51 (edited by unpluggged 2024-05-21 09:03:17)

Re: ADI-2 DAC: Slow filter at higher sample rates

switch6343 wrote:

Again, ChatGPT gave me the answer, I was looking for.

I hope you don't refer to ChatGPT for medical advices roll

52 (edited by ramses 2024-05-21 09:27:44)

Re: ADI-2 DAC: Slow filter at higher sample rates

Sorry, I don't get involved in every topic. I just wanted to remind you of what KaiS said above in posting #43 about ChatGPT and share my experience with it.

KaiS wrote:

Wow, this had to happen sooner or later - have to argue with ChatGPT:
A good example how ChatGPT mixes true and wrong assumptions it has collected from … - where?

Frankly, I would rather not read increasingly in forums what an AI is rambling on about, knowing that you can't rely on the statements.

Coupled with a certain belief in technology/AI, this will simply lead to “exciting” or should I better say "unnecessary" discussions and irritations in the sense of “the AI said it”.

It will also make working in the forum more difficult. Instead of simply being able to answer a question, you have to deal with all the AI's "verbiage", which is peppered with wrong, partly right and right answers. But it's far too much work having to review this information.

Above all, you may have to argue why you and not the AI are right. Let's be honest, nobody needs that.

AI is helpful in areas with more control over the quality of training data, e.g. medical area where you also have to meet certain quality standards.

BR Ramses - UFX III, 12Mic, XTC, ADI-2 Pro FS R BE, RayDAT, X10SRi-F, E5-1680v4, Win10Pro22H2, Cub13

Re: ADI-2 DAC: Slow filter at higher sample rates

ramses wrote:

It makes no sense to bring the answers from KI tools into discussions. Trust the real experts instead. I have asked the AI in areas I know a lot about on a trial basis and the quality of the answers is just scary.

Think about it, the AI is fed with so much data, who knows where it all comes from and no one can tell me that something like quality could be ensured with so much different data and data sources. You can also see it in the quality of the answers. Some things may be true, others not. As in many areas, “shit in, shit out”.

I didn't receive any answer from an "expert" to my question.

I expected from so called "experts" on this forum to respond rationally and without prejudice. Why not respond objectively and constructively? Why responding non-factually and emotionally, why exposing personal prejudice to the audience at large?

I did not ask for an opinion on ChatGPTs answers to prompts. 

It seems some respondents are bathing themselves in self-indulgence.

Re: ADI-2 DAC: Slow filter at higher sample rates

switch6343 wrote:
ramses wrote:

It makes no sense to bring the answers from KI tools into discussions. Trust the real experts instead. I have asked the AI in areas I know a lot about on a trial basis and the quality of the answers is just scary.

Think about it, the AI is fed with so much data, who knows where it all comes from and no one can tell me that something like quality could be ensured with so much different data and data sources. You can also see it in the quality of the answers. Some things may be true, others not. As in many areas, “shit in, shit out”.

I didn't receive any answer from an "expert" to my question.

I expected from so called "experts" on this forum to respond rationally and without prejudice. Why not respond objectively and constructively? Why responding non-factually and emotionally, why exposing personal prejudice to the audience at large?

I did not ask for an opinion on ChatGPTs answers to prompts. 

It seems some respondents are bathing themselves in self-indulgence.

Sorry, but I don't reply to such unobjective and emotional comments.

BR Ramses - UFX III, 12Mic, XTC, ADI-2 Pro FS R BE, RayDAT, X10SRi-F, E5-1680v4, Win10Pro22H2, Cub13

55 (edited by Kubrak 2024-05-21 14:21:21)

Re: ADI-2 DAC: Slow filter at higher sample rates

@switch6343

That ChatGPT answer may be right, or not. IMHO it is just guessing.... One would have to look at specsheet.

The fact is that when DA is performed the filter should remove frequencies above the half sampling frequency. But designer may decide to place filter at lower frequency...

Re: ADI-2 DAC: Slow filter at higher sample rates

unpluggged wrote:
switch6343 wrote:

Again, ChatGPT gave me the answer, I was looking for.

I hope you don't refer to ChatGPT for medical advices roll

I guess you are a self proclaimed, would-be, expert who fell and falls victim of your hubris and vanity. You're ignorant and your response proves you're uneducated and badly mannered. You behave like a parrot. The substance of your responses is zero. In fact, you have no clue whatsoever.

Re: ADI-2 DAC: Slow filter at higher sample rates

The problem of ChatGPT is, that it mixes true, educated guesses and totall nonsence so well that it is hard to ditstinguish what is what cattegory....

It is just great Baron Munchausen....

58 (edited by KaiS 2024-05-22 06:36:35)

Re: ADI-2 DAC: Slow filter at higher sample rates

The KI’s answer is unprecise:

Sample rate vs. Sharp Filter cutoff start:
SR 44.1 kHz - 20 kHz
SR 48 kHz - 22 kHz
SR 96 kHz - 44 kHz

So the filter kicks in at ca. 1/2 sample rate -10%, the margin it needs for the transition from pass- to stop-band.

ADI-2 Pro‘s manual page 86, 34.14 DA Filter Curves 44.1 kHz shows the frequency responses of the various DA-filters.


I can‘t hear any sound difference between the filters at SR‘s above 48 kHz.
So the selection of filters is irrelevant from 88.2 kHz and above.

Re: ADI-2 DAC: Slow filter at higher sample rates

switch6343 wrote:

Again, ChatGPT gave me the answer, I was looking for.

This was my prompt:
MY RME DAC has an DA Filter SD Sharp, which filters from about 20 kHz upwards with sample ratels of 44.1 Khz and 48 kHz. However, when the sample rate is for instance 92 kHZ, does the SD Sharp Filter filter from 48 kHz upwards?

...........
Explanation deleted
..........

So, to answer your question: Yes, when the sample rate is 96 kHz, the SD Sharp filter will start filtering from around 48 kHz upwards. This scaling ensures that the filter continues to remove unwanted frequencies and noise that are above the audible range and could cause issues with digital audio processing.

False, as it is around 44 kHz, not 48 kHz. ;-) ChatGPT did not know, just guessed that it is about half of 96 kHz....

Re: ADI-2 DAC: Slow filter at higher sample rates

KaiS wrote:

The KI’s answer is unprecise:

Sample rate vs. Sharp Filter cutoff start:
SR 44.1 kHz - 20 kHz
SR 48 kHz - 22 kHz
SR 96 kHz - 44 kHz

So the filter kicks in at ca. 1/2 sample rate -10%, the margin it needs for the transition from pass- to stop-band.

ADI-2 Pro‘s manual page 86, 34.14 DA Filter Curves 44.1 kHz shows the frequency responses of the various DA-filters.


I can‘t hear any sound difference between the filters at SR‘s above 48 kHz.
So the selection of filters is irrelevant from 88.2 kHz and above.


KaîS, thank you for your response. That's what I was looking for.