PCM61P DAC chip info?

[garbulky]

I'm trying to figure out if this is a true multibit dac or a delta sigma dac? Burr Brown PCM61P
Context it's from a Denon DCD 660 which says it's a " 20-Bit 8x Oversampling DAC".
DAC is a "Serial Input 18-Bit Monolithic Audio
DIGITAL-TO-ANALOG CONVERTER"
more info
www.lampizator.eu/lampizator/LINKS%20AND%20DOWNLOADS/DATAMINING/pcm61P.pdf

[audiobill]

From The Elitist Journal:

Burr-Brown: Multibit:

PCM51, PCM52 – 16-bit predecessors PCM53, a ceramic package
PCM53 16-bit, one of the old, but good (1983-86 years), with a parallel input
PCM54 16 bit multibitnik normal (slightly older PCM56), parallel input
PCM55 16 bits, the same as the PCM54, only in a small SOIC package
PCM56 16 bit, normal quality multibitnik
PCM58 18-bit, a very good DAC
PCM61 18-bit, the quality of a PCM56, but 18-bit
PCM63 20-bit, one of the best of the world’s Audio
PCM64 18-bit, one of the best, with a parallel input
PCM65 – one of the best, most expensive and rare multibitnikov, with parallel input
PCM66 16-bit DAC-2 and, huddled in a single housing, the quality – so so
PCM67 hybrid (multi + one-bit), 2 DACs and in one case (stereo)
PCM69 hybrid (multi + one bit), the same as the PCM67
PCM1700 18-bit, 2-DAC and in one case (stereo), the quality level PCM61
PCM1701 18-bit, one of the best, the successor technology PCM58
PCM1702 20-bit, the quality just below the PCM63

[KeithL]

It's an 18 bit multi-bit DAC.
(Remember that CDs are only 16 bit... so a 16 bit DAC is fine for a CD player.)

Do bear in mind that multi-bit DACs are NOT something new and innovative...
ALL really old DACs are multi-bit... because it's "how a basic DAC works".
When the performance, and the price-to-performance ratio, of multi-bit DACs proved insufficient for modern audio equipment, Delta-Sigma DACs were developed...
Delta-Sigma DACs have pretty much replaced multi-bit DACs for audio applications, which is why folks like Schiit have to adapt DAC chips intended for other applications nowadays.
(Multi-bit DACs are mostly only used these days where high-speed is required and really good accuracy and linearity are not especially important.)
According to its data sheet the PCM61P was discontinued in 1999.

Apr 17, 2018 15:45:04 GMT -5 garbulky said:

I'm trying to figure out if this is a true multibit dac or a delta sigma dac? Burr Brown PCM61P
Context it's from a Denon DCD 660 which says it's a " 20-Bit 8x Oversampling DAC".
DAC is a "Serial Input 18-Bit Monolithic Audio
DIGITAL-TO-ANALOG CONVERTER"
more info
www.lampizator.eu/lampizator/LINKS%20AND%20DOWNLOADS/DATAMINING/pcm61P.pdf

[Casey Leedom]

So KeithL, when a Delta-Sigma (ΔΣ) DAC claims that it's "32-bit", I assume that that's referring to the reconstructed accuracy? Just curious about terminology.

Casey

[garbulky]

Apr 18, 2018 10:38:48 GMT -5 Casey Leedom said:

So KeithL, when a Delta-Sigma (ΔΣ) DAC claims that it's "32-bit", I assume that that's referring to the reconstructed accuracy? Just curious about terminology.

Casey

I always thought it just meant it could decode 32 bit files?

[ÈlTwo]

Isn't 32 bits just $4?

[garbulky]

Thanks KeithL and audiobill I think you guys are both right on the specs. It is a true multibit dac.
I also got this from Audio science review from restorer john.
"The PCM-61P is a current output, 18bit true multibit DAC. It used nichrome thin film resistors, laser trimmed and is a pin for pin 18 bit version of the PCM-56P.

