Thoughts on a bunch of DACs (and why delta-sigma kinda sucks, just to get you to think about stuff)
May 10, 2015 at 8:07 PM Post #5,191 of 6,500
The hard part was finding a NOS, TDA1541A-based DAC that looked good on paper OR wasn't some cheap design, so I really didn't have any other options beyond the Audial Model S.

 
You may have told me this already, but was there another DAC that was even close to consideration VS the Audial? 
 
ZMFheadphones ZMF headphones hand-crafts wood headphones in Chicago, USA with special attention to exceptional sound and craftsmanship. Stay updated on ZMFheadphones at their sponsor profile on Head-Fi.
 
https://www.facebook.com/ZMFheadphones https://twitter.com/ZMFheadphones https://www.instagram.com/zmfheadphones/?hl=en http://www.zmfheadphones.com/zmf-originals/ contactzmf@gmail.com
May 10, 2015 at 8:22 PM Post #5,192 of 6,500
 
You may have told me this already, but was there another DAC that was even close to consideration VS the Audial? 

 
There were a couple cheaper options I considered. Not all were Philips/TDA-based, though, as a side note, I do think there are too many high-priced TDA1543-based DACs on the market (it's a budget chip and performs as such! C'mon!). Most I ruled out for one reason or another after research. One that I tried, but will not name, was a no-go. Then there was the potential of used TotalDACs, but those are quite pricey, plus the possibility of them not really reaching beyond 14-bit accuracy due to the 0.01% resistors despite great static measurements concerned me for such a high price. I almost considered the Metrum Pavane until the price was announced.
 
I got the Model S with RCA and XLR output, transformer coupled (so, most expensive version of it). Cost about $2.6K, and given I wanted something very specific to my tastes, I feel I got my money's worth. Consider that the base Metrum Hex model is at least, and often more than, $2.5K without USB input and with stock output transformers, and the Model S blows it away, IMO.
 
BTW, Pedja of Audial informed me the USB version of the Model S has lower jitter than the SPDIF version and sounds a bit sharper too, though with an ever so slightly less rich tone. He and I both agreed the SPDIF version would probably suit my hardware chain and personal tastes better.
 
May 10, 2015 at 9:05 PM Post #5,193 of 6,500
@hans030390, sounds like a stunning little DAC. Do you also have a Yggy? I think I remember you saying you did, and if so, is the Audail the one you currently use more? I must admit to knowing nothing about the Audail and as 2016 is my year for an upgrade in DACs I just may have to learn more about it. Cheers.
 
May 10, 2015 at 9:07 PM Post #5,194 of 6,500
No, I have the Audial Model S and Theta Gen Va. No Yggy. Waiting on trickle down tech, since non-oversampling is my go-to. Really no need to spend so much money on what would be my backup, complementary DAC. Also, Model S isn't what I'd call little. :) Not as big as the Theta, but more in line with DACs that size than not.
 
BTW, if you have not yet heard a non-oversampling DAC, best to do so either with a loaner, at a meet, or sampling a budget offering before diving in too deep. Most are better served by and prefer oversampling DACs. I have particular tastes for a variety of reasons.
 
May 10, 2015 at 9:28 PM Post #5,195 of 6,500
I feel that NOS is something that you either really like it or don't. It sounds can grainy but at the same it's pleasant. I think most people who started their head-fi hobbies in the past 10-15years, would have grown used to the Oversampled dac sound.
 
May 10, 2015 at 9:34 PM Post #5,196 of 6,500
Generally NOS for me is the opposite of grainy, with some exceptions. Though the guy I sold my Hex to said it sounded rough in comparison to the PWD, which was the exact opposite of my experience. Can't say I'm entirely sure what factors cause those different perceptions.
 
May 10, 2015 at 9:38 PM Post #5,197 of 6,500
Hopefully I can try a NOS as I have always been intrigued. Correct me if I'm wrong, but the Yggy does oversample or have I gotten my information mixed up? So many DACs that I have been reading about (and sadly not hearing).
 
