John Curl's Blowtorch preamplifier part III

Status
Not open for further replies.
Jump to Latest
#33,102
Thx Jam... a bit wordy but makes several points not often considered... .

"There is a huge caveat to the foregoing approach, however: this approach only works when the signal is steady, with a known and steady frequency.
.... This approach would not work well at all on any unstable power source, ....

.... this approach has the drawback that it simply will not accurately catch transient events, or rapidly changing line levels. Transient events ...... have their own spectra that adds to the series of spikes of the basic power; rapidly changing line levels act to spread out the spikes in the power spectra, as does changing line frequency. Any of these phenomena can completely destroy the accuracy of the measurement....

-----------------
rerun ---> To accurately measure the frequency of a signal, we need a sampling rate of at least twice the highest frequency in the signal. This concept is known as Nyquist's theorem. To get the shape of the signal, you will need a sampling rate of at least ten times higher than the highest frequency in the signal. Jan 8, 2019

----------------
It was Demian Martins DAC that let me down-load and hear higher sampling rate and more accurate sound which got me interested again. Invested in new gear and better speakers etc. There was hope for better and I lived long enough to see it happen. Everything is more promising again but not at 44KHz for audio. We can do a lot better now.



THx-RNMarsh
 
Last edited:
#33,104
RNMarsh said:
rerun ---> To accurately measure the frequency of a signal, we need a sampling rate of at least twice the highest frequency in the signal. This concept is known as Nyquist's theorem. To get the shape of the signal, you will need a sampling rate of at least ten times higher than the highest frequency in the signal. Jan 8, 2019
Click to expand...


This is not true, but if we say it often enough, maybe it will become true.


Good night to all,
Chris
 
#33,107
Not mentioning the CD BW design limit of 20KHz. ... but, ..

Anyway, a minimum sampling rate of 192KHz or higher would be fine. 384K would be perfect IMO.

Many sites, like HDTracks, offer music at 192/24. And DSD also. These are more accurate. CD's 44.1/16 is indefensible against such improved sound and measured performance.



THx-RNMarsh

[PS -- IMO]
 
Last edited:
#33,110
All these things you describe below is not a "fault" of anything - it's the result of the limited sampling frequency. We all know that 44,1 cant do 25,6 Khz, you dont have to rub that in. 44,1 is fine if you are contempt with music signals reproduced up till 20kHz. Are you OK with that? I am.

Justs to be sure that you understand this because out of what you write one can get the impression that you just dont really get it - Any transient or fine detail - the ones you talk about that are amiss in 44,1 are.... wait for it....

...details in music which waveform parts are composed by flanks that represent higher frequencies than 21,05 kHz.

So below, you lecture us on that you to have discovered that a 23kHz detail is lost by a 44,1 ksps PCM system - wow! This makes sense if one undertands the relation between frequency and "slew rate".

Your technical analysis of a PCM system ability to recreate an analog signal is not correct. 2 samples a 44,1 will recreate a perfect sine. If there was more than a sine there at AD conversion time, it was lost due to the Fs limitation and not to some FFT-denial evil ghost. The PCM system will extract anything "sine" out of any signal waveform and reproduce that sine. It's a filter :wave: you know.

Whats a repetition? One sample will reproduce a sinc. Thats half a sine - OK? Its that non-repetitive enough for you?

Your last remarks indicate that you haven't had a decent DAC at hand. Or the rest of your system is tuned to compensate for all the flaws in record playing so that when a high quality source is introduced, these compensations deny you from hearing the source. Do you visit acoustical events regularly? Maybe you don't recognise reality? Can you handle the truth? Or are you anchored to the old warping rumble and H2?

//

RNMarsh said:
How many times have we heard this as Thee reason 44.1Khz sampling is fine:

"Just like Nyquist tells that the sample frequency Fs should be at least two times the highest signal frequency for correct reconstruction of the signal,"

Dont you know that only applies to a repetitive waveform? You cant construct a waveform from just 2 data points in one waveform cycle.

Anyone ever use a "digital" scope? Try to capture a single transient event with 2 samples. You need at least 5-10 data samples to even get a clue what the waveform shape and timing is. More data points and more detail is shown.

Nyquist works in both directions... more sampling needed for single cycle and fewer sampling points for many repeating cycles (2 being the minimum) With music it is of transient and never repeating waveform SO .... This is why (IMO) it sounds worse as the music content has higher frequencies... like cymbals. Not enough data points or too low sampling rate.

So, I answered the question of which is responsible for the improved sound... bits or rate. its mostly sampling rate.

Ok. sharpen your pencils... I am bracing myself for another wiz-bang EE101 thing. Remember though... keep in mind the central point -- Music is not repetitive waveforms ... and the phrase AT LEAST 2 points are needed really means with music waveforms you will need more than 2. More than 44.1Khz.

Someone told me at LLNL that if you run the sampling rate up high enough you will have the original analog signal. Seems he was right. But his problem with digitizing ultra fast transient signals was the sampling rate always had to be much higher than the signal being measured and they could barely keep up using wideband fast rise time analog techniques.

TEKtronix, when they first came out with digital scope techniques tried to use 2 times to make thier BW spec look better. They were called on it by people who really needed accurate waveform capture. Now they use 5 times. My TEK scope claims 300Mhz BW using 2 GHz sampling rate.

If you use the 5 times number, 44Khz barely gets you out of the midrange before falling apart. And that is exactly how it sounds, too.




THx-RNMarsh
Click to expand...
 
#33,112
I think what Richard might overlook is that a signal with a fundamental of let's say 1kHz can contain many higher frequencies that determine the specific wave form. In order to faitfully reproduce that signal, obviously, you need to sample at twice the frequency of the highest frequency you want to discriminate within that signal.

The simplest test is: can you hear the difference between a 10kHz sine and a 10 kHz square wave. If not, Redbook is good enough for you.
 
Last edited:
#33,118
scottjoplin said:
Fourier denial is still a good marketing tool to confuse the gullible audiophile, I don't see that ever changing.
Click to expand...

You do not understand Fourier, I think. Or maybe you think music only have 20kHz sine wave, or 10kHz signal only have H2 distortion, etc.

If you are poor just do not buy the expensive audio gear. Or if you can not tell the different about the quality of cheap gear and expensive gear, just buy cheap one. Why do you blame the salesman? 😀 😀 😀
 
Status
Not open for further replies.
Top Bottom