NOS vs OS: An Audiophile Culture War?

Depends on what flavor you are after.....which neither is truly accurate to real life. Some like it dry, some like it wet, its just flavoring with no objectivity actually. There is no golden standard.

 

NOS or OS is a preference. It is no secret the difference in sound as highlighted by @Magnetostatic_Tubephile. It is just there are now more devices that can do this with a flip of a switch ad @k4rstar pointed out. To do NOS on many DACs from the past required DIY modifications.

I don't know either why why upsampling/oversampling has to be done in a separate box with its own set of issues. Any attempts to explain this objectively is more or less what you have to tell yourself.

It is also no secret the sound or "magic" from vintage tube amps...unregulated power supplies, tube rectification, saturated output transformers, inpur power chokes, big Sprague caps, etc. We are talking about technology from the 1950's. As far as the sound, rounded off square waves and even harmonics.

@k4rstar, you seem to have a preference, and that is okay. We all have our preferences. Some just like a sharper sound, some like a more organic sound. Reel to reel tape also strikes this balance, although leans more organic because tape heads and tape saturation with more of those lovely even harmonics. I think also there is no one here rushing for material or scientific explanations. You know that SBAF is not measurebators. Yet we are not purely subjective either. I think it is okay to strike a balance, and that is what this community is all about.

Back to NOS - We can see effects of the signal with and without, and we can hear it. I think I read somewhere, (maybe here on @Hands NOS thread) that as you increase oversampling, the sound gets sharper and more glary. 2x, 4x, 16x, etc.

I keep saying this, but when AKM founded his company, his benchmark goal for an DS chip was the TDA1541. And much of what DS has to overcome is this 128x or even higher oversampling artifacts. I am not in love with the "Velvet Sound", but I think they were more on track with the 4396 and even 4399, but I digress.

I will end with I did do the Pepsi Challenge with OS and NOS with the RU6. Changes were as expected. However, the was something odd in the EQ I found with the OS mode. Akin to comb filtering or something. It could be just a poorly designed and implemented filter, or just OS vs NOS. Either way NOS was my preference.

 

My thread was never to espouse my own preferences. It was about encouraging others to discover theirs, by rejecting mainstream notions and dogmas. You just cited some of them, by the way

I think I was at least partially successful. The topic has 60k views and I have dozens of PMs across several forums from people thanking me after deciding to try different things they previously hadn’t considered. It makes me feel warm and fuzzy inside

 

R2R vs OS will be a debate forever. Just like Tubes vs SS, Horns vs Tweeters, Analog vs Digital, Room treatment vs Plants, Dsp vs analog eq, analog crossover vs digital, ect

I use both. Both can be good. Thats the whole debate.

the “wars” are being driven by those who have $$$$$ incentives to do so

 

Inspired by the wonderful comments here, decided to try HQplayer oversampling into the NOS DAC I'm using (192khz) and I do like the sound. Seems a little more 'lively' and 'filled out'. Once I get the Holo Spring 3 KTE loaner I'll be able to play around a little more because I think it can accept higher rates.

 

But wouldn't the LPF filter sill be there? You would still need a way to bypass the DSP or analog filter. I could be wrong.

Either way, I believe there are file converters that will alter a file and up-sample without interpolation. And all the Schiit MB bypass the filter at 192.

 

Exactly. It comes down to the effort and intention of the designer and how much effort they put into voicing the dac to make it sound the way they wanted. The dac chip or other components matter, but perhaps not as much as the execution. I had a Chi-Fi Sabre dac that sounded more like a modern Metrum than the Gungnir that followed it. Even "Metrum sound" isn't very specific because Cees has tuned a range of sound character into his dac efforts.

As others have said, the dac is just one component in the system and the whole system combines into the sound. When you combine components into a system, you may think you are choosing them to be the truest to the original, but in the end you are seasoning the broth to your taste.

 

Possibly. I've only used OS DACs until I got the Paradisea+ but I'll know more when I get the Spring 3 and Yggdrasil LIM.

 

BTW, isnt Yggdrasil and the other schiit multibit dacs technically NOS? Since they preserve the original samples? Like a NOS/OS hybrid?

