[71 dB]

ok, let's go back to square wave. do you think the following is a bandlimited signal?



Yes? No? Maybe?

Yes.

Have a look of the following:



Do you know what I just shown above?

Of course I do. This is very basic stuff for someone with my education.

If not, have a look of the following:



The above graph show a band limited 100 Hz "square wave". Of course, it is not a perfect square wave but it is good enough for a lot of analysis.

Yep. Now, do you have a point? Where is this going?

 

[sunjam]

Not your fault, you weren’t to know that you picked a specific area of digital audio which is currently a hot marketing point for snake oil products and will therefore attract audiophile trolls with a vested interest, who pretend to be engineers or combating pseudoscience while doing the exact opposite.

No, it’s identical. Upsampling isn’t a new thing, some of the first generation of CD players back in 1984 were upsampling. The trolls typically ignore the fact that upsampling and downsampling is a required part of music recording, mixing and mastering. In fact typically, the commercial recordings you’re listening to have been upsampled and downsampled again numerous times throughout the process, sometimes several dozen times. If, as has been falsely claimed, upsampling is “night and day noticeable”, then several dozen upsampling and downsampling cycles should make the recording utterly unrecognisable but obviously that is not the case, as any real engineer would know!

Of course not, how is that even a sensible question? The thread you started was closed down precisely because you refused to have a “constructive discussion” regarding various points, including square waves. You just lied, cherry-picked, misrepresented and misquoted what was stated instead, which of course completely precludes any possibility of a “constructive discussion”! How is that not obvious?

How exactly do you think that you presenting more BS and pseudoscience in a science discussion forum is a “help”?

I can only assume that your incessant trolling is a deliberate attempt to get yourself banned, seeing as castleofargh has already warned you several times. Presumably so you can play the victim again and claim how you were banned for some unknown reason.

G

Not all upsamplers are created equally.

Do you consider the 100Hz "square wave" I posted is band-limited?

 

[sunjam]

Yes.


Of course I do. This is very basic stuff for someone with my education.


Yep. Now, do you have a point? Where is this going?

Cool, someone finally agreed that it is a band-limited "square wave". Yeah.

I want to highlight that testing with a simple sine wave input could miss a lot of details as our music is not a pure sine wave. We need input with multiple frequencies as some artifacts can only be seen easily with input that have more than one frequency.

A band-limited "square wave" is full of harmonics and it is a good input for real analysis.

p.s.: I believe that not all people in audio science forums has your level of education. I think it is good to share this knowledge.

 

[danadam]

For your impulse (i.e. input), looks to me it is a single frequency with decreasing amplitude.

Somehow I'm not surprised that you would think that. It's actually all the frequencies:

Is it possible to use a band-limited 100 Hz "square wave" that shown earlier in my reply as input?

I'm worried that you'll take it and misrepresent in your usual fashion, but ok. Everything as before, I only changed the Ref Level in ADI-2 from +13 dBu to +4 dBu because the max output from the laptop is small enough. A 100 Hz square wave, band-limited to 20 kHz, generated at 352.8 kHz:

Left channel delayed, converted to 16/44k:

Played on laptop, captured at 705.6 kHz:

 

[sunjam]

Somehow I'm not surprised that you would think that. It's actually all the frequencies:



I'm worried that you'll take it and misrepresent in your usual fashion, but ok. Everything as before, I only changed the Ref Level in ADI-2 from +13 dBu to +4 dBu because the max output from the laptop is small enough. A 100 Hz square wave, band-limited to 20 kHz, generated at 352.8 kHz:




Left channel delayed, converted to 16/44k:



Played on laptop, captured at 705.6 kHz:

Excellent, thanks a lot again for your reply.

Could you tell me which software you used to generate the 100Hz sqare wave and how you band-limit it to 20kHz?
And which software you used to down sample the 352.8 kHz to 44.1 kHz?

I'm really happy to see your reply as it prompts me to think of an easier/better way to present the information I wanted to show.

