The last time I had a problem with my DAC, it was because my crappy wifi connection to the source device kept stuttering without enough throughput and created a low buffer cutout problem that the device wasn't really designed to deal with, and the wifi source wasn't compensating for well enough.
Pause>>DAC Off>>Source Restart>>DAC On>>Play
Pause>>DAC Off>>Source Restart>>DAC On>>Play
No-contest totally better sound: great extended range, superb dynamics, very detailed mid-range and high end. It makes you hear sounds that weren't noticed before as well as performances for example in very fast, rapid succession transients with percussions (lower volume in the background: Sade Live - Smooth Operator) and guitars (louder volume, at the forefront: Masters of Acoustic Guitar - Track 1).
With live music recorded with vocals or not, friends and ourselves often say the musicians are in the room with us. It still often happens we get startled by some sounds and people, thinking momentarily something happened or someone else is in the apartment (despite knowing it can't be).
It has all the analogue you'd expect from vinyl, but has none of the issues:
No pitch distortion (my Technics has very little, it's a direct drive), no clicks/static and pops at all.
Nothing gets rid of the clicks and pops in vinyl.
With DS128 and above, I am happily re-discovering all my collection and that of my significant other as well.
Despite the DR on vinyl often being higher than digital after it was compressed to death in mastering? sad but true.
Yes, that's a big issue with some Redbook CDs. There are some nifty utilities on the web where you can look at databases of CDs and their measured Dynamic Range. That can help avoid some known bad releases. For instance, I know some people who have both the CD and the vinyl of Depeche Mode's "Playing the Angel", say that the vinyl sounds way better. I would probably find the same result here in my rigs, as I hate the lack of dynamics in some overly compressed music.
Another case in point: Kavinsky's album... Everything is so *loud*.
Thus, in the DM example above, I'd certainly prefer the vinyl version with the more proper DR, despite the clicks and pops absent from the digital version.
On the other hand, I don't think compressed DR is the issue with proper, native DSD releases that you can get from Blue Coast Records, nativedsd.com (check the JustListen section for free files) and Superhirez.com.
These guys come from the angle of a very picky audiophile background for their customers, where people do their best to get as close as possible to the live experience of a full band (often a full orchestra).
So that's a very good test to do: compare original DSD128 and above (you can sub-divide this as native DSD recording or DSD derived from the Master Tapes) versus a good vinyl release.
The Yarlung Quad-rate album I have - Smoke & Mirrors - Varnish - is stellar. I never heard vinyl sound that good, and I don't expect it to.
You could also do a comparison with DXD native recordings from 2L.no - there are a few sample files there as well.
This said, I love the fact I can get my hands on very, very rare records for small prices on vinyl.
Last edited:
That's Manna!:
Do a DIY PS Audio Directstream, in the sense that your Redbook is up-converted to DSD128 and above and plug that into a native DSD DAC like the iFi iDSD Nano.
I up-convert offline (takes a lot of space!), but you could also use HQ Player and up-convert in real-time using its wonderful filters and algorithms. (Never, ever use JRiver to do that).
Another thing you could try for free is the same but with Foobar2000 on Win.
Or, on Win, you could also try BugHead Green on WAVs or BugHead Emperor/Infinity Blade. The quality with just normal Redbook WAVs is astonishing.
PeterSt's XXHighEnd can be a worthwhile try (never tested this myself).
Why the excitement about DSD? 24/96 is on paper better. Currently I feel no need to upgrade my DAC to play a lower resolution format with a bigger marketing budget.
My DAC (like most) oversamples and yet to find a coherent reason why upsampling external to the DAC should perform any useful function. Is there something the upconversion algorithms do that is superior?
My DAC (like most) oversamples and yet to find a coherent reason why upsampling external to the DAC should perform any useful function. Is there something the upconversion algorithms do that is superior?
All examples of how digital should come across - but often doesn't. The dull, disposable quality that one often hears is a distortion, but it's hard to put one's finger on what's actually not quite right, unless one goes through a few iterations of hearing correct, vs. incorrect sound.
An easy, in your face, way of hearing the unpleasantness of off-kilter digital sourced sound is to put on a highly energetic, heavily layered mix of pop sound with lots of reverb thrown in, and wind up the volume a bit. If this is an impossible, highly disturbing mess to listen to, then the problems of the playback system are screaming at you ...
I also have all of the same questions and after a lot of reading I still have not found a satisfactory answer.
I am convinced that it is BS. Upsampling is trying to get something from nothing, and typically results in a reduction in quality unless lots and lots of CPU cycles are involved. The best case scenario with upsampling results in identical sound, so it is a wasted effort.
I see converting from PCM to DSD as being equally futile, and I have never seen a logical explanation that proves otherwise.
