budget DAC to compliment the JDS Labs Atom ?

[Raimei Templar]

Not necessarily the board's fault. Finding a case where the front panel does not introduce any ground loops does not seem to be trivial. I've had good luck with some C2D era Fujitsu-Siemens Celsius machines, so maybe those guys care for their audio at least. Other prebuilts, not so much. (And it's not just prebuilts. I've seen complaints about rather expensive cases, too, like some Lian Lis if memory serves.) This is what I just found on a ca. 2006 Asus midtower:
Audio ground from HDA cable connects to one patch of ground, which then goes to chassis ground via the LHS retainer (screw - bracket - screw).
Headphone and mic jack ground joins USB and misc switch / LED ground, going to chassis ground via the RHS retainer.
In practice, this means there is a ground loop running over the front panel audio cable, and worse yet, headphone ground return is shared with USB and whatnot, with some more ground loop action.

Boards by themselves can be messed up, too, of course - subpar grounding schemes, sharing 3.3 V between DVDD and AVDD, less than ideally clean supplies...

Implementation is key with these things.

Now I don't know about modern chips, but one notoriously weak spot of Realtek codecs has always been the mic input. More or less noisy mic bias, usually covering up that the amplifiers themselves are far from state of the art in input noise either. (I know that these chips are supposed to eliminate external components and conserve board space as far as possible, but one capacitor would go a long way in solving this problem.) Periodic ripple in the ADC digital filter also tends to be at rather uninspiring levels in all but a few chips. Basically, everything that doesn't reflect in big numbers in the spec sheet is rather neglected. Recording in multiples of 44.1 kHz does not seem to be a good idea either, it seems to result in massive amounts of jitter with substantially degraded SNR and aliasing. The playback side is far more robust.

In any case, providing a decent line-level output for a headphone amp or similar is the area where I'd see the best chances of onboard audio working well. On the recording side, make sure you stick with multiples of 48 kHz (don't forget to set up sound devices properly).

I would even say most. Typical output ranges from 1.0-1.2 Vrms on 5 V AVDD and about 0.8 Vrms on 3.3 V AVDD.

Oh. What sort of issues have you had with them? The most serious problem I ever found was sample rate being set up all wrong when selecting 44.1 kHz in driver versions from about 2011, which was subsequently fixed. (Generic implementation also seem to be limited to 16 bit samples on the recording side even if the ADCs themselves would definitely support 24 bit, which is silly.) Mind you, I have little experience with their latest chips and Microsoft's bug collection version 10... there have been a few changes and issues with sound drivers in general since 7.

I've seen a lot of sound driver troubles over the years, and honestly, the Realtek drivers would seem to be working pretty well to me. At least there are no weird suspend / hibernation issues, that sort of stuff is a total dealbreaker for me. I still have an ESI Juli@ floating around - a no doubt good and unique card that just stops working after hibernation. And don't even get me started on the hoops I had to jump through to get my Asus Xonars to produce the sample rates I wanted, or theiir other driver quirks. (Hint - for Xonar D1 & D2, best use driver builds 1800 or older.)

My most recent (very fun issue) was that Windows decided on its own to update my realtek drivers to a version that my motherboard doesnt actually support, so when my computer rebooted after updating I was greeted to the loudest static explosion noise I have ever heard in my life. It scared the absolute shit out of me and actually caused physical pain to my ears, I thought my headphones were destroyed for sure though luckily they survived. Anytime additional sounds tried to play I would get a static popping noise rather than sound.I had to uninstall the drivers and go manually get the ones from the Asrock website to fix it. Then it was such a nice experience that windows kindly did the same thing again the next day. I dont know if its Microsoft or Realtek who screwed up on marking my motherboard as using that version but needless to say I was less than pleased.

I have also had trouble with the virtual sound profiles randomly loading and making everything sound like crap on multiple machines.

My most consistently annoying experience was with a laptop that had a shared mic in and line in port. When you plugged in a Mic the realtek control panel was supposed to pop up and allow you to select if you were using a mic or line in but it almost never did. Even going in manually to change it, I would frequently switch back to line in mode and the mic would not work properly. It did the same with the headphone port too, wanting you to tell it if you used headphones or speakers and it refused to remember your selection. So I spent way too much time in that control panel as a result.

I will say that the optical out usually works without issue and my troubles have been confined to actually using the speaker or headphone outs on a motherboard, so connecting a external dac via optical out is generally a trouble free experience.

I will also give credit to Realtek for having better drivers than Creative, who consistently has the absolute worse drivers I have ever encountered.