BTW, given you use the 100Hz band limited "sqaure wave" in your experiment, I'd assume that you agree the 100Hz band limited "square wave" is a legitimate input for digital audio (instead of illegal as someone mentioned earlier in this thread). Am I correct?

Cheers

 

[bigshot]

Digital audio wasn’t designed to reproduce square waves. It isn’t a signal that is allowed in commercially recorded music. It doesn’t matter because music doesn’t contain square waves. Complaining that a DAC doesn’t handle square waves well is like complaining a camera doesn’t reproduce x rays well… it was never intended to do that, and even if it did, your snapshots wouldn’t require it.

 

[sunjam]

Digital audio wasn’t designed to reproduce square waves. It isn’t a signal that is allowed in commercially recorded music. It doesn’t matter because music doesn’t contain square waves. Complaining that a DAC doesn’t handle square waves well is like complaining a camera doesn’t reproduce x rays well… it was never intended to do that, and even if it did, your snapshots wouldn’t require it.

Is electronic music not allowed in commercially recorded music? I just showed a drum machine made by Korg that can produce sqaure wave.

Here is the photo again:

You made a very good point about camera, I'm just complaining a camera doesn't have a good sensor.

Full-frame 135-format digital camera is good enough for a lot of usage. However, we are talking about 120-format full-frame professional digital camera back vs 135-format full-frame consumer-grade digital cameras. Personally, I am happy with my Olympus 4/3 system. LOL (if you are into digital camera, you should know what I am talking about).

 

[bigshot]

You can generate a signal but that doesn't mean that digital audio was designed to reproduce it. Square waves are irrelevant to home audio. Home audio is designed to produce a perfect signal for the purposes of listening to music with human ears.

 

[sunjam]

You can generate a signal but that doesn't mean that digital audio was designed to reproduce it. Square waves are irrelevant to home audio. Home audio is designed to produce a perfect signal for the purposes of listening to music with human ears.

If I understand your statement correctly, it means digital audio and home audio systems are not designed for electronic musics (which use a lot of signal generators, aka synthesizer)?

In other words, based on your statement, electronic music live in the analog domain but not in digital domain.

Am I correct in understanding your statement?

 

[danadam]

Could you tell me which software you used to generate the 100Hz sqare wave and how you band-limit it to 20kHz?
And which software you used to down sample the 352.8 kHz to 44.1 kHz?

I'm using SoX for nearly everything.
For generating the band-limited square wave I have this script, which uses SoX internally.

instead of illegal as someone mentioned earlier in this thread

Earlier you showed non-band-limited square wave and that's what's illegal:

I got the following graph from a fellow member.

For comparison, 100 Hz square wave at 44.1 kHz sampling rate, bandlimited vs not bandlimited:

 

[sunjam]

I'm using SoX for nearly everything.
For generating the band-limited square wave I have this script, which uses SoX internally.


Earlier you showed non-band-limited square wave and that's what's illegal:

For comparison, 100 Hz square wave at 44.1 kHz sampling rate, bandlimited vs not bandlimited:

Wonderful, thanks for your reply. I'll play around these and see if I have a better way to present the information I wanted to show earlier.

 

[sunjam]

You can generate a signal but that doesn't mean that digital audio was designed to reproduce it. Square waves are irrelevant to home audio. Home audio is designed to produce a perfect signal for the purposes of listening to music with human ears.

Hmm.. having a second thought isn't that the Monty's version of "sampling theory" claim "[band-limited] analog signal can be reconstructed losslessly, smoothly, and with the exact timing of the original analog signal"?

Now, you are saying there are more limitations as band-limited "square wave" is not included. So you also agreed that the claim is not correct?

 

[bigshot]

Digital audio is designed to be perfect for the purpose of reproducing recorded music in the home to be listened to with human ears. Fixating on signals that the system wasn't designed to reproduce and don't exist in recorded music is mental masturbation. It doesn't matter if a DAC reproduces it. You don't need it and it isn't in the music you're playing.

