DAC output stage vs technology

[@user-od9iz9cv1w]

Great answer, thanks. I never knew why it was called R2R. I'd go so far as to say the chip is the least important to sound. The power supplies, the output stage, I/V, the clocking of the I2S signal all impact sound more than the DAC. Heck, just the vibration dampening and EMI and electric noise shielding will impact the sound. With all of those things optimized then the choice of DAC chip/process is a matter of taste like a fine wine. In the end if you like the sound, you are done.

[@mikeg2491]

I have several DACs based on the tda1541 chip and they all sound very different from one another

[@user-od9iz9cv1w]

@@mikeg2491 I can imagine. I have used the TDA1541a chip now for 15 years and really like it. I have had a couple of CD Players and build numerous DAC's around this trusty old chip. The more effort you put into it's implementation the more you are rewarded with natural engaging beautiful sound.

[@tristanjones7735]

Ah yes output stages. This is the single biggest advantage DSD has over everything else. With DSD, you only need a low pass filter. That's it. With R2R you need a current to voltage converter. And let me be the first to tell you, they are a living nightmare to design. One thing everyone loves to forget about dacs is that power supply noise on the output will reduce your effective bit depth. Output stages HAVE to have less than 1uv of noise in order to maintain -120b SNR at 1v out. So trying to design a complicated IV stage when the parts themselves have a higher than 1uv self noise is almost impossible.

[@ShawnWrona]

Thank you… I learned something new today

[@adotopp1865]

Don't concern yourself with the technology just listen to the thing.

[@AWEG-qu8bz]

Thanks Paul. That was much needed. So the next step for the manufacturer is to explain to the customer how the analog output Stage is designed. That might slow down the DAC reviewer madness a bit. With some manufacturers you'd think they bring out 4 new DACs a year. Crazy.

[@SteveD-m6z]

Perhaps George of Houston has purchased a Denafrips R2R DAC. The DAC design has an unbuffered output stage. This makes the DAC sensitive to the input impedance of the pre-amp. The interconnect cables will also influence the performance.

[@mikerodriguez5868]

Yes, the output stage definitely defines the DAC. There are some wonderful sounding older DACs. Also, some wonderful sounding DACs using simple chips like the PCM1794 chip.

[@cjimcook]

@Paul McGowan I don't understand the difficulty of producing a precision resistor ladder. I remember seeing samples of variable resistors in a SIP package with an adjustment screw on the end back in the early 1970s, if not the 1960s. Say you are trying to produce a 100 ohm step, if the adjustment is too coarse on one - say +/- 1ohm on a 100ohm SIP - then put it in series with a second 1 ohm SIP that was +/- .1 ohm and so on, finer and finer, until you got the required level of adjustable precision. What am I missing?

[@DP-qp8wr]

Thanks Paul. Now I understand the DAC magic.

[@paulrs2975]

I thought that R2R refers to an arrangement where only 2 resistor values are used: some value, and twice that value. The resistors are then switched in, sometimes multiple resistors, to create the proper current output for that word.

[@johnstone7697]

I have a Gustard DAC using a pair of AKM 4499's and a discrete output stage. I also have a Theta GenVa DAC from 1996 that used the legendary Burr Brown PCM 63 ladder DAC chips in a quad balanced configuration, with a bunch of fancy proprietary DSP in front of them. . I've compared these 2 DACS using 16/44 files on numerous occasions, and never once have I been able to reliably pick out which was which on a blind test. Given that these two DACs are nearly 25 years apart, it tells me that digital decoding has been a "solved problem" for a long time. Even though the Gustard measures far better, the Theta still measures well below the threshold of audibility for distortion, noise, jitter, etc. So it comes as no surprise that audible differences-if any- would be extremely hard to discern. I really wish these companies claiming their DACs sound so much better, even though they have no concrete specs to back that up, would show some proof to back up their claims. But then again, this is high end audio, and they'll say "We don't need no stinkin' proof".

[@stevemd8947]

I own an Adcom GDA 600 DAC (1994) with the BB PCM 63. Recently, I compared my DAC to a D/S dac ($4000). For me the Adcom was superior. I listen to Classical Piano Music exclusively. The D/S DAC stripped the piano of its natural harmonic richness etc. So, I am very happy with my 30 year old DAC. There is something special about the BB PCM 63 chip. I also, own a BAT CD Player made in 1998 with the same BB PCM 63 chip which is also great for a front end.

[@babubabu12345]

Very Good Information, Thanks Paul Sir...🙏

[@JingoLoBa57]

So a tube output stage should be sweet… certainly sounds like it in various Audio Research cd and DAC designs.

[@SuperMcgenius]

My 10 year old Burson DAC which has Separate power supplies for digital and analog, and all Discrete circuits still holds its own. Yes there better sounding DAC’s today, but the difference is small like changing cables. Note : this is a full range system in the 25 k range. Yes just use your ears. 😊

[@geoff37s38]

Put your $4000 towards quality loudspeakers and attend to room acoustics. This will yield real results.