The Denon implementation used a bit shifting system not unlike Yamaha's HiBit 16-18bit shifter around the same time. Technics also were bit shifting around the same time with 16-18bit advertising. It was deep in the OS and bit wars of the late 80s.""
So interesting stuff like the whole laser trim and stuff.
I haven't listened to it yet.

[KeithL]

Multi-bit DACs require a series of precisely matched resistors if they are to perform even decently.
Back in the old days 1% separate resistors were considered to be "precision" - and it was really hard to make closely matched resistors inside an IC.
LASER trimming was about the only way to get reasonable precision with on-chip resistors back in those days.
(But, when it was first developed, it was a big deal, and worth mentioning.)

Remember that, before 24 bit files, the best resolution you needed was 16 bits..... because CDs were 16 bits.
Therefore, all the machinations about 18-bit performance were simply aimed at delivering better performance with 16 bit files.
(Having extra bits to play with can also help you avoid rounding errors when you do the interpolation involved in oversampling.)

However, just to put the performance we're talking about into perspective:
- a 24 bit DAC is NOT "one and a half times as precise as a 16 bit DAC".
- a 24 bit DAC delivers 256 TIMES the resolution of a 16 bit DAC.

That means that, using the same design topology, compared to a 16 bit DAC, you would need resistors 200x as precise to deliver full resolution performance in a 24 bit DAC.
(And Delta-Sigma DACs deliver equivalent or better resolution with MUCH less stringent parts and fabrication requirements.... which is why they replaced the older type so quickly.)
Of course, though, if the DAC is inside a CD player, then it will never be called upon to play 24 bit files anyway.

Apr 18, 2018 11:48:40 GMT -5 garbulky said:

Thanks KeithL and audiobill I think you guys are both right on the specs. It is a true multibit dac.
I also got this from Audio science review from restorer john.
"The PCM-61P is a current output, 18bit true multibit DAC. It used nichrome thin film resistors, laser trimmed and is a pin for pin 18 bit version of the PCM-56P.

The Denon implementation used a bit shifting system not unlike Yamaha's HiBit 16-18bit shifter around the same time. Technics also were bit shifting around the same time with 16-18bit advertising. It was deep in the OS and bit wars of the late 80s.""
So interesting stuff like the whole laser trim and stuff.
I haven't listened to it yet.

[audiobill]

Before we get too lost, here's another informed point of view:

Things to be aware of in choosing an Audiophile Level DAC

It’s a complicated world out there in the land of digital audio with cheaper and cheaper solutions coming out on the market with higher and higher sampling rates! Many manufacturers are offering very high priced DAC’s and others have offerings that cost next to nothing! – Some DAC’s come in large rack mount cases – others are part of the USB connector!– There is a lot of confusion out there about sampling rates, dither , jitter, oversampling, digital filters, quantization errors, bit widths, USB, SPDIF, 24/96, 192, 384, SACD……. and where is it all going.

Read on to understand what is really going on in digital Audio. Hopefully this article will help you to make an informed decision when it comes to buying a DAC and understand more about the current DAC you have. Lets start by discussing some major points first and get into the detail later!

MAJOR POINT #1

“The actual conversion process from digital to analog is very critical” and this is what is at the heart of any DAC – Now listen closely, there are two major methods of accomplishing this conversion from digital information stored on a computer or a disk – the original method and the one currenly used by xxxx today along with a few other manufacturer’s is by “hardwiring” the digital numbers through an electrical circuit made up of resistors, voltages & currents and outputting a voltage driven directly from the input – as the digital numbers change at the input of our circuit the output voltage changes. This method is referred to as a Resistor Ladder DAC or R-2R.

Its really beautifully simple and effective:
◾No filters in the digital domain
◾No filters in the audio domain
◾No altering or modifying the signal
◾Just pure conversion

Of course there are complexities in the precision resistors and the finer points of the circuits but bottom line is it’s a direct HARDWIRING of digital numbers to actual output Voltage. For an audiophile level DAC this is correct strategy. END OF STORY.

