Schiit Happened: The Story of the World's Most Improbable Start-Up
Mar 27, 2023 at 4:12 PM Post #114,602 of 155,168
Mar 27, 2023 at 4:15 PM Post #114,607 of 155,168
Redbook or not a DAC always has a LPF. Always. And yes, on a scope a sine wave (within the audio band) at the output of any DAC does not have stair-steps (that is a bad graphical representation).
Methinks you should revisit Jason & my replies. If you don't have a brickwall you may get stairsteps. Plus the link I posted clearly shows your statement is false. NOS DACs do exist that don't have an LPF.
 
Mar 27, 2023 at 4:15 PM Post #114,608 of 155,168
Ehm... PCM, PDM (as in DSD), PWM - isn't that the end of what you can do with digital? If we define digital as running on two fixed levels and a fixed clock. Maybe one of them can represent an audio signal in some non-standard way that somehow makes it interesting?

Or maybe interesting schiit happens if we mess with the entire definition of digital and start modulating the clock. Not sure about that one :thinking:

Or maybe in which way the digital signal represents the analog one isn't even what he's talking about. But if so I've got no idea what he's talking about!

Very intriguing :L3000:
 
Mar 27, 2023 at 4:43 PM Post #114,613 of 155,168
NOS DACs do exist that don't have an LPF.
(it's just unfathomably stupid)

In my not-so-humble opinion.

If the sampling rate is high enough (which means OS, probably a lot) that the ultrasonics are far beyond what even poses a risk of making the tweeters vibrate at all, and all amplification components will tolerate them without ill effect, and you have cabling and connectors that are standard for HF stuff not audio stuff so that you mitigate the reflections and what not, then... and only then, I think... maybe removing the LPF makes sense. All that together should, I think, result in a system with way-better-than-normal transient response (if you have an extremely good upsampling filter). Worth trying, at that point (I mean if you do have such equipment for other reasons, buying it to test this is definitely not worth it).

But for NOS? Nope. Ultrasonics anywhere close to the audible range are a bad thing, I'm quite sure of that.
 
Mar 27, 2023 at 4:46 PM Post #114,614 of 155,168
(it's just unfathomably stupid)

In my not-so-humble opinion.

If the sampling rate is high enough (which means OS, probably a lot) that the ultrasonics are far beyond what even poses a risk of making the tweeters vibrate at all, and all amplification components will tolerate them without ill effect, and you have cabling and connectors that are standard for HF stuff not audio stuff so that you mitigate the reflections and what not, then... and only then, I think... maybe removing the LPF makes sense. All that together should, I think, result in a system with way-better-than-normal transient response (if you have an extremely good upsampling filter). Worth trying, at that point (I mean if you do have such equipment for other reasons, buying it to test this is definitely not worth it).

But for NOS? Nope. Ultrasonics anywhere close to the audible range are a bad thing, I'm quite sure of that.
Yeah, I'm with you. I'd rather always have an LPF in a DAC. Just sayin, it's not always there and in the case of Schiit DACs in NOS mode, even when upsampling, you can still get stairsteps as I understand it, bc the LPF is optimized for 8x but the highest you can effectively feed a Schiit DAC is 4x.
 
Mar 27, 2023 at 4:48 PM Post #114,615 of 155,168
Methinks you should revisit Jason & my replies. If you don't have a brickwall you may get stairsteps. Plus the link I posted clearly shows your statement is false. NOS DACs do exist that don't have an LPF.
I disagree. And not because I ‘think’ so. I am very familiar with the technology, theory and practice.
 
Last edited:

Users who are viewing this thread

Back
Top