[@glenncurry3041]

It can not be argued that all else equal, an R to R DAC will be more accurate in reproducing the individual samples in PCM. A properly designed R to R DAC will literally reproduce the exact voltage captured from the source ADC. It will do so for each sample as they are sent to the DAC. While a Sigma Delta has no idea what the actual voltage of any particular sample is. All it knows is on a single sample (multiple bits) level whether to turn on for that clock cycle or not. The DAC will try to slowly align with the moving signal and hopefully eventually ride the sample to sample changes correctly. But at no point in time does a Sigma Delta actually reference, verify it's output compared to the specific voltage value of any sample. IOW if you start playing/ sending a PCM file to a DAC and the first sample is a high voltage, the R to R will reproduce that voltage while the Sigma Delta's output will just be a single on bit for the sample duration.

[@johnstone7697]

That makes absolutely no sense. What measurement can you point to that shows the superiority of a R2R DAC over a Delta Sigma DAC? It seems to me that if what you said was true, the biggest issue would then be that the Delta Sigma would have measurably worse sample value accuracy than the R2R. But just about any Delta Sigma DAC will show the exact opposite with respect to linearity. The majority measure virtually perfect in this regard. Kindly point out where the measurements reveal the flaws that you claim.

[@glenncurry3041]

@@johnstone7697 It makes complete sense to anyone that knows how to read schematics and understand electronic designs. Send one sample to an R to R and it will output that voltage level until the nest sample. Send one sample into a Sigma Delta and you will get one sample duration pulse... or not.

[@wty1313]

The issue isn't necessarily the DESIGN of an R2R Dac. If anything, it's easier to design properly than a delta-sigma Dac. It is, after all, one of the oldest Dac circuits. The main problem, as Paul alluded to, is the MANUFACTURING and IMPLEMENTATION. In order for R2R to TRULY work correctly as intended, it takes incredibly accurate resistors within the ladder that are matched to such a degree that it's actually not possible with current manufacturing processes without insane time and expense - no matter what any R2R manufacturer says.

[@glenncurry3041]

@@wty1313 OK. Has nothing to do with basic design level discussion. Which is exactly why I set mine up that way. Thus "all else equal". By pure definition a R to R is based on each samples value, while a Sigma Delta has no idea what that sample's value actually is.

[@johnstone7697]

@@glenncurry3041 The only thing that is important in a DAC is the accuracy of the output analog waveform relative to the digital input signal, period. And that accuracy is determined upon an established set of measurements. Where in those measurements of a Delta Sigma DAC does this supposed "deficiency" of yours show up? It either accurately converts the digital stream, or it doesn't. If it can't reconstruct the analog waveform correctly, that will show up in the measurements, plain and simple. Your statement that D/S DAC's 'have no idea what a sample's value actually is" is utter tripe. Were that the case, the damn things wouldn't work at all, much less produce SOTA results.

[@reasoned007]

now, R2R dacs at 24 or 32 bit are common

[@cengeb]

They all sound different,so which one is most accurate? And why is someone's imagination making one sound better,better referenced to watch. Random random

[@doylewayne3940]

ummm... sorry Paul you lost me on that explanation

[@DrNoahBoddy004]

I understand all too well! And it’s why I’ll n e v e r, Never, NEVER leave a n a l o g sound sources: phono, tuner, and tape. CD, and for that matter, HiMD will never lose their convenience of portability which is great! But this business of chasing DACs across the galaxy? N😢t by any stretch!

[@5starmaniac]

Tape.. ?! 😮

[@DrNoahBoddy004]

@@5starmaniac Amazon has access to many linked sources of analog cassette tape.

[@5starmaniac]

@@DrNoahBoddy004 I'm more concerned about the sound.. But as a nostalgic trip, I can understand it, as I grew up during the 80's 😃

[@DrNoahBoddy004]

@@5starmaniac my only demands as far as SQ, have already been solved in the use of full tri-amplification for the speaker systems, and the use of electronic crossovers. New drive belts, clean tape heads, and full demagnetization have always cleared the way for their continued use, in my set-up.

[@5starmaniac]

@@DrNoahBoddy004 As long as you're happy with it, that's all that matters 👌😊

[@stevenholquin2127]

If Some Guy From Huston Texas is Asking a Question Like This His Doesn’t Need Paul To Answer It He is Well Enough Knowledgeable About How The DAC Works and Its Characteristics and Parameters

[@razzman2987]

It is the denafrips hype the chinese make everybody crazy 😅

[@maffs6270]

can i suggest he's underwhelmed because he's spent $4000!! you can get an amazing dac for 10 times less, at the end of the day, it's only changing digital to analogue, it's not going to make you a cup of coffee and wash your car. welcome to the world of realising that spending thousands on hifi isn't necessarily going to make your music sound any better (unless you've tricked your brain in to thinking it does to justify the spend).

[@markritacco270]

If you can not hear the difference, check your hearing at a doctor before BLOWING tons of money!

[@mikeg2491]

And an amp is “only” amplifying a sound signal. But let’s be real my McIntosh system is going to sound better than an op-amp altoid can amplifier lol. You don’t need to spend big bucks for a good system but you should also expect to invest a little too for something not absolutely janky.

[@tothemax324]

Please don't criticise people that have more experience or a deeper wallet than you.