MAJOR POINT #2

The second method of performing this critical conversion is through computer horse power – basically the computer hardware reads in the first bit of data and continues to read in bits and analyze via complex mathematical algorithms, digital signal processing , quantization , dither , noise injection & feedback and the computer will determine a voltage that has been carefully computationally arrived at. This is referred to as a DELTA-SIGMA.

You may think this is not a very good process compared to good old fashioned hardwired R-2R, but the truth the computer processing method has many advantages:
◾It is inexpensive compared to its R-2R counterpart & very convenient for many musical applications : Internet radio, Home theater, 5 or 7 channel movies with surround sound, Mp3, iPods, digital video cameras, even SACD uses this method.
◾It allows for great versatility, filtering, sound shaping, room correction etc – incredible flexibility and for 99.5% of the music world DELTA-SIGMA is a great way to go.
◾Its cheap ( Seen the cost of many 24/96 and 24/192 and 384Khz sampling DAC’s) – ( DELTA-SIGMA DAC’s love high resolution formats because its easy to process them compared to R-2R). Its easy to manufacture chips with tons of processing, thus SIGMA DELTA dacs are here to stay!

Even companies like Analog Devices that made a range of R-2R DAC chips in the 90’s now have a complete suite of DELTA-SIGMA chips with tons of features. Basically all home theater gear, CD players, video cameras etc.etc. will use DELTA-SIGMA processing. Virtually all modern DAC’s are DELTA-SIGMA based to handle the 24/192 supposedly high resolution formats.

So why did Manufacturers move away from R-2R chips that they developed in the 90’s? Well basically it is an expensive process to create integrated circuits with laser trimmed precision resistors to correctly implement the R-2R circuit. Its much easier to throw a few million gates at the problem! The Analog Devices AD1865 chip used in the Audio Note DAC’s is the highest bit resolution DAC chip that was made with 18 bits and provides 108db of potential dynamic range. So basically when you put a 24/96 signal through an Audio Note DAC you have a potential dynamic range capability of 108db which is approaching twice the dynamic range ever recorded.

So which is the right DAC for you – R-2R or DELTA-SIGMA? For a true audiophile level DAC we highly recommend R-2R!

If you are truly concerned with the very best of 2 channel audio the R-2R DAC’s will clearly outshine the computer processing DELTA-SIGMA dacs for obvious reasons, If you have a high end audio system such as a Single Ended Tube 300B or 2A3 amp or possibly CLASS A push Pull tube amplification , transformer coupled pre-amplifiers, high efficiency speakers, it’s the only way to go.

Listen to a saxophone with an R-2R DAC and you are going to pick up all the subtle nuances associated with this instrument – the attack , the timbre, the decay, these subtle nuances that immediately allow the brain to identify the instrument are absolutely critical to an audiophile.

[Casey Leedom]

It looks like the source of the above article is from ANK Audiokits in their paper What you need to be aware of in choosing an AudioPhile Level DAC - March 2013 Part1.


Casey

[KeithL]

Yes...... mostly....... but.........

In basic terms, a 16 bit DAC is a DAC that can convert 16 bit numbers into analog audio, and a 24 bit DAC is one that can convert 24 bit numbers.
The issue is that there are certain types of small errors that become very important when it comes to DACs.
For example, I could have a 16 bit DAC - which accepts 16 bit numbers and converts them with near perfect accuracy.
But, how about if I had a 24 bit DAC - which accepted 24 bit numbers - but was only ACCURATE to 16 bits.
So, on the first DAC, you feed in any 16 bit number, and the right voltage comes out.
But, on the second DAC, you can feed in any 24 bit number.... BUT THE OUTPUT VOLTAGE WILL ONLY BE ACCURATE AS FAR AS THE FIRST 16 BITS (and the remaining 8 bits will be more or less garbled).
We would properly term this "a 24 bit DAC that only has 16 bits of precision" (or, colloquially, as "a 24 bit DAC that really doesn't work any better than a 16 bit DAC")
And there are several sorts of "in-betweens"..... for example, our 24 bit DAC might handle the top 16 bits perfectly, but, for those bottom 8 bits, while smaller numbers generally give us smaller values, the steps are a little OFF.
We might say that it's "a 24 bit DAC with some linearity errors near the bottom".... or we might do some math and decide that our 24 bit DAC "was only resolving 21 bits correctly".
(And, if the bottom numbers are horribly inaccurate, but bigger numbers always give us bigger outputs, we might say that "it has good monotonicity but poor precision".)

