cymbal crash or a segment of music this effect is pretty much invisible. The original wave and the played back on CD and recorded at 44.1 khz waves are near identical with no marked beat artifact.

I have done this a lot, try it yourself or look at these , the 2nd is phase inverted. NB this was done from an iPod to my desktop crappy soundchip so it is not high quality and there is a slight amount of temporal slippage but the pattern is still easy to see

Original

Sampled at 44.1

 

[JaZZ]

Quote:

Originally Posted by nick_charles /img/forum/go_quote.gif You forgot to quote this bit

Actually I didn't forget it -- I left it out intentionally. Oversampling doesn't change a bit on the issue (as demontrated on several occasions; see e.g. Symphony graphs). It just helps simplifying the analog filtering, i.e. signal path and material effort.


Quote:

The Beat phenomenon is most marked on simple waves Sine and more so with square waves, on my recordings of square waves I found a lot of this effect and when you compare the generated pure square wave with the recorded 44.1 khz samples of the same there is a marked effect. However this does not tell the whole story. When you sample a complex waveform such as a cymbal crash or a segment of music this effect is pretty much invisible. The original wave and the played back on CD and recorded at 44.1 khz waves are near identical with no marked beat artifact. I have done this a lot, try it yourself or look at these, the 2nd is phase inverted. NB this was done from an iPod to my desktop crappy soundchip so it is not high quality and there is a slight amount of temporal slippage but the pattern is still easy to see

Yes -- I do a lot of sound editing and am aware that there's no clear indication of ringing/time-smearing in the complex signals ripped from music CDs (hence recordings meant to have passed an anti-aliasing filter before sampling). I'll do my own test with recording from a CD using D/A and A/D conversion, though.

Your two test samples differ quite a bit (although I couldn't hear a clear difference). Also, the original suffers from temporary clipping, which the recorded copy doesn't show (as digital zero, at least). Moreover some conspicuous 1-sample peaks could be a hint for resonance effects caused by the additional conversion and filtering, but that's just an assumption. All in all the difference is quite obvious -- either it's the quality of the D/A-A/D convertor(s) or the additional (!) ringing or both, hard to tell. In any event, complex signals like these make it hard to detect ringing, because it might be masked by the following signal samples.

Come to think of it: I could handcraft my own samples -- making tone burst out of hearing-test signals -- and record them.
.

 

[nick_charles]

Quote:

Originally Posted by JaZZ /img/forum/go_quote.gif Your two samples differ quite a bit (although I couldn't hear a clear difference). Also, the original suffers from temporary clipping, which the recorded copy doesn't show (as digital zero, at least). Moreover some conspicuous 1-sample peaks could be a hint for resonance effects caused by the additional conversion and filtering, but that's just an assumption. All in all the difference is quite obvious -- either it's the quality of the A/D-D/A convertor(s) or the additional ringing or both, hard to tell. In any event, complex signals like these make it hard to detect ringing, because it might be masked by the following signals.

It is a very old recording of a segment of the UK "Danger Man" theme from the 1960s and not terribly high quality to start with and it was played back on a 5G iPod (in WAV format) and recorded 16/44.1 on a SoundMax chip via Audacity, I did not match the levels so it has everything against it. You can see that the timing is not perfect

In your screenshots each major division is 0.5ms and the minor divisions are 0.05ms - do any two recordings of the same track and you will get random variations of the same magnitude. I have been recording "identical" musical samples repeatedly for the last month and a half (on a slightly better ADC) and I see that level of variation all the time. For my purposes I trim down to
0.01ms levels

That they were not easily audibly different is interesting ( I ABX'ed them successfully 10/10 with AKGK501 , TBAAM and a Nikko TRM-210, though it did require rapt concentration and multiple repeats on my part), the sample is 1.2db quieter overall (0.8db at 17K and averages 1.87db down above 19K (not that I can hear 19K) , 1.17db down from 13K to 15K which is perhaps more relevant) and is phase inverted, do not ask me how that happened , could be the iPod, Itunes (used for importing it) or the ADC ? - that is bizarre, any way it does indicate how relatively insensitive human hearing is.

The point was that the beat effect really does not show up. I could post some square wave samples I have recorded and the picture is pretty bad even as low as 500hz. It would be interesting to do actual listening tests on generated square waves and recorded ones, maybe when I get some spare time..

 

[JaZZ]

Quote:

Originally Posted by nick_charles /img/forum/go_quote.gif The point was that the beat effect really does not show up.

The beat effect can't show up because there's no sine wave with which the sampling frequency could interfere.

Moreover all «decently designed» DACs have a standard brick-wall filter to prevent it (by means of time smearing)

.
.

 

[nick_charles]

Quote:

Originally Posted by JaZZ /img/forum/go_quote.gif The beat effect can't show up because there's no sine wave with which the sampling frequency could interfere.

So where is the problem

 

[milkweg]

Quote:

Originally Posted by PhilS /img/forum/go_quote.gif "The general BS spread around here." LOL. You're so blinded by your neuroses that you just can't help but prove the very point I'm trying to make. Hilarious.

You have no point.

You may think you are doing a service by making wild unfounded claims about audio gear but I see you as doing a disservice because new members give you more credibility than you deserve.

 

[PhilS]

Quote:

Originally Posted by milkweg /img/forum/go_quote.gif You have no point.

Good one. I don't think I've ever been pwned so badly before.

Quote:

Originally Posted by milkweg /img/forum/go_quote.gif You may think you are doing a service by making wild unfounded claims about audio gear . . . .

What claims are you talking about? Most of my claims on this thread are not about audio gear, but about what a mean-spirited fellow you seem to be on this forum. I suspect, or at least hope, that this is not the case in "real life," but it's the "Head-Fi persona" you seem to want to cultivate.

Quote:

Originally Posted by milkweg /img/forum/go_quote.gif . . . . but I see you as doing a disservice because new members give you more credibility than you deserve.

I don't know how you know what new members think about my credibility. I'm sure the vast majority are pretty intelligent and not easily duped by anything foolish that I say or anyone else says. Then again, since you think almost everyone on Head-Fi is stupid, you presumably think everyone here gives too much credibility to everyone else except you.