Here is a hint to you. Read what I wrote. I didn't suggest a perfect square wave exists. I'm well aware of rise and fall time.
Pulses are not a sine wave, and even with rise and fall times, are not a sine wave.
You can make up a square wave with "numerous summed sine waves"
But that is not what "makes" them in a digital switch, and the implication is incorrect. So you clearly have no idea.
you are correct, I have no idea what you are on about
a digital switch and its electrical output, the actual real object, is subject to the very same rules. the digital interpretation ignores this. you do not need an oscillator to create a sinewave.
hey are you trying to destroy the only thread that draws any parallel to analogue?
anyway is this another attempt to shift a discussion of one thing, by taking a snippet out of context and attempting to discredit the premise? or is just a petty and unsuccessful attempt to score points?
a digital switch and its electrical output, the actual real object, is subject to the very same rules. the digital interpretation ignores this. you do not need an oscillator to create a sinewave.
hey are you trying to destroy the only thread that draws any parallel to analogue?
anyway is this another attempt to shift a discussion of one thing, by taking a snippet out of context and attempting to discredit the premise? or is just a petty and unsuccessful attempt to score points?
Last edited:
No, Fourier says that a square wave is a series of sines. How you made it is entirely irrelevant.erin said:
This reminds me of the great sideband debate in the early 20th century. Mathematics (actually, trigonometry) says that a modulated carrier wave consists of an unchanged carrier plus sidebands. Many engineers disputed this, claiming that sidebands were merely a matter of mathematics and not reality, and would not believe it until someone managed to actually detect and measure the sidebands. Now you can make an AM signal either by modulating a carrier or adding the sideband signals to a fixed carrier but the end result is the same. AM is carrier plus sidebands. In the same way a square wave is a sum of sines.
not because you have some knowledge I dont, but because you are not making any sense
I am glad the discussion has returned and not just devolved into snippy comments.
I say that the way you make it is very relevant.
So in this regard I believe that Fourier was incorrect, the way you make
the square wave is very relevant.
I agree that a genuine square wave is not possible due to rise and fall times.
But to say that all square waves are made from sine waves is also incorrect.
A circle viewed from a distance looks like a circle because we see it in its entirety, but when you look at a close up of just one edge it can look like a straight line.
Fourier was discussing square waves from a mathematical point of view in regard to ways of expressing them mathematically, but ignoring the simple reality of day to day operation of circuits.
I'm not disputing sidebands, and noise.
So I come back to the battery operated circuit controlled by a switch, or any TTL CMOS etc. chip.
There is no sine wave "making" the square wave. Its a DC (yes i know there is ripple) signal being turned on and off, with a rise and fall time.
It is possible to over think things. Digital is not "made up" from sine waves.
you should write a paper to prove Fourier wrong then, rather than just saying I think hes wrong because... (insert handwaving anecdotes)
anyway if you prove yourself correct, you have then dismantled the only analogue mechanism youve got left.
the mechanism doesnt matter, it could be clapping hands
so the ripple, why are these sine waves then? where did they come from? some energy is lost from the edge of the transient and the ripple is the result of that.
anyway if you prove yourself correct, you have then dismantled the only analogue mechanism youve got left.
the mechanism doesnt matter, it could be clapping hands
so the ripple, why are these sine waves then? where did they come from? some energy is lost from the edge of the transient and the ripple is the result of that.
Last edited:
If playing from SD card (direct connected to the CPU), SD card (connected via USB), SATA HDD (direct connected), SATA HDD (galvanic isolated), SATA SSD (direct connected), SATA SSD (galvanic isolated), NAS etc. and the played source are one original that are copied to the other sources - they all play bit perfect, but they sound slightly differently...
Nothing strange - as it is the differences in the PSU / GND noise spectrums we are able to detect.
The differences are not only caused by the differences in hardware, but also due to the differences in the software code that runs and if the data are moved in/out with DMA or PIO etc..
If video (for a visual image) are used there will also be slightly differences in the video quality due to the same issues.
Thus I have custom hardware with galvanic isolated SATA, USB, I2S, I2C, clocks (also 1ps jitter clocks for CPU, USB, ethernet etc..) etc. and custom modified and compiled kernel, ALSA and MPD - for my music player. And power supplies with hundred of thousands uF and multiple levels of regulation / filters.
And last but not least custom cables..
And due to there are no real distinction between "digital" and "analog" components or signals - I use "analog" components in "digital" circuits and "digital" components in "analog" circuits.
Simply because sometimes the available "digital" components creates more noise, adds more jitter and are slower (rise and fall times) than an alternative "analog" component..
As an example - a "digital" inverter can be used as a "analog" amplifier as the "digital" inverter are made up from one P and one N channel FET etc..
"Many" - not all - maybe 20% - are able to hear the subtle differences in fidelity - the remaining 80% do not..
Often the ones that have the ability to hear the subtle differences do not have the technical skills or the "reporter" skills needed to present their "observations" correctly. Thus it normally end up like in this thread -> a complete "mumbo jumbo" discussion where the cause and the symptoms are mixed up and all kind of "OT" "mumbo jumbo" pollutes and destroys.
Nothing strange - as it is the differences in the PSU / GND noise spectrums we are able to detect.
The differences are not only caused by the differences in hardware, but also due to the differences in the software code that runs and if the data are moved in/out with DMA or PIO etc..
If video (for a visual image) are used there will also be slightly differences in the video quality due to the same issues.
Thus I have custom hardware with galvanic isolated SATA, USB, I2S, I2C, clocks (also 1ps jitter clocks for CPU, USB, ethernet etc..) etc. and custom modified and compiled kernel, ALSA and MPD - for my music player. And power supplies with hundred of thousands uF and multiple levels of regulation / filters.
And last but not least custom cables..
And due to there are no real distinction between "digital" and "analog" components or signals - I use "analog" components in "digital" circuits and "digital" components in "analog" circuits.
Simply because sometimes the available "digital" components creates more noise, adds more jitter and are slower (rise and fall times) than an alternative "analog" component..
As an example - a "digital" inverter can be used as a "analog" amplifier as the "digital" inverter are made up from one P and one N channel FET etc..
"Many" - not all - maybe 20% - are able to hear the subtle differences in fidelity - the remaining 80% do not..
Often the ones that have the ability to hear the subtle differences do not have the technical skills or the "reporter" skills needed to present their "observations" correctly. Thus it normally end up like in this thread -> a complete "mumbo jumbo" discussion where the cause and the symptoms are mixed up and all kind of "OT" "mumbo jumbo" pollutes and destroys.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.