[Piffles]

Hi,

Here's a question that's been bugging me for a while: you can get DACs for just about any price, from 50€ to 5000€. What is the difference between a good DAC and a not so good one?

For instance, I look at the expensive ones, e.g. Musical Fidelity X-DAC V8 which upsamples everything to 24bit/96kHz before converting to analog. I don't get how this can be a good thing. If you're playing primarily 16bit/44.1kHz stuff, what's the point in upsampling? Does it somehow reduce noise to have more samples in the signal before converting, even if they're just extrapolated samples?

But then, there are also cheap ones that do 24bit/96kHz like my M-Audio Fast Track Pro. I can set the max bit resolution and max sample rate in windows. So, if I set it to 24bit/96kHz will it upsample all my audio? And if yes, is that a good thing? Or will this high setting mean that 16bit/44.1kHz stuff will not be upsampled but allow 24bit.96kHz audio not to be downsampled?

Now, what happens once the signal is converted and I have my analog audio? All it has to do is route it to the output, with the lowest possible impedance I suppose. Seems pretty simple to me and yet the Musical Fidelity boast a switchable tube OR solid state output. I thought tubes were used only in amplifiers, pre and/or power. There's no amplification going on in DACs, is there?

The reason I ask all these questions is that I just finished putting together my hifi set. I put 1300€ in the amp and 2300€ in the speakers but only a mere 230€ in the DAC/soundcard. If I compare an Lame Standard MP3 playing through my soundcard to an original audio CD playing in my dad's Marantz CD67 SE, I can't tell any difference at all. I can however hear the difference between an 800€ amplifier and a 1300€ amplifier. Same for the speakers. That's why I put so much money into those elements of the chain. But I'm surprise that so cheap sources do the job for me. Am I missing something?

Thanks for reading. I'd really appreciate some info from people who have experience with several different DACs (CD players, soundcards, external DACs, DAC amp combos etc.)

See you,
Piffles.

 

[cerbie]

PCM data (everything but older SACD) says, "output at this level at this point in time." Next point in time, it may be at another level. If these changed instantly (zero order hold, IIRC), it would sound bad, being all stair-steppy. The DAC takes those points, figured out about how it's changing, based on previous samples, and makes the change fit that curve.

That last bit happens on the analog side. You've got four basic things to consider:
1. DAC
2. Power
3. Layout/quality
4. Jitter

The DAC chip itself makes the basic analog signal. It's going to matter. Not much more to say.

You need clean power. Power that has a bunch of noise will raise the noise floor, and possibly mess with the amp portion of the DAC (which most will have, be it in the IC or another chip).

The circuit board quality, its layout, and quality parts backing the chips up will give it a cleaner signal, and better power. Matching them up synergistically can be better than just going with the best separate parts.

Jitter may or may not be audible. Let Big Shot and everybody else worry about that part. But, it is there. The DAC operates from a clock. It's nothing more than a signal that rises and falls in voltage. But, noise from all around can get in there, and make not a perfect wave, o the DAC may see it reaching high or going low at times when it isn't quite there yet.

Basically, jitter is a huge digital audio issue, but many thousands of engineer-hours have gone into dealing with it, so now that there's good proof it was totally audible way back when, it may or may not be in good modern equipment.

Upsampling is due to clocking. You've got things to play in 44.1kHz, and then 48kHz multiples. It's hard to get a good clock signal that can match both. But, due to possible jitter, you really don't want to pass the clock straight through from outside (SPDIF, FI). So, you asynchronously upsample. 44.1khZ to 96kHz may be more like 44.092kHz to 96.038kHz. It's not dead on. But, they've got asynchronous resampling to such high quality now, that most of us either can't hear the difference, or prefer the mild dynamic range loss to more annoying jitter artifacts (again, assuming you hear jitter). Jitter still gets mapped over to the new signal, but not in the same way. In reclocking, the DAC can also get an average of the incoming jittered clock, so it's even less bad. The DAC will be tied to a super-clean 96kHz clock, so is analog jitter is not audible even based on the jitter believer limits. You can find a thread about all this over at diyaudio, but I think you need to be a math person to read it all the way through (I did

).

But, then, you've got ASRC (asynchronous sample rate conversion) going on, and whether that's better than jitter or making a super-clean source (Pace-car, FI), is a huge debate that my Kosses and I are going to sit out on, but with some popcorn (made on the stove in a steel pot, no microwave crap). Mmmm, budget-fi.

All that said about resampling, doing so in hardware is a totally different animal than software. In Windows, you're best using a good player, like FB2K, and having it resample for you, or at least check to see what your specific device will do to incoming audio (and if it bypasses kmixer). If you can just pass 44.1, try it that way, and resampled in software, and see if you can hear a difference.

A DAC will have an amplifier (but not what we generally consider a power amplifier) in it in like 99% of all cases. However, having a tube amp on it seems like a silly gimmick, even if you like tubes, and is probably not offering the best output.

What's best for you, nobody can answer but you. I found an amp for my cans to be fantastic, and much better than getting better cans (my whole system, total, costed about as much as your DAC, though, too

). All the extra resolving detail wasn't the difference to me that the tightening up and smoothing out of a decent amp was. You can get good tips from others, but not direct answers as to how best put your wallet's extra thickness to use for your own enjoyment of recorded music.

So, you may be missing something, but don't sweat it. Your brain hasn't learned to pick out the differences, or something with your system--who knows. You know there are people who can't tell apart their MP3s with big bad systems, too (and likely many who can't but won't admit it). *shrug* Your ears and/or brain aren't exactly operating like anybody else's. Your not hearing iit doesn't mean it's not there, but also doesn't mean you're defective for not hearing it.

 

[tot]

The analog signal coming from the DAC has to be low-pass filtered below the Nyqvist frequency (half of the sampling frequency). Since perfect brick-wall filters are impractical because they take infinite amount to operate, you can not filter perfectly and more or less phase distortion is present on audible range.

If you upsample let's say 44.1kHz into 96kHz, the analog filtering is pushed into higher, inaudible frequencies.