[Audiophiliac]

Apps don't necessarily change audio. A lot of computers play back audio through their built in audio board. It doesn't matter what app is being used, it all goes through the same playback hardware.

One of the misconceptions about digital audio is that everything has an audible effect on the sound. That isn't true. Most modern digital audio equipment and software are audibly transparent... what goes in sounds exactly the same as what comes out.

Dithering affects the depth of the noise floor. It doesn't affect the sound above that at all. You can find a link to a comparison of dithered and undithered audio in the video link in my sig file. It isn't a great deal of difference, especially if you don't listen to your music very loud.

By the way, it's a lot easier to have a conversation with you when you make shorter posts like this one. When you do those long stream of consciousness rants flying all over dozens of different subjects, it's difficult to know where to even start, so it makes me want to just dismiss the whole thing.

Apps don't necessarily change audio? You haven't hear the differences between players? Or tried Kernel Streaming output to bypass all but the tiny first bit of the OS with your app?
Sorry for the long streams all over the place.

 

[Audiophiliac]

Audiophiliac: Staircases are just a way to visualize digital audio. Physically staircase signals are impossible, because they require infinite bandwidth. DACs have reconstruction filters which make the signals completely smooth.

But the staircases visualized of you look at the diagram are representing the time between the sound changing at all. Further, those diagrams only ever show you low notes wobbling. The problem is with the higher frequencies, where the higher they go, the more likely they are not steps on the way up or down, but are actually wobbling so fast that they could be on the downswing instead of stepping up. Fake noise.
No, it can't get that bad, because it takes 2 samples to play a sound. Hence a 22khz upper limit with redbook as a crude both ends of the wave.
Bassheads will still be ok with digital compared to the rest of us. Below 20hz will be totally smooth. You can always argue it could be smoother though.

 

[bigshot]

Apps don't necessarily alter sound. They may have filters that can be turned on or off that do process the sound, but in a lot of cases, the digital audio passes through the same playback system as if you weren't using the app. Every app on my Mac passes the sound through Apple's own audio processor, and yes, they do all sound identical with no processing applied.

Staircases are not an accurate way of visualizing digital audio. If you have an audio editing app, you can feed a tone into it and look at the waveform blown up huge and there are no stair steps. Waveforms are smooth from point to point. Also, if there was a problem with that, it would show up in high frequencies like 20kHz where there are only two points to define a frequency, not in bass notes where there are tens of thousands.

To PERFECTLY reproduce a frequency, it takes double the sampling rate. 44.1 can PERFECTLY reproduce every frequency up to 20kHz, which is the upper limit of human hearing. This is part of the basics of digital audio. You might want to read up on Nyquist Theory at wikipedia. https://en.wikipedia.org/wiki/Nyquist–Shannon_sampling_theorem

Folks here who know all about it are happy to answer any questions you have. If you aren't sure, it's best to phrase them as questions, rather than as statements that are based on guesses about how things work. Knowing what you don't know can be more important than knowing what you do know sometimes.

 

[Audiophiliac]

Apps don't necessarily alter sound. They may have filters that can be turned on or off that do process the sound, but in a lot of cases, the digital audio passes through the same playback system as if you weren't using the app. Every app on my Mac passes the sound through Apple's own audio processor, and yes, they do all sound identical with no processing applied.

I can hear a difference between all players, so far. The quest for the best player goes on. Tidal's default is the least bad of streaming services I've tried so far. My favorite player is Audirvana. It sounds the least of all like there is a player in there somewhere. With the kernel streaming output, it sounds like it passes through less OS software than WASAPI, and sounds more natural because of it. Natural will seem like exactly the right word to use.

Staircases are not an accurate way of visualizing digital audio. If you have an audio editing app, you can feed a tone into it and look at the waveform blown up huge and there are no stair steps. Waveforms are smooth from point to point. Also, if there was a problem with that, it would show up in high frequencies like 20kHz where there are only two points to define a frequency, not in bass notes where there are tens of thousands.

Staircases are not an accurate way of visualizing audio, but with digital audio, staircases are an accurate way of looking at what is going on during each sample. It's ugly, compared to what you're telling me it should be, right? That's why we want higher resolution. With samples, they will never be gone. The quest is to get the jagged steps to look like you want them to be. They'll never get there as samples, increasing resolution will improve infinitely.

