MaxwellsEq
Major Contributor
- Joined
- Aug 18, 2020
- Messages
- 1,807
- Likes
- 2,753
Please can you unpack this?
-
WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!
Best No oversampling dac to buy??
- Thread starter KehaDNb
- Start date
I've never learned about the Nyquist-Shannon sampling theorem...however, this conversation never happened...Tell me you've never learned about the Nyquist-Shannon sampling theorem without telling me that you've never learned about the Nyquist-Shannon sampling.
I'd observe that theory is one thing, but real life perfect implementation is another. I think some excellent recordings can be made at 16/44 but that aside let us postulate that filtering does cause problems. I keep asking this question for those who might have actual data without answer, but my *belief* is that there is very very little energy actually getting through from a room to a microphone to an electric signal. Then especially if the sampling is high in real life maybe you don't actually need anti-aliasing filtering at all (?!?)
MaxwellsEq
Major Contributor
- Joined
- Aug 18, 2020
- Messages
- 1,807
- Likes
- 2,753
I blame Stephen Butterworth.
when you are sampling at 192kHz, or even DSD, there is no need to do more over-sampling. When you over sampling, you then have to get rid of aliasing by applying a digital domain filter, which is destructive. You either have to use a linear phase filter to get rid of alias or minimum phase for punchy bass but risking some aliasing getting through, neither is perfect.
What?
minimize pre-ringing.
Why would that affect bass?
Killingbeans
Major Contributor
staticV3
Master Contributor
- Joined
- Aug 29, 2019
- Messages
- 8,281
- Likes
- 13,315
Oversampling doesn't create aliasing.
It reduces aliasing by moving images up in frequency, thereby relaxing the requirements of the analog LPF.
Aliasing is at its worst when using NOS.
restorer-john
Grand Contributor
Gotta love a resurrected zombie thread.
I was thinking the same thing! I’m just amused that there’s still interest in discussions about NOS DACs. Or DACs in general, really—they were perfected to audible transparency years ago with the optimization of delta sigma DACs and even the ES9038Pro was overkill—now we’re into the ES9039Pro and AK4499EX with these crazy SINADs, and folks still want to talk about costly R2R and NOS DACs because why? The only discussion left to be had is to topple myths. I guess there are a lot of folks left to enlighten, but boy do they seem to resist marketing pseudoscience!
Please spend some more time on this forum, and open your mind to the wisdom shared here. There are a lot of very wise folks on here who will save you a lot of money. I speak from experience in that regard. And be very wary of claims made on that other big forum—there’s a lot of hogwash making the rounds over there, and beware of corporate funded websites. Your wallet will thank you.
I’m in a mini audiophile ChatGPT phase to see how well it gets it right. Let’s ask it what it thinks about that last statement!
L
I’d say not bad.
LI’d say not bad.
Last edited:
I’m kind of digging that new haircut Brad. You nailed the 60’s look.
oleg87
Senior Member
Pre-ringing has nothing whatsoever to do with bass, or frequencies an adult human can hear for that matter.
I guess for the manufacturers of pricey NOS DACs the stuff that exists outside of the range of human hearing is what pays the mortgages on their McMansions.
Filtering causes "problems", but in general they're not serious problems because they've been worked on for 30+ years now.I've never learned about the Nyquist-Shannon sampling theorem...however, this conversation never happened...
I'd observe that theory is one thing, but real life perfect implementation is another. I think some excellent recordings can be made at 16/44 but that aside let us postulate that filtering does cause problems. I keep asking this question for those who might have actual data without answer, but my *belief* is that there is very very little energy actually getting through from a room to a microphone to an electric signal. Then especially if the sampling is high in real life maybe you don't actually need anti-aliasing filtering at all (?!?)
When it comes to acoustic signals, you are correct that there isn't a lot of ultrasonic energy coming in to the recorder.
However, when it comes to DACs, basically you need the filter to get rid of aliasing... when you have frequencies in a digital signal that are higher than 1/2 the sample rate, they "wrap around" and start making weird noises at lower frequencies.
This thread has some very shocking real-world examples of an AVR that fails to filter properly, with audible aliasing as a result. There's no excuse for this happening in any gear built after 1990 IMO. And this is also why NOS DACs are a terrible idea and (IMHO) only favored by people who don't really understand and/or accept the realities of digital audio.
What I meant was transient attack
At sampling rate of 192 kHz, this moves the Nyquist frequency to 96 kHz and that helps a lot.
Last edited:
Similar threads