May 10, 2015 at 9:57 PM Post #5,199 of 6,500
  Hopefully I can try a NOS as I have always been intrigued. Correct me if I'm wrong, but the Yggy does oversample or have I gotten my information mixed up? So many DACs that I have been reading about (and sadly not hearing).

 
Yggy's digital filter/oversampling is one of its main selling points. 8x oversampling, IIRC. No rate converter/upsampling prior, though.
 
May 10, 2015 at 10:25 PM Post #5,200 of 6,500
I already pointed you towards the right track some posts ago. Do some research on Dac accuracy, it's part of the specs, and expressed in LSB INL.

 
The dac chip chosen by MM is accurate to 1ppm IIRC.
Are you sure you're not referring to oversampling vs upsampling?
 
May 10, 2015 at 11:34 PM Post #5,201 of 6,500
On the topic of the sound of R2R and D-S:
 
Generally R2R is more fluid sounding, has stronger tone density, more texture, more distinct richer timbre. D-S DACs may have treble glare, hash, rasp, stridency, treble grain, or a temporary grain / lack of fluidity. 
 
The very good D-S DACs minimize the bad stuff, but it's always there to some extent. For example, I find that the M51 and M1 have a very fine treble grain, but it's not annoying at all. (But it's still there, so I guess it is annoying.) Also, I have never heard a D-S DAC that has the a tactile sense of body with dense tones / timbre. Maybe Berkeley, PS Audio, Bricasti come somewhat close, but still not R2R. Other D-S DACs come close to R2R fluidity and lack of treble hash and grain like Lavry DA11 or Gungnir, but unfortunately, they don't resolve as well the Berkeley, Bricasti, etc. There's always some trade-off. Either very high resolution, but with treble crap; or lower resolution with less treble crap. And never having the deep tones and textures of R2R.
 
Most of the R2R of DACs of recent I do not like. Some expensive units from big big names costing thousands of dollars. The reasons is they weren't resolving enough, and often they were too laid-back. Overly polite and forgiving. Blame the PCM1704 or some industrial chips with really crappy specs that should never have been used in audio. Or perhaps blame intentional voicing of DACs to sound like bad turntables.
 
The gymnastic and ballerina analogy is a good one, but oversimplified.
 
The R2R DAC that I first (recently) found rather very interesting was the Audio-GD M7. The M7 by itself with its built-in USB was too ballerina for me. Just typical PCM1704 syrupy bass, smooth lush sound, moderately resolving, but nothing that special. Certainly not worth the hype on HF anymore since the price shot up to well over $2k. It was only until I figured out the i2s pinouts to use the M7 with the Off Ramp 5 converter where I finally felt satisfied with the sound, and made the up/sidegrade from PWD2 to M7/OR5. DACLadder implemented an LVDS HDMI receiver for the i2s on M7. This was the final touch. A combination of both gymnastic and ballerina traits. That is what I wanted. The best of both worlds. Or more precisely, a good portion of both worlds. 100% ballerina, 50-75% gymnast. Would personally like more gymnast, but hey, I can't complain. 
 
I was happy for a while until I started to hear rumors of some DAC wizard with white hair and a big belly who kind of looked like a gruff Santa Claus who kept babbling about missiles and medical equipment. After doing some research, I figured out this dude was the dude at Theta.
 
Yggy gets confirmed by Schiit, I hear marketing speak on mega-burrito filters or whatever. I didn't want to wait, and I also wanted to see if this dude was full of **** about filters, et. al. Also, I hear talk that the PCM63 was the chip for R2R and that in fact, the PCM1704 sucks. (Everything BB after TI bought them sucked more and more. This happens in audio, much like when Harmon bought JBL.) The fact that someone actually dared say that the PCM1704 sucked (I always felt it kinda sucked, but accepted it as a necessary for lack of choice in R2R), made me want to hear the PCM63. (You see, this is why you should never get your panties in a bunch when someone says something sucks.) Anyways, my only other experience with PCM63 was with Linn's Karik/Numerik, and that DAC sucked, so I was doubtful the PCM could really be all that good. But then again, I knew that Linn back in the 90s didn't know what the **** they were doing with digital.
 