 

[71 dB]

Cool, someone finally agreed that it is a band-limited "square wave". Yeah.

Who doesn't agree to that? Of course band-limited square waves are band-limited! All square waves are band-limited. An ideal square wave is just an abstract concept like the number infinity.

I want to highlight that testing with a simple sine wave input could miss a lot of details as our music is not a pure sine wave. We need input with multiple frequencies as some artifacts can only be seen easily with input that have more than one frequency.

Sine waves are used, because it is very easy to graphically see what is happening to the signal, but you are right. The behaviour of a linear, time invariant system is described by its impulse response. The measurements of impulse responses use signals consisting of all frequencies (say 0-20 kHz).

A band-limited "square wave" is full of harmonics and it is a good input for real analysis.

Not really. It contains only the fundamental frequency + harmonics and nothing else. MLS (Maximum Length Sequence) is often used to measure the impulse response of a system, but there are other options such as sinusoidal sweeps.

Step signal can be used to study the behaviour in time visually. It is often much easier to see all kinds of things that are actually inaudible, because visual representations of sound can be zoomed to reveal things we can't hear and because seeing is so different from hearing. Signals such as square waves can be misleading as to how audible the problems are. If you take real music and do some damage to it the result is more difficult to tell apart from the original.

p.s.: I believe that not all people in audio science forums has your level of education. I think it is good to share this knowledge.

There are people in the world who know more than me and people who know less than me about audio science.

 

[gregorio]

Not all upsamplers are created equally.

How can we have a “constructive discussion” (as you falsely claimed you want) if all you do is deflect from all the points put to you, how is that any sort of discussion, let alone a constructive one? And in response to your deflection, So what?

[1] … our music is not a pure sine wave. …
[2] A band-limited "square wave" is full of harmonics and it is a good input for real analysis.

1. That’s a strawman argument, as no one stated “a pure sine wave” what was stated was “sine waves”.
2. What do you think those “harmonics” are, if they’re not sine waves? And, if a square wave is band-limited then by definition it is NOT “full of harmonics” (and is therefore not strictly a “square wave”)!

I'm really happy to see your reply as it prompts me to think of an easier/better way to present the information I wanted to show.

Exactly, that is precisely why we can’t have a “constructive discussion” (as you falsely claimed) and why you’re a troll! The information you “wanted to show” is BS pseudoscience and therefore you are “really happy to see your reply as it prompts me to think of an easier/better way to present” BS pseudoscience! That is why your previous thread was closed and why you’ll likely be banned from this one.

I just showed a drum machine made by Korg that can produce sqaure wave.

Sure you did, you presented a 25 year old drum machine by Korg that according to you had infinite bandwidth and produced actual square waves. $250 for a magical drum machine that broke the laws of physics is a bargain IMO! lol

If I understand your statement correctly, it means digital audio and home audio systems are not designed for electronic musics (which use a lot of signal generators, aka synthesizer)?

Again, whenever you say “am I correct” or “if I understand correctly”, the answer is ALWAYS “No”! You are not correct and you do not understand correctly because you are not trying to understand or be correct, you’re just trying to find ways to present BS!

To answer your question, digital audio and home audio systems ARE “designed for electronic musics” but they’re not designed for magical synthesisers that break the laws of physics and don’t actually exist!

Now, you are saying there are more limitations as band-limited "square wave" is not included.

Of course he’s saying a “band-limited square wave is not included”, for the same reason as digital audio also doesn’t include unicorns, fairies or leprechauns, there’s no such thing! A square wave contains an infinite number of (odd) harmonics, if it’s band limited then it obviously doesn’t contain an infinite number of harmonics and therefore is not an actual square wave.

So you also agreed that the claim is not correct?

How many times? Whenever you post a question like this, the answer is ALWAYS “No”! “No” he did NOT agree, “no” you are not correct! But don’t let the facts stop you from lying and trolling.

G