oversample or NOS DAC ??

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Jump to Latest
Bernhard

Bernhard

Member
Joined 2002
Some more thoughts.

Most speakers use frequency dividing networks. Not as steep, but right in the middle of the audio band.

Sony's most sophisticated players have switchable digital filters. Why ???
Because there is no digital filter that sounds "right".


Besides all technical trade offs, my personal preference is non os. :)
 
S

SoNic_real_one

Member
Joined 2008
Key said:
Ah but what if I go up to 28,224,000 before going back down to 192,000?
Click to expand...

Still no go... Artifacts are still going to be present if you do not use specialized circuits like Asynchronous Sample Rate Converters. Such circuits use an internal 28 bit data path and advanced digital filters.

Bernhard said:
And the linearity of a 20 bit multibit DAC isn't even 16 bit.
Which means that if the noise floor is very low, the distortion becomes even more audible.
Click to expand...

Haha, it IS better that that and still much better that a 16 bit DAC. BTW, you know that even an "R-2R" DAC has only a limited ladder inside.
The real "option" on hi-res are the modern 24 bit delta-sigma switched capacitor DAC's.
 
V

Vadim

Member
Joined 2003
Bernhard said:


I have owned a few DCD-2560 and was not impressed.
Click to expand...
Bernhard,

I am not sure what is there not to like about AD1862. Perhaps you might want to say what exactly you did not like.

For my money, AD1862 was pretty good. I did replace the N-grade chips with K-grade and then used the adjust pin to fix the linearity as good as I could get it, I know that those K-grade chips were doing nearly 19 bits. Try to get that from modern 24-bit chips! Considering most modern Delta-Sigma D/A, you can’t see the last 5-6 bits on any scope, - noise city!

By the way, the AD1862 was a true ladder type D/A unlike what the Burr Brown used to make, so comparison is not all that fair here. The AD1862 was laser trimmed to achieve excellent linearity. In any event, I see nothing wrong with it. I use AKM4396 D/A these days, which is one of the better 24-bit chips on the market and still it is impossible to tell it apart from AD1862K. Presently I have my eyes set on Sabre product, - will see.


Bernhard said:
Most speakers use frequency dividing networks. Not as steep, but right in the middle of the audio band.

Sony's most sophisticated players have switchable digital filters. Why ???
Because there is no digital filter that sounds "right".
Click to expand...
Your comment regarding passive cross-over is well taken and I agree they are a major problem far more problematic then any DAC may bring to the table. This is why I do not use them, - I only have actively crossed speakers and I use digital cross-over only.

Finally, why are you saying that “…there is no digital filter that sounds "right….”? I do not get it, what is wrong with digital filters

Vadim
 
V

Vadim

Member
Joined 2003
SoNic_real_one said:


Come on, that's your scope fault, do not blame the DAC's :)
You can't see the last 3-4, I agree with that (120-123dB). Depends of the I/V converter thou...
Click to expand...

I say 5-6, you say 3-4...well, maybe your Faraday box is better then mine, it does not matter. What matters is that even 16-bits resolution is not that trivial to get out of your CD. And to get that one would need all the tricks, - the oversampling, the power regulation, layout, parts, etc.
 
D

dzzmiller

Member
Joined 2008
I suggest whatever you do, look at pro sound equipment for your DAC needs. Audiophile DACs have large markups and a lot of marketing who haa.

Two choices of pro dacs, those used in music production, and those used in performance. The performance DACs are speaker management systems that also have other functions, such as active crossover and EQ.

At any price point you will get much more value with pro equipment. Also more fun stuff.
 
K

Key

Member
Joined 2009
okay tried to compare algorithms. And I can actually hear a difference in sin(x)/x from linear and lagrange.

I still have no idea which is best but I hear a difference. Sin sounds to me like it has a slight distortion on high frequency attacks. Of course with this particular sound that I can hear it on I don't know if it's supposed to be that way or not. But I do hear a slight difference.

At first on program material it seemed like sin(x)/x was giving me a slightly extended tail on reverbs but I am skeptical if it was a placebo effect or not - it was very subtle.

I know I am going to get crap for this signal chain though. In soundforge I took an EAC secure rip (44.1 16-bit)>64-bit Float 44.1kHz>176.4(linear, lagrange,sin (x)/(x)>played back in soundforge at 176.4kHz 64-bit float (file and DSP)>176.4 24-bit playback ASIO(I assume truncated without dither)> Mackie Onyx 400f> Neutrik 1/4" balanced to XLR> Wharfedale Diamond 8.2a Pro
 
S

SoNic_real_one

Member
Joined 2008
Key said:
okay tried to compare algorithms. And I can actually hear a difference in sin(x)/x from linear and lagrange.
Click to expand...

I was reading this and I was thinking "dude, you are imagining things"... then I read the signal chain you have and I changed my mind. You might did hear something with that, I cannot say that with my system :)

Anyway, a hardware DAC will have only linear interpolation.
 
T

Telstar

Member
Joined 2007
SoNic_real_one said:
Well... maybe you are right here. I didn't like the sound of the "brickwall" - maybe is not the phase shift that bothers me most but the incomplete rejection of the image. The best analog filters can do little to nothing for that. I don't know exactly, but also the numbers are not in favor of NOS. And it's not only me, all the audio industry people think the same since they make only OS DAC's now. They must be hearing something too.
Click to expand...

The industry mostly like numbers. And the measurements are way in favour of OS.
Audiophiles like me instead uses their hears.

All this big discussion about OS/NOS is much, much less important than the quality of the converter (and its sourrounding parts).

I have recently auditioned an excellent pcm1704-based DAC, with switchable 8x OS and analog filter.
Surprisingly, I hated the analog filter, while the 8x oversampling did just a little change (for the worse imo) in the sound.

It's a hot issue but for me it's a moot point :)
 
T

Telstar

Member
Joined 2007
Vadim said:

I use AKM4396 D/A these days, which is one of the better 24-bit chips on the market and still it is impossible to tell it apart from AD1862K. Presently I have my eyes set on Sabre product, - will see.
Click to expand...

Yes, the AK4936 is one of the few delta-sigma dacs that sound reasonably well.
When you get the sabre, go for the 9018 (the 32 bit version), it has a lot of needed improvements.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Top Bottom