Topping D30 DAC Measurement and Review

[Julf]

Evidence of what? Observations that after switching all devices, supporting this, to MSI-mode; setting the Windows operating system thread manager to “Long, Fixed, No foreground boost” intervals of context switching; setting the system directs all drivers and all executables to use Large-Page memory regions; setting of Heap De Commit Free Block Threshold and Heap De Commit Total Free Threshold parameters increased by several orders of magnitude higher than the officially supported by M$; increasing the size of the default Heap Segment to the maximum - in total led to the sound of the DAC became as if I had changed it to something from a completely different market niche and price category?
Do the same, maybe then you will have something to discuss from your own experience. This is the essence of the "scientific method". And not in the choral bleating of the mantra: "double blind, double blind".

Yes, evidence for exactly that. All we have is your claim. An important part of the scientific method is "verify". How did you do that? How did you rule out expectation bias and placebo effects? You already stated that you only matched levels by ear. An important part of science is also to make sure that your test actually tests for what you think it is testing.

 

[Vosya]

An important part of science is also to make sure that your test actually tests for what you think it is testing.

Well, take it and check it, as it is customary to do in the scientific world after publication: setting up an experiment, first of all, one that can give a result that refutes this theory. If you have enough knowledge of the listed terms.

 

[Julf]

Well, take it and check it, as it is customary to do in the scientific world after publication: setting up an experiment, first of all, one that can give a result that refutes this theory. If you have enough knowledge of the listed terms.

The burden of proof is with the one making claims.

 

[Vosya]

The burden of proof is with the one making claims.

I'm sure you just didn't understood what was changed in the operating system. It is pointless to prove something to such a contingent.

 

[Julf]

I'm sure you just didn't understood what was changed in the operating system. It is pointless to prove something to such a contingent.

You seem very sure of a lot of things.

 

[Vosya]

You seem very sure of a lot of things.

Do you have anything to say on the topic of discussion? Something technical / technological where you actually have special knowledge? I was tired of wasting time on bickering with windbag.

 

[Julf]

Do you have anything to say on the topic of discussion? Something technical / technological where you actually have special knowledge?

I suggest you read back in this thread. I am done, if someone has to resort to silly, childish insults any discussion is pointless.

 

[Vosya]

childish insults any discussion is pointless.

Well, quite the expected result. Lack of knowledge encourages attempts to disguise this under any pretext. I have not received a single counterargument regarding the functioning of Windows. Making a reasonable conclusion about your complete ignorance of the functioning of this operating system, I will wait for the next opportunity to communicate with someone who has a completely opposite opinion, but at least has a basis for it.

 

[Julf]

Well, quite the expected result. Lack of knowledge encourages attempts to disguise this under any pretext. I have not received a single counterargument regarding the functioning of Windows. Making a reasonable conclusion about your complete ignorance of the functioning of this operating system, I will wait for the next opportunity to communicate with someone who has a completely opposite opinion, but at least has a basis for it.

Indeed, I am extremely happy to have nothing to do with Windows, as I prefer better operating systems. You did make general statements such as "USB doesn't use DMA", when I guess what you meant was "the driver I am using on Windows doesn't use DMA". Even a non-DMA diver is interrupt driven with buffering.

 

[Vosya]

Even a non-DMA diver is interrupt driven with buffering.

Prove it, certified Windows connoisseur. Come on, disgrace yourself further.

Р.S. It seems to me that if such as ignorant people do not shut their mouths in time, then the last remnants of the common sense of this resource will lose their meaning. There is not much of it left.

 

[Julf]

Prove it.

That is like asking me to prove gravity exists. Please describe how you think a device driver fro USB works.

 

[Vosya]

Please describe how you think a device driver fro USB works.

I'm not a missionary to preach to the ignorant. For these purposes, there is https://docs.microsoft.com/

 

[Julf]

I'm not a missionary to preach to the ignorant. For these purposes, there is https://docs.microsoft.com/

I asked for how *you* believe it works. I suggest you educate yourself in how the Windows USB I/O subsystem (and DMA) actually works. How about starting here?

"As others have explained, the lowest-level USB driver that talks to the
root hub does DMA. "

Instead of silly insults, how about providing actual evidence for your claims? Post jitter spectrums with different kernel parameters.

 

[Julf]

Evidence of what? Observations that after switching all devices, supporting this, to MSI-mode; setting the Windows operating system thread manager to “Long, Fixed, No foreground boost” intervals of context switching; setting the system directs all drivers and all executables to use Large-Page memory regions; setting of Heap De Commit Free Block Threshold and Heap De Commit Total Free Threshold parameters increased by several orders of magnitude higher than the officially supported by M$; increasing the size of the default Heap Segment, System Pages, Paged Pool Size, Non Paged Pool Size to the maximum - in total led to the sound of the DAC became as if I had changed it to something from a completely different market niche and price category?

Can you post jitter spectra before and after those tweaks?

 

[AnalogSteph]

I was actually relieved to find out that DMA had been introduced with USB 2.0. My old PC (dual PIII Coppermine on i440BX, USB 1.1 only) had a habit of dropping parts of USB keyboard input under load, which obviously never happened with a PS/2 keyboard (interrupt-driven instead of polling). More modern machines being much faster obviously would be helping, too...

 

[Vosya]

I was actually relieved to find out that DMA had been introduced with USB 2.0.

Universal Serial Bus (USB 1.1) OHCI Host Controller CAN use DMA to reduce memory latencies. Whether this potential was used by the operating system is the question. This is exactly what is being discussed here.

My old PC (dual PIII Coppermine on i440BX, USB 1.1 only) had a habit of dropping parts of USB keyboard input under load, which obviously never happened with a PS/2 keyboard (interrupt-driven instead of polling). More modern machines being much faster obviously would be helping, too...

HID such as mice, keyboards - not a bulk endpoints. With Full-Speed interrupt transfer at 12 Mbps any wretched hardware coped. IBM ThinkPad R31, which I have and which still works - handles external usb keyboard and mouse without any problem.

 

[Vosya]

Can you post jitter spectra before and after those tweaks?

Do I have to run to the store, buy an APx555 to do this?
Suggest investigate this to those, who posts a peerless jitter spectrum of a DAC, that does not match the real conditions of users.

 

[Vosya]

I asked for how *you* believe it works. I suggest you educate yourself in how the Windows USB I/O subsystem (and DMA) actually works. How about starting here?

I don't care about hastily found gossip from the first forum you see. Point me to an authorized statement from microsoft that Windows 7 x86 USB uses DMA.

 

[Julf]

I don't care about hastily found gossip from the first forum you see. Point me to an authorized statement from microsoft that Windows 7 x86 USB uses DMA.

Ah, moving goalposts by narrowing the target. I think this is the first time you mention that your claims are specifidc to Windows 7 x86.

 

[Vosya]

Ah, moving goalposts by narrowing the target. I think this is the first time you mention that your claims are specifidc to Windows 7 x86.

These are not my claims, but the very first mention of the operating system, on the basis of which you tried to teach me something with your link. Although, yes, Windows version: 32-bit (x86) or 64-bit (x64) was really not indicated there.