The reality is that a 24 bit signal has a dynamic range of something like 140 dB.
Another way of looking at it is that, if your DAC chip has a S/N of 135 dB - YOU WON'T BE ABLE TO TELL IF IT'S 24 BITS OR 32 BITS BECAUSE ALL THE DIFFERENCES WILL BE BELOW THE NOISE FLOOR.
All of the extra information that your 32 bit file contains, but that's missing from the 24 bit file, will produce analog differences that fall below the noise floor where you can't hear them.
(Of course, even if you could make a DAC with a S/N of 180 dB, no amplifier has that much dynamic range, and neither do your ears.)

However, when you get to COMPLICATED DACs it becomes even more complicated.
For example, Sabre DACs are rated to handle 32 bits.... but they have a S/N of somewhere around 135 dB.... so the difference between 24 bits and 32 bits SEEMS useless.
Except that Sabre DACs have some DSP-type features, like a digital volume control.
And, being able to accept a 32 bit digital audio signal, AND PERFORM THE VOLUME CONTROL CALCULATIONS AT 32 BITS INSTEAD OF 24 BITS, helps avoid some significant rounding errors that might crop up in the math.
And, likewise, it's POSSIBLE that being able to do the oversampling at 32 bits instead of 24 bits might eliminate some rounding errors there.
Note that we've drifted away from "DAC functions" and into "a bunch of extra stuff they put in DAC chips these days".

To be honest, unless you want to REALLY dig into the details, it's best NOT to become too mired in the details.
Just accept that "a 24 bit DAC will accept 24 bit inputs" and then look at the ANALOG specs to see how good a job it makes of the whole project.
(It doesn't really MATTER if the DAC is only delivering 22 bits of accuracy because of digital limitations or analog limitations.... either way you're getting an analog output with "22 equivalent bits of accuracy".)

Always remember that, regardless of what goes on inside a DAC, a stream of numbers goes in, AND ANALOG AUDIO COMES OUT.
So, while it's may be academically very interesting, as long as the result is good, it doesn't actually matter how the numbers get sliced, diced, converted, and compacted along the way.
All that counts is how close the analog audio that comes out is to what it theoretically should be.

And, considering that nobody makes a DAC with the dynamic range to produce analog audio with 32 bits of resolution.....
a "32 bit DAC" really is "just a DAC that accepts a 32 bit input signal" or maybe "a DAC with DSP that accepts 32 bit signals and does its calculations at 32 bits".

Apr 18, 2018 11:38:19 GMT -5 garbulky said:

Apr 18, 2018 10:38:48 GMT -5 Casey Leedom said:

So KeithL , when a Delta-Sigma (ΔΣ) DAC claims that it's "32-bit", I assume that that's referring to the reconstructed accuracy? Just curious about terminology.

Casey

I always thought it just meant it could decode 32 bit files?

[Casey Leedom]

Thanks KeithL. I was also looking at the product page for the Asahi Kasei Microdevices AK4490EN which I believe is what you're using in the RMC-1 (and probably for the front three channels of the XMC-1G3 and maybe even the DC-2). Similar to the ESS Sabre DAC example you offered, the AK4490EN is "32 bits" but its Data Sheet shows it as only having a Signal/Noise Ratio of 123dB in Mono Mode.