To PERFECTLY reproduce a frequency, it takes double the sampling rate. 44.1 can PERFECTLY reproduce every frequency up to 20kHz, which is the upper limit of human hearing. This is part of the basics of digital audio. You might want to read up on Nyquist Theory at wikipedia. https://en.wikipedia.org/wiki/Nyquist–Shannon_sampling_theorem

Again, they are not a smooth curve, the 20khz sounds are the most jagged possible to still be able to even play. It was the minimum they went with, otherwise someone would complain they couldn't hear something that was supposed to be there.
Watch out for all that stuff that talks about you at 30khz, it all sees you.

"Folks here who know all about it are happy to answer any questions you have. If you aren't sure, it's best to phrase them as questions, rather than as statements that are based on guesses about how things work. Knowing what you don't know can be more important than knowing what you do know sometimes."
If I have a question, I'll ask. Thank you for telling me what you were thinking. I wish the waves weren't staircased, but you have to have steps when it's only samples. Samples just say you have to play that at some point, they're not what's going on, your chip just wings that. Gimme a break it's not smart just to be playing what that sample works out to for as long as a sample lasts, though. It should work out to hearing a more accurate starting and stopping of things constantly if you increase resolution.

 

[bigshot]

If you’re not interested in responses, I don’t have to respond. This is pretty basic stuff and I know what I’m talking about here. I even provided you with links to verify what I said. You just dismissed what I shared with you without acknowledging it. When you just blow by information and go back to repeating your misconceptions as if I wasn’t here, it doesn’t encourage me to share information with you. Shall I just dismiss your post like you just did mine, or are you actually interested in what I have to say?

 

[Audiophiliac]

If you’re not interested in responses, I don’t have to respond. This is pretty basic stuff and I know what I’m talking about here. I even provided you with links to verify what I said. You just dismissed what I shared with you without acknowledging it. When you just blow by information and go back to repeating your misconceptions as if I wasn’t here, it doesn’t encourage me to share information with you. Shall I just dismiss your post like you just did mine, or are you actually interested in what I have to say?

Your posts have only ever been dismissing my posts.
You can go ahead and think your chip is smart enough to make it sound like what the recording actually sounded like, using samples, if you like. If you want to think your 44.1khz frequency of data knows what the curves are in between, why use more than 1 sample per second?
This is you thinking you're a bigshot because your digital has been perfect since 18 years ago, when you first joined.

 

[sander99]

If you want to think your 44.1khz frequency of data knows what the curves are in between, why use more than 1 sample per second?

Because with 1 sample per second only the "curves" of all frequencies below 0.5 Hz can be reconstructed.
(With 44100 samples per second the "curves" of all frequencies below 22050 Hz can be reconstructed.)

 

[gregorio]

I never said foobar2k was a dac, moron.

Exactly, moron! Foobar is not a DAC, it does not convert the digital data to an analogue signal, moron!

I don't care about turning on dithering, I'd rather it play direct.

Why do you have it turned on?

You should be careful about which one of us looks stupid.

I am but obviously you don’t mind looking stupid at all. In fact your posts indicate you deliberately want to look stupid.

I haven't been saying I don't understand something.

Exactly and that’s the problem, clearly you have no understanding but complete ignorance isn’t stopping you from arguing about it.

Unaltered stair steps for me, I guess.

And where are you going to find these “unaltered stair steps” that don’t exist?!

The problem with upsampling must be that the higher the frequency of the soundwaves, the faster they wobble, so the more likely they are supposed to be on the downside by the next sample, could be anywhere, causing it to sound like a chip and noisy.

More complete nonsense, well done!

But the staircases visualized of you look at the diagram are representing the time between the sound changing at all.

There are no staircases and the graphic visualisations depicting staircases do NOT represent the waveform between the sample points. Just more nonsense!

G

 

[Audiophiliac]

Because with 1 sample per second only the "curves" of all frequencies below 0.5 Hz can be reconstructed.
(With 44100 samples per second the "curves" of all frequencies below 22050 Hz can be reconstructed.)