Last edited:
Conversions may result in better subjective sound, because the new format 'suits' the replay mechanism better - it distorts less objectionably while processing the data. But, it's a poor way of solving the problem, for all the obvious reasons - a system should be able to transparently handle any format, all versions should sound identical.
So you are saying, for example, that a particular DAC chip may perform so much better with 96/24 streams than it does with 44/16 streams, that it may be beneficial to utilize the CPU and software in the playback device to upsample to 96/24. I can see this being true in very few cases (if at all) and even so the software upsampling would need to be truly excellent and the CPU power proportionally adequate.
The CPU and software is used prior to, offline, from the playing, outputting a new file on disk - it can do a "perfect" job, say, using current versions of Audacity. A desktop I played with quite a bit in this regard benefited from this approach. But this is obviously a right pain to use, it's a dead end for use in an ongoing fashion.
Unless one has tangible evidence that their DAC chip performs measurably better at a specific sample rate and bit depth, upsampling (or downsampling) one's entire library is a ridiculous act of futility.
However, there is no end to audiofoolery and HDTracks.com has built a huge business on doing precisely what I have just described.
DSD is the next fad. I'll give it a miss, like all the other snake oil, until proven otherwise.
However, there is no end to audiofoolery and HDTracks.com has built a huge business on doing precisely what I have just described.
DSD is the next fad. I'll give it a miss, like all the other snake oil, until proven otherwise.
Various multibit DAC chips ( PCM1704 to name one that I examined) perform worse at higher sample rates, based on the specs in their datasheets. Yet in commercial DACs its fashionable to run these devices as fast as their specs allow.
The most important factor I've found with getting good sound is assuming nothing. That is, every single part of a replay chain is possibly faulty, or has a weakness, no matter how brilliantly engineered or impregnable to problems it supposedly is, certified as such by all the "best authorities", 😛. Doing that, I quickly get to the heart of the problem, and progress is made ... 🙂.
I don't listen to music by looking at a piece of paper. On paper, with mp3 compression you were told you wouldn't hear anything lost either.
Yes, it could be the algorithm is better (HQ Player has great sounding algorithm and filters).
Secondly, there can be an advantage to have the least processing possible near the DAC, letting the DAC do only what it should do, that is the D/A without being perturbed by additional noise induced by any processing near the DAC chip.
Another thing you can find is that you can then possibly disable the filtering at the DAC or use a gentler one. That can help as well, especially with phase/transient issues.
That said, this is all very system-dependent: you could certainly have a PCM/DSD DAC doing better processing that what your ears are telling you with Audirvana+ as an example.
This is why once you have a versatile DAC (like mine), one of the first things to do is to work on comparisons with all the possibilities you can get from your audiophile player and the DAC combination, including the different ways of ding the over-sampling or up-conversion and the different ways of doing the filtering.
There usually is a 'sweet-spot' where the DAC shines best. It so happens that I love how my DAC sounds with DSD128 and above: a very fluid, analogue-like sound.
Your own experience can vary greatly with the components and materials chosen.
Last edited:
Yes and it figures if you examine the digital chain we are used to as computer-audiophiles: noise in the main computer, RFI/EMI, Jitter, PDN noise at the USB receiver, higher power consumption when USB signal integrity is less than ideal -> more noise near the DAC chip and the DAC clock.
There is a wide variety of phenomena right inside the computer as well: motherboard traces and the very way memory is constructed, refreshed and accessed makes a lot of difference. Motherboard traces, as well as internal graphics cards, when switching, also emit interference.
So when you've worked to reduce the 'digital' in that chain, it can help greatly. DSD is particularly suited, IMO, but PCM can sound fantastic too.
My rig isn't even near what can be done in addition to what I did and which some people have actually performed, so I know I can improve on it. But currently it is very satisfying.
For example, I could use HQ Player rather than Audirvana in client-server mode, with a network-attached audio device doing just the buffering and sending near the DAC: this prevents having a big computer near the DAC.
Another useful device on my wish list is the USB Regen by Uptone Audio: It cleans up and regenerates the signal near the DAC so that there's less noise from the DAC PHY when trying to work more to compensate the lack of signal integrity.
I should also one day install some good linear power supplies rather than use switching-mode ones. A cool thing to DIY.
Last edited:
Really? What do most DACs already do internally anyway?
Most are SDM and do internal over-sampling of your PCM signal.
But maybe the gentler filtering allows for better sound as perceived by some, more specifically here, it's about transient reproduction and attack dynamics and soundstage rather than something having to do with frequency range. In implementation terms, it's more to do with the phase response and things like group delay and filter ringing characteristics (pre-ringing and post-ringing).
- Status
- Not open for further replies.