So I take the leap and find a DAC with the mega-burrito filter (at least an earlier version) and the PCM63. That would be the Theta Gen V. I didn't know if the Gen V had the mega-burrito filter or not, but I took a guess, and later confirmed with Jason / Mike and that indeed, it was the same filter, but less complex than what was planned in the Yggy because the processing power of chips 25-30 years ago sucked compared to now. (Before this, I actually picked up a Sonic Frontiers DAC and got rid of my M7 immediately.)
 
So basically, the Theta Gen V is like Ho Lee ***. 90% ballerina, 100% gymnast. AND MORE... add magician. The magician stuff is the holographic imaging (best appreciated on speakers, but translates to headphones too.) So a lot starts to make sense. These digital filters are important to space, soundstage, imaging. Anecdotal evidence, but probably explains why the M7's stage is horribly shallow with fuzzy imprecise images. (M7 has its own custom digital filter, but obviously it sucks.)
 
Half a year later, I have a mini-meet. Jason posts, "Hey, can I come? I'll bring Yggy." and I get a few sneak peaks here and there. zerodeefex gets it first for two days. I get it second for seven days to allow for full warm-up as recommended by Santa (yeah, and how many people thought I was full of crap on the warm up stuff now?). The rest as they say is history.
 
So basically when I say delta-sigma sucks, this does not imply R2R doesn't suck. A lot of R2R, especially the current PCM1704 stuff, sucks (too ballerina, not enough gymnast). The only thing I've concluded is that Yggy and some select vintage R2R DACs sound better than the best D-S DAC ($8000) I have heard. The second meaning behind delta-sigma sucks is that designers have gotten lazy and rehashed delta-sigma evaluation boards / kits into some very expensive DACs, leading to a homogeneous enviroment of D-S DACs - more choice is always better. The third reason is that if you do not speak out on what sucks, things will never get better or at least different (REPEAT: You see, this is why you should never get your panties in a bunch when someone says something sucks.) 
 
Finally if you are true audiophile seeker, the minute you hear something better, the gear that was worse instantly and immediately sucks. Only false audiophiles do the circle jerk thing and try to feel good about ill-informed purchases or old-gear which has been surpassed by better (not necessarily newer) designs. I sincerely hope a lot of audio companies and chip makers take note of what Schiit is doing. 
 
P.S.
 
I told Mike Moffat of my reaction (using speakers) on the polarity button. His reply was "It's like whether the image sucks or blows."
 
May 10, 2015 at 11:36 PM Post #5,202 of 6,500
   
Rankings are my own based on my own preferences. Your rankings may be different. I believe I said something to that effect in the first post. Maybe I didn't.
 
 
I have heard DACs with minimum phase and linear phase filters. Some D-S DACs I have owned or borrowed for extended periods offered several filters. PWD2, Gamma2, Vega, LH, etc. (too many to recall.) In these cases, I've almost always preferred the linear phase filter. Smoother to my ears; however I still wouldn't say analog sounding or particularly smooth compared to R2R. The other ninjas in the evaluations preferred other filters. I'd say the filters are more a matter of "different" than better or analog.
 
Using your logic, the smoother "analog" sounding R2R DACs of yesteryear and Yggy would sound very digital because they have linear filter characteristics with both pre and post ringing. BTW, if you didn't know, most of the R2R DACs of yesterday also used cheap pre-canned filters. 
 
Finally, the Audiolab DAC Optimal Transient filters do have pre and post ringing. So there goes your theory. I'd be happy to provide measurements for you if you don't believe me. Don't get bamboozled by their marketing literature. It's impossible to not have any kind of ringing with filters.
 
Sorry to say it, but you don't know what you are talking about.

 
I also prefer linea phase filters - I'm not sure if it's that they sound smoother, but IME more coherent as the minimum phase filters I have been playing with seem to throw out really weird imaging I couldn't live with.  
 