 

From Mike Moffat:
"It is a digital filter/sample rate converter designed to convert all audio to 352.8 or 384KHz sample rates so that it may drive our DACs. You get it uniquely from us; it is our filter. It took five people many years to design and perfect at the dawn of digital playback, way back in the early eighties. It keeps all original samples; those samples contain frequency and phase information which can be optimized not only in the time domain but in the frequency domain. We do precisely this; the mechanic is we add 7 new optimized samples between the original ones. All digital filters multiply the original audio signal by a series of coefficients which are calculated by a digital filter generator. Over the years, before Theta Digital was born (my original company), we developed this filter design/generator. The common digital filter method is a Parks-McClellan algorithm, which has been used in all of the older oversampling chipsets, and persists to this day as the input filter in most Delta-Sigma DACs. Why? I assume it is because it is royalty-free, and the algorithm is widely available as are digital filter software design packages to aid in a cookbook approach to the design. Now Parks McClellan an open form math solution, which means that the coefficient calculation is a series of approximations which always get halfway there. This of course, means it never completely solves. The worse news is that all original sample are lost, replaced by 8 new approximated ones. Further, the Parks McClellan optimization is based on the frequency domain only – flat frequency response, with the time (read spatial) domain ignored. Our filter is based upon closed form math – the coefficients are not approximations, the equations solve; the matrices invert and the math is done. The filter also optimizes the time domain."

 

I find it a bit frustrating as I prefer the sound of OS subjectively but fatigue quickly from every OS DAC I've tried so far. It becomes a point on if I want to hear my music the way I prefer it to sound(airy & defined) vs listening comfort.

More annoying is that I don't have any solid reasoning as to why I find NOS less fatiguing, for every argument discussing the perfect impulse response there's a solid counter-argument that NOS creates phase shift or timing confusion with aliased images. I've owned NOS DACs which sound both lightning fast and syrupy slow which share the fatigue advantage, I have no clue what subjective trait I'd assign to the fatigue factor if any.

All in all I've been exploring modern SD DACs for the past few years to no real avail. The Gustard X16 I'm using now matches what I thought I wanted near perfectly but I don't find myself enjoying music all that often with it. I expect to suck up my 'preferences' & return to NOS this year. Or in other terms not arguing with what works & brings me joy.

 

Funny, the very reason I don't get along with NOS DACs without HQPlayer upsampling is that they give me headaches. The May KTE sounds great in NOS right up until I get a headache. If I use HQPlayer to upsample with a linear phase filter, the May KTE sounds better AND I don't get a headache....even after 6 hours of listening.

 

I also struggled with listening fatigue for some time with OS DACs and always returned to NOS. Until I tried the Chord Qutest and I totally agree with GoldenSound's assessment, that the Qutest (and probably other Chord DACs as well) is exceptionally non fatiguing.

 

Can someone explain how it is that Park McClellan is said to "ignore" time domain? Given that PCM is two 'things', a measure of loudness combined with a "time domain", that is the sample rate (let's stick with 44.1hz, so 44,100 samples per second). These two 'things' are combined to reconstruct the one waveform (to rule them all) that is sound which is nothing more or less than a pressure wave. Since these two things are "digital", measures represented by numbers, by applying "over-sampling" you are by definition applying math, algorithmically "interpreting" what should go in-between the given samples, the relationship between the samples, and if/when the original samples will be preserved "exactly" - exact preservation being "closed form" if I'm not mistaken.

So what does Mike mean when he says that non-closed form over-sampling by definition "ignores" the time domain? Does he mean that because that because the math is open and the OG sample is not preserved accurately (though it is precise right!?), that the resultant wave form is not accurate to the OG...but then it has to be said that the resultant waveform is both time and frequency domain inaccurate (though again, very very precise given that we recognize the result - the Rolling Stones are still clearly themselves) and the two always go together.

OR, is there something specific about Park McClellan that alters phase, similar to how so called "min-phase" filters change the relationship of timing of higher frequency's relative to lower frequencies?

 

Switching between OS and NOS gives me a headache every time. It feels like getting a new eyeglasses prescription.

 

thats weird cause its the exact same experience. I find OS have tighter bass, clearer soundstage, more bite, faster, but eventually, I get worn out. I just cant relax and yeah...
I get the same experience when I go from vinyl to digital though