Discussion in 'Digital: DACs, USB converters, decrapifiers' started by GoldenOne, Jul 12, 2021.
EU but makes no difference to me, just takes a bit longer.
Awesome, I'll shoot you a DM shortly.
Main issue is shipping value/insurance cause gotta pay VAT on imports now (thanks brexit.....)
I should have clarified, I was ignoring the first gen Metrum DACs and only referring to anything starting with the Pavane or after.
You know, those Metrum DAC modules are upgradable/rollable. And with their 3rd gen DAC modules out (ignoring modules from Morpheus), in theory I think you could test a Metrum DAC with three generations of their DACs. Could be interesting, since it would appear they've changed their glue logic with each design. But costs can add up quicly.
Anyway, I am with you on a preference for DAC ICs. Accuracy and precision more or less guaranteed there with less need for tricks.
It may not necessarily be unpleasant. I have this working theory that it's the NOS DACs (in the current modern sense) lack of a proper construction filter / analog LPF (and inevitable aliasing) that results in this "denser tone colors presentation" - noting that not all NOS DACs sound like this and some are in-between or shades of grey.
I can find the sound enjoyable, but ultimately it feels like putting butter on steak.
Which seems to be a thing people enjoy. At least it is not blue cheese ...
Ya, that's fair. I wonder if the photography-related metaphor can be of help here, or is our processing of visual and audio too different. Much of the digital photo sharpening is basically convolution filters that reduce resolution but do make the image feel sharper. Similarly turning up contrast results in greater color saturation (up to a point). We are likely seeing similar effects here - simplification of the data makes it more vivid.
Those illustrations of the effect of say, my Venus II, over even something like the Convert-2 or DAC3, resonate with me (no pun intended.) Guitar tones, which I am particularly fond of, just have an extra gear to them but they can sound a tiny bit overcooked by comparison. I'm also battling a little bit of richness in the lower midbass that's kinda driving me a little nuts, but it took a really long time and some unrelated changes to really stand out as unfixed as of yet.
We'll see how May sounds by contrast, in a 6-7 weeks. Thanks for breaking it down for simples like me
This kinda sums it up for me.
I love the option of having NOS. It's something that I do like to use sometimes. For certain genres or just in the mood. But normally listen with HQP.
To me the appeal of NOS is that I find it easier to follow along with transients in fast music, the thickness has always been an unfortunate downside(and the reason I ventured outside of NOS territory.) I'd even describe entry level Metrum(Flint DAC1 modules) as cleanly resolved in the temporal domain despite displaying a syrupy laid back sound.
Even now with a fast clear Gustard X16 Sabre DAC I don't think the speed is quite at NOS levels although it's much sharper. Said Sabre DAC also has it's own 'NOS mode' which I'm not even sure is working as I hear no difference with it.
Three pages in and I'm surprised no one has noted the lack of ringing on the ARES 'not quite NOS' impulse response. This has always been the major selling point of NOS designs and likely why they labelled it as such. I'm curious if the sound in this mode would be closer to real NOS R2R or the modes modern S/D DACs are increasingly using, and how fast it would sound subjectively.
In regards to the ringing this is something that can be done with OS too.
A very slow sinc filter. Cubic/polynomial interpolation etc will have no ringing or almost no ringing
Clean/fluid transient and rhythm reproduction is a perfect description of one thing real NOS does well.
I think there is more to it than ringing, take for example the 'super slow' filter option that recent AKM DAC chips support, it has a NOS-like stair stepped waveform (no ringing, perfect impulse response, treble roll off) but it sounds like complete shit, it only captures the negative qualities of NOS with the same base character as any of the other filter modes. Totally different to comparing a real NOS DAC with and without OS.
Because of this I have my suspicions part of the real NOS sound could be related to differences in clock speeds and possibly the complete absence of a digital processing stage, no idea why this would be though.
This has been a fascinating read! I've had the Ares II for a few weeks now and have been playing with OS vs NOS and liked OS better, and I couldn't really place why. Maybe now I have more insight. In NOS mode it seemed like the mids took a slight backseat, maybe it was this sharpness on the high-end I was hearing and drowning out the mids.
Depends upon what you mean by "real NOS".
I owned a first generation Sony CDP, no oversampling. Tonally, it sounds more like a Modi 3+ with the opamp replaced with an OPA1656 (less phat, less warm, more neutral). That Sony "NOS" CDP I wouldn't describe as fluid sounding either.
The Holo NOS presentation is quite different from Metrum, Denafrips, AGD NOS specials, etc. The latter have that denser tone color and liquid presentation - similar to MSB DACs' presentation, which are OS.
The Holo NOS Spring sounds surprisingly normal, actually a tiny bit on the zippy side, but with an NOS twist with respect to soundstage.
This is one of the things I quite like about the holo dacs. The nos mode is actually really really good.
I know a lot of people would object to having a low-pass filter in nos but honestly I can only assume that's part of why it sounds the way it does.
I still prefer hqp for most of the time. But nos on the may is as you said, surprisingly 'normal' sounding.
Msb stuff I'm interested to try. What do you think about those? Are they just a meme or genuinely top tier stuff?
Real NOS is just.... NOS, no oversampling, what NOS has always been before Denafrips.
NOS or OS is not a big determining factor of the sound in the grand scheme of things as you mention.
You can transform any NOS DAC into an OS DAC with external OS, that's the only meaningful way to make a comparison between them.
I havent heard a NOS DAC with a steep analog reconstruction filter (unless you count ears), it was different DIY stuff always with simple 1st or 2nd order filtering, same as Holo DAC filtering it seems.
The amp you use and how it handles the ultrasonic stuff plays into it I think.
Yeah I imagine the rest of the chain will be a notable factor in terms of the resulting sound with NOS. Some amp/transducer combos might not respond well to ultrasonic content whereas others might be fine
Several times i have had the feeling that (more than) some of the people that really like the Spring are not the biggest NOS sound, and it's disadvantages, fans.
I remember, as i explained at the time, that it was hard for me to choose between the Spring 2 (not KTE, i believe it was LVL2) and the Jade, which i ended up choosing in the end.
After all this time i enjoy my setup, because it just works very well with my (altered) HD800's but i still miss that different sound in some types of music, specially with my HD600's.
Most certainly on the last point, let the rest of the chain be the analog reconstruction filter.
With respect to Denafrips, I have to laugh. One can argue that they intentionally implemented aliasing (via connect the dots) to get their desired "denser NOS tone" while running a higher sample rate. Either that or incompetence. Or both.
As I mentioned, the result could be heard as a pleasing - measurements be damned!
Have there been any further investigations/understanding around the weird spike with the slow filter's impulse response measurement?
Soundstage just published a video about the Denafrips DACs, specifically the Ares 2 and Termi Plus. The video is not a review per se (they have written reviews on each DAC on their sites), but more talking about Denafrips in general, their DAC measurements, and how they like the sound of their DACs a lot.
In the video, they give a shout out to @GoldenOne for uncovering the Ares II NOS mode funkiness, but basically saying they wouldn’t use their DACs in NOS mode anyhow because of the HF rolloff and aliasing.
Eh, Ares 2 even with fake NOS still sounds better than their truly atrocious built-in OS solution. If the "NOS" mode isn't for you then it's recommended to run this with some form of software upsampling.
Separate names with a comma.