I have been doing this from software using HQPlayer which has an excellent range of filters, dither options, and IMO sounds more transparent than any other software for some reason.  My second preference after HQPlayer would be JRiver, with Foobar in last (I found the WASAPI plugin the least objectionable and digital sounding)
 
I find that I generally prefer to oversample and music by at least 2 x as this seems to reduce some of the nervousness and grain on some poorer recordings, even hi-res formats.  I suspect this is leveraging some kind of coloration and there is slight time domain blurring from the filter I suspect but is sounds more like music.
 
I also wonder how much difference there is between noise filtering in both the digital inputs, sensitivty of DAC chip to noise, as well as the analog output sections.  The Sabre DAC I use shows dramatic differences in noise levels corresponding to changes made to the computer.  I really am struggling to adjust to using a switching PSU after getting used to the linear powered computer, but I will have to as I have damaged the linear supply by overloading.
 
It is one of those things - at first it took my brain a while to get used to lower noise from the digital input - my thoughts were the sound was too laid back, but after a while either burn in took place, or I began to realise that the linear supply was more transparent and detailed, less blurred, less noisy, such that even very good switching supply was unlistenable.
 
I suspect maybe something similar happens with R2R DAC's once they are on a good enough technicalities footing, even very good delta sigma switching DAC's are difficult to listen to.  Previously I had little to complain about in my system using the switching ATX supply, I think the brain becomes used to the noise and just ignores it as much as possible.
 
BUT I still have to figure out what exactly is causing the noise in Sigma Delta, and why it is absent from Multibit.  Especially why Sabre, which sets out to eliminate the problem areas of delta sigma sets multibit fans off so badly given how low level the noise producted by the chip itself is, at least with most of the measurements I have seen, including those on ESS white papers.  There are some very high end, cost no object DAC's which use Sabre chips, often in multiples, presumably to try and filter out what remaining artefacts are created by the chip, and probably not just using the reference design from ESS.
 
I have a suspicion that part of the problem with Sabre is that the noise levels are too low, and that because of this, other noise from the digital inputs, power supply etc are easily heard, while other DACs which more noise in the audio band do a better job of covering up these noise sources.  This is just my own daydream and not something have tested or formulated from heaps of listening experience so take with grain of salt.
 
The other though I have had is that I would not mind NOS DAC with high bandwidth digital input, as when using a computer, there is heaps of unused processing power that can be used to implement very good digital filters.  This would reduce costs for a lot of DAC's as they can forgo using complex filters in high end FPGA chips, and would not need the space in the chasis etc.  Maybe it is better to do some of these things near the DAC chip, but FPGA circuits especially powerful ones also product electrical noise AFAIK.
 
May 10, 2015 at 11:43 PM Post #5,203 of 6,500
  Regarding Schiit's marketing approach, you must feel very strongly about this for you to join Head-fi today just to post both about it and your bad experiences with Schiit.  
 
First, welcome to HF (I think) and secondly, some people get their marketing approach and some don't and some folks have bought their products and loved them and some haven't.  I understand, you don't like them and I'm happy for you that your Auralic stack works well for you.
 
Yggy is the first piece of Schiit that I've owned, so I'm not a fanboy and I also can't comment upon Bifrost's sound, but it is the best DAC I've had in my main system...beating out some DACs that retail for 6X+ Yggy's retail price.  Yes, to my ears it is that good and if it wasn't, I'd say so.
 
It looks like you have some disposable cash, so get a Yggy and decide for yourself if it's hype or not.  If you're in the states, you're only out a couple of hundred bucks if you don't like it. 

 
Not from the States. Where I live, auditioning, shipping, and distribution are a pain in the ass. Disposable cash is currently going towards a tube-fund, been dazzled by the tube aesthetic for a long long time now, so a Yggy or any other piece of gear is not on my roadmap.
 
May 11, 2015 at 12:15 AM Post #5,204 of 6,500
Are you sure you're not referring to oversampling vs upsampling?


Seriously ?
 
I'll point out another word you can look for in google to try and understand : DAC monotonicity. After that you're on your own. Goodbye
 
May 11, 2015 at 12:30 AM Post #5,205 of 6,500
Seriously ?

I'll point out another word you can look for in google to try and understand : DAC monotonicity. After that you're on your own. Goodbye
You got no evidence alright.
 

Users who are viewing this thread

Back
Top