Also, I've never seen any digital audio file with 32-bit samples. That would e pretty crazy ...

Casey

[garbulky]

audiobill
One thing they fail to mention with R2R DACs when they talk about "approximations" in delta sigma. If you actually look at the delivery of the signal, the "approximations" can look to be more accurate and linear in output vs the R2R in low levels.
The R2R's have problems with low level linearity and accuracy. So that one voltage per bit may not be as accurate in low level signals as one may imagine.

Here is the $2300 Ygdrassil which uses four 20 bit true multibit R2R DACs together to get 21 bits of equivalent number of bits. (very impressive)
Here's where they judge using linearity how much it varies out from an ideal straight line.
Now I'd like to mention that I don't think that there is a formula that says "this is what counts as 16 bits of resolution". So I take issue with that declaration on the graph. But what it does do is basically say after 16 bits this is when variation starts outside of +/- 0.1 db/


Here is the $250 Topping d50 usb delta sigma dac


As you can see this delta sigma "approximation" is way more linear in to much finer resolution.

However, below 50 to 60 db the noise floor is essentially silent in regular music. So I don't know if these stunning low level linearities really make a difference in listening.

[KeithL]

I've GOT to insert my two-cents here.....

This article is actually quite informative, but it also has quite a few places where it jumbles together facts, conclusions, and opinions - somewhat haphazardly.
For example, the author clearly LIKES R2R DACs, and so he considers them to be "beautifully simple and effective" except for "...a few complexities..." and "... the finer points of the circuits...".
The problem is that what this particular author sees as "trivial complexities and finer points" many other engineers see as MAJOR ISSUES AND DRAWBACKS.
And he also obviously considers some things as "critical for audiophiles" that many audiophiles don't care about at all (so they're simply critical in his personal opinion).

He also clearly doesn't especially LIKE Delta-Sigma DACs......
Even though he never actually says anything specifically bad about them (other than that they are cheap to manufacture, have obvious benefits, and are clearly here to stay).

He also uses a LOT of "spin" in the way he phrases some things....
For example, to paraphrase, he says "it's easier to throw a million gates at the problem than to design a properly implemented R2R ladder DAC".
From the way that's phrased, even though he didn't say it, YOU are clearly supposed to infer that "throwing a million gates at the problem" is a shoddier solution than "a properly implemented R2R DAC".
And you are obviously supposed to believe that, if the DAC manufacturers have figured out how to give you what seems to be a better product for a lot less money, there must be some reason why an audiophile obviously WOULDN'T want to own it. 
He sort of glosses over the fact that the solution that involves "throwing a bunch of cheap gates at the problem" actually DOES A BETTER JOB according to all the metrics that most people consider important.
(And, if he's got an argument for why that isn't true, then he should tell us what it is....)

The he sort of jumps from there to: "For an audiophile level DAC this (an R2R DAC) is correct strategy. END OF STORY."
The problem is that he hasn't actually explained WHY this should be true.
He says "we should prefer R2R DACs for obvious reasons" - yet he doesn't give us those reasons.

And he suggests that "if you own a high-end SET tube amp you should prefer an R2R DAC" - yet I don't see a specific reason mentioned there either (and you all probably know how I feel about SET tube amps).

However, once you get past the assumptions and unsubstantiated statements.... all he really says is that, IN HIS OPINION, R2R DACS SOUND BETTER.

(And, yes, his "informed" point of view is party informed by the fact that he's promoting a non-oversampling R2R DAC product - not a big surprise).

Apr 18, 2018 13:33:55 GMT -5 audiobill said:

Before we get too lost, here's another informed point of view:

Things to be aware of in choosing an Audiophile Level DAC

It’s a complicated world out there in the land of digital audio with cheaper and cheaper solutions coming out on the market with higher and higher sampling rates! Many manufacturers are offering very high priced DAC’s and others have offerings that cost next to nothing! – Some DAC’s come in large rack mount cases – others are part of the USB connector!– There is a lot of confusion out there about sampling rates, dither , jitter, oversampling, digital filters, quantization errors, bit widths, USB, SPDIF, 24/96, 192, 384, SACD……. and where is it all going.