Would you be able to draw a connect the dots picture, if they made the Sistine Chapel into them? Sure, you could do it perfectly. Would it be identical to the original Sistine Chapel? I bet you'll keep saying, yeah but if I had more dots...

 

[gregorio]

Would you be able to draw a connect the dots picture, if they made the Sistine Chapel into them? Sure, you could do it perfectly. Would it be identical to the original Sistine Chapel?

That’s a nonsense analogy, a DAC chip isn’t drawing the Sistine Chapel, it’s just recreating a simple sinusoid. If the Sistine Chapel were just a simple sinusoid then yes it would be identical.

I bet you'll keep saying, yeah but if I had more dots...

Of course I’ll keep saying “yeah”, as it was proven over 70 years ago. Can you prove that the Nyquist/Shannon Theorem is incorrect? If not then you’re just spouting nonsense again! And “NO” more dots make absolutely no difference at all!

Watch the video supplied to you and stop with all the utter nonsense!!

G

 

[Audiophiliac]

Exactly, moron! Foobar is not a DAC, it does not convert the digital data to an analogue signal, moron!

If you show me where I said foobar was a dac, i'll be the moron instead of you.

Why do you have it turned on?

I said I'm not going to turn it on, show me where I said I did. If I had, I would be the one saying that dithering is doing something.

I am but obviously you don’t mind looking stupid at all. In fact your posts indicate you deliberately want to look stupid.

Then why can't you make me understand? Not enough posts?

Exactly and that’s the problem, clearly you have no understanding but complete ignorance isn’t stopping you from arguing about it.

And where are you going to find these “unaltered stair steps” that don’t exist?!

The original file is unaltered stair steps. That's how I want it to arrive at my dac.

More complete nonsense, well done!

There are no staircases and the graphic visualisations depicting staircases do NOT represent the waveform between the sample points. Just more nonsense!

Your dac does not know how to curve between samples. Why do you think they show you a graph with steps in the first place, if there was no difference between digitizing and recording the file identically to what it originally was?

 

[Audiophiliac]

That’s a nonsense analogy, a DAC chip isn’t drawing the Sistine Chapel, it’s just recreating a simple sinusoid. If the Sistine Chapel were just a simple sinusoid then yes it would be identical.

Audio is not usually a simple sinusoid, at all. A simple sinusoid is 1 frequency playing only. You could probably only do that electronically. Most sounds are a range of a frequency. Things can sound better that way. That's why stringed instruments are hollow, wood, and curved the way they are, for the range and tone of a good sound.
And if you tried to connect the dots of a simple sinusoid, would you try straight lines to each dot, or do you try to make your own curve in between? Which would work better?

 

[bigshot]

Your posts have only ever been dismissing my posts.
You can go ahead and think your chip is smart enough to make it sound like what the recording actually sounded like, using samples, if you like. If you want to think your 44.1khz frequency of data knows what the curves are in between, why use more than 1 sample per second?
This is you thinking you're a bigshot because your digital has been perfect since 18 years ago, when you first joined.

I haven't dismissed your posts.. I took the time to explain the way digital audio works. That's contributing to the conversation. But if you consider that to be a dismissal, I'm happy to follow your wishes and dismiss your incorrect statements. I'll save energy and just point out when you're wrong without explaining why. If you'd ever like to find out why you are wrong, I'm happy to oblige if you want to start respecting my contribution to the conversation.

 

[bigshot]

The original file is unaltered stair steps. That's how I want it to arrive at my dac.

This is incorrect

Your dac does not know how to curve between samples.

This is incorrect.

 

[71 dB]

You can go ahead and think your chip is smart enough to make it sound like what the recording actually sounded like, using samples, if you like. If you want to think your 44.1khz frequency of data knows what the curves are in between, why use more than 1 sample per second?

The problem here is you don't understand how little you know about digital audio. You don't get it that you could learn from us.

Simply put: The DAC "knows" what shape the original analog signal had "between" the sample points because there is only ONE analog signal possible that is bandlimited to half of the sampling frequency and goes through the sample points. The reconstruction filter gives you the smooth original signal automatically, because that's how the math works. In order for this to works, we need to take samples at least twice the highest frequency in signal. That's why 44.1 kHz sampling for audio that goes up to 20 kHz.