The question is simple🙂! Which do you prefer and feel free to make comments on why. Might also be useful to say what DACs you base your opinions on.
Fran
Fran
It often depends on the material... nice to have a switch (so Fran, i added another option, hope you don't mind)
dave
dave
Not at all Dave, but from the poll so far we might not have needed it! Interesting results so far hut let's have s few more explanations.....
Fran
Fran
I think NOS has more appeal, and is more prevalent in DIY form, because it is accessible to most. Technically, its easier to do. The involved frequencies involved make implementation simpler, and less sensitive to jitter.
You can knock up a TDA1543 DAC on strip board, and plumb it into a CDP with an I2S signal in it quite easily, and be pleased with the result. Proud in your achievement.
You loose the high end quite a bit, so the sound can sound as if there is quite a pronounced bass, in strange wooly What Hifi terminology - 'warm, laid back, etc'.
I think those are the main reasons for why it is liked.
I started with NOS, and then moved on to 4xOS. The first thing I thought was 'where has all the bass gone', but after acclimatising, I noticed so much more detail, I've never looked back.
Anyway, there's loads on this forum about this stuff.
Phil
You can knock up a TDA1543 DAC on strip board, and plumb it into a CDP with an I2S signal in it quite easily, and be pleased with the result. Proud in your achievement.
You loose the high end quite a bit, so the sound can sound as if there is quite a pronounced bass, in strange wooly What Hifi terminology - 'warm, laid back, etc'.
I think those are the main reasons for why it is liked.
I started with NOS, and then moved on to 4xOS. The first thing I thought was 'where has all the bass gone', but after acclimatising, I noticed so much more detail, I've never looked back.
Anyway, there's loads on this forum about this stuff.
Phil
If someone does not compensate the hf rolloff and does not use anti alias filter, what will he expect from something that is just technically totally wrong ?
You would not rectify ac, supply it to your amp without caps and then complain ? 😉
Yeah, but, if you do it properly, then surely the huge nth order brickwall filter would scupper the sound completely? Its certainly not ideal. As an alternative, I think oversampling is a better compromise, and easier to implement well. I get the impression that most NOS out there is not correctly filtered (e.g. those 4xOS CD players out there with their digital filters bypassed).
no 🙂
Anybody who wants to oppose to that, please post the schematic of the applied filter.
yes 🙂
Even more all those NOS DACs.
Hello
There is a third one, the shift-registers dac, kind of analog Over-Sampling, there is one from ecdesigns and Cambridge audio use it in the 90's and there was another cd player brand but I've forgot it.
NOS are best with small musical groups, like jazz and chamber music, but with big orchestral music you lose some precision. This loss of precision are more because of the HF noise causing intermodulations in the I/V amp, the preamp or the amp, or you would need a analog brick-wall 20khz low-pass filter with lot of phase shifts. This loss of precision are much less because of the HF rolloff.
OS still quite much sensitive to jitter, a good synchronous reclocker can do the job by reducing jitter and maby a better OS filter than the SAA7220.
Bye
Gaetan
There is a third one, the shift-registers dac, kind of analog Over-Sampling, there is one from ecdesigns and Cambridge audio use it in the 90's and there was another cd player brand but I've forgot it.
NOS are best with small musical groups, like jazz and chamber music, but with big orchestral music you lose some precision. This loss of precision are more because of the HF noise causing intermodulations in the I/V amp, the preamp or the amp, or you would need a analog brick-wall 20khz low-pass filter with lot of phase shifts. This loss of precision are much less because of the HF rolloff.
OS still quite much sensitive to jitter, a good synchronous reclocker can do the job by reducing jitter and maby a better OS filter than the SAA7220.
Bye
Gaetan
Last edited:
I voted both, because I prefer R2R dacs w/o analog filter. I do not mind much of the digital filters, if done well.
High-end NOS DAC
Most of listening comparisons were made with low cost NOS DAC and high-end OS DAC.
The digital filter is only a part of DAC; everything is important digital receiver, digital filter, analog DAC chip, power supply...
After listening and comparing various high-end OS DAC, I implement my DIY high-end NOS DAC. see pictures of my Jundac One
Most of listening comparisons were made with low cost NOS DAC and high-end OS DAC.
The digital filter is only a part of DAC; everything is important digital receiver, digital filter, analog DAC chip, power supply...
After listening and comparing various high-end OS DAC, I implement my DIY high-end NOS DAC. see pictures of my Jundac One
Last edited:
Linear interpolator
Cambridge CD2 16bit x 16 ovsl. It performs a linear interpolation, fortunately, after a standard 4x oversampling.
I like PCM1794. Played with a NOS 1543 a few years back and generally hated it for anything but the simplest music. Piercing mids and no highs.
Oh man thats harsh!!! (pardon the pun🙂🙂)
Directly the opposite to my experience with 1543 though....
Please make more comments about the DACs you are basing your opinions on (or at least some details if you don't want to mention a make or model)
Fran
I'm basing my findings on NOS TDA1541A (admittedly not properly brickwall filtered, just 3rd order), then adding SAA7220P/A+B (tried both chips), then on to PMD100. Each step improved the quality IMO.
I think NOS sucks. I tried it so many times with TDA1543 and 1541 and others. it kind of remind me the situation with true hi-end audio vice DIY. Most of DIY really sucks ,too. It's all becouse most of DIY think they can build hi-end but they can't due to lack of years of expirence and money for research and prototypes. Yes maybe it is funky, groove and cool this NOS idea but for me it sucks and I can hear it right away. Gone is clearance, details, highs, some mess is all the time there, ok maybe it's tiny bit more natural in the midrange...so what when all the rest sucks. too much waisted time with it.
Bartek
Bartek
Last edited:
I think you heard the wrong NOS dacs 😉
But I agree that DIY a state of the art converter is VERY hard and that most project make too many compromises, as much or more as mid-level commercial units.
When comparing OS and NOS DACs, the only correct way is to use the same DAC: first in OS mode, then in NOS mode. Comparing two different products or even different units is misleading. The easiest comparison can be done by using the old Philips players with SAA7210, SAA7220 and TDA1541(A). The SAA7220 is responsible for oversampling, and it is relatively easy to bypass (even switchable in/out).
BTW, what about the -3 dB rolloff of the NOS DAC, is it real or hypothetical? I mean I can generate a full-scale 22.05 kHz square wave test signal on CD, and its fundamental harmonic is a 22.05 kHz full-scale sine. Where is the -3 dB coming from then?
BTW, what about the -3 dB rolloff of the NOS DAC, is it real or hypothetical? I mean I can generate a full-scale 22.05 kHz square wave test signal on CD, and its fundamental harmonic is a 22.05 kHz full-scale sine. Where is the -3 dB coming from then?
- Status
- Not open for further replies.