Read on to understand what is really going on in digital Audio. Hopefully this article will help you to make an informed decision when it comes to buying a DAC and understand more about the current DAC you have. Lets start by discussing some major points first and get into the detail later!

MAJOR POINT #1

“The actual conversion process from digital to analog is very critical” and this is what is at the heart of any DAC – Now listen closely, there are two major methods of accomplishing this conversion from digital information stored on a computer or a disk – the original method and the one currenly used by xxxx today along with a few other manufacturer’s is by “hardwiring” the digital numbers through an electrical circuit made up of resistors, voltages & currents and outputting a voltage driven directly from the input – as the digital numbers change at the input of our circuit the output voltage changes. This method is referred to as a Resistor Ladder DAC or R-2R.

Its really beautifully simple and effective:
◾No filters in the digital domain
◾No filters in the audio domain
◾No altering or modifying the signal
◾Just pure conversion

Of course there are complexities in the precision resistors and the finer points of the circuits but bottom line is it’s a direct HARDWIRING of digital numbers to actual output Voltage. For an audiophile level DAC this is correct strategy. END OF STORY.

MAJOR POINT #2

The second method of performing this critical conversion is through computer horse power – basically the computer hardware reads in the first bit of data and continues to read in bits and analyze via complex mathematical algorithms, digital signal processing , quantization , dither , noise injection & feedback and the computer will determine a voltage that has been carefully computationally arrived at. This is referred to as a DELTA-SIGMA.

You may think this is not a very good process compared to good old fashioned hardwired R-2R, but the truth the computer processing method has many advantages:
◾It is inexpensive compared to its R-2R counterpart & very convenient for many musical applications : Internet radio, Home theater, 5 or 7 channel movies with surround sound, Mp3, iPods, digital video cameras, even SACD uses this method.
◾It allows for great versatility, filtering, sound shaping, room correction etc – incredible flexibility and for 99.5% of the music world DELTA-SIGMA is a great way to go.
◾Its cheap ( Seen the cost of many 24/96 and 24/192 and 384Khz sampling DAC’s) – ( DELTA-SIGMA DAC’s love high resolution formats because its easy to process them compared to R-2R). Its easy to manufacture chips with tons of processing, thus SIGMA DELTA dacs are here to stay!

Even companies like Analog Devices that made a range of R-2R DAC chips in the 90’s now have a complete suite of DELTA-SIGMA chips with tons of features. Basically all home theater gear, CD players, video cameras etc.etc. will use DELTA-SIGMA processing. Virtually all modern DAC’s are DELTA-SIGMA based to handle the 24/192 supposedly high resolution formats.

So why did Manufacturers move away from R-2R chips that they developed in the 90’s? Well basically it is an expensive process to create integrated circuits with laser trimmed precision resistors to correctly implement the R-2R circuit. Its much easier to throw a few million gates at the problem! The Analog Devices AD1865 chip used in the Audio Note DAC’s is the highest bit resolution DAC chip that was made with 18 bits and provides 108db of potential dynamic range. So basically when you put a 24/96 signal through an Audio Note DAC you have a potential dynamic range capability of 108db which is approaching twice the dynamic range ever recorded.

So which is the right DAC for you – R-2R or DELTA-SIGMA? For a true audiophile level DAC we highly recommend R-2R!

If you are truly concerned with the very best of 2 channel audio the R-2R DAC’s will clearly outshine the computer processing DELTA-SIGMA dacs for obvious reasons, If you have a high end audio system such as a Single Ended Tube 300B or 2A3 amp or possibly CLASS A push Pull tube amplification , transformer coupled pre-amplifiers, high efficiency speakers, it’s the only way to go.

Listen to a saxophone with an R-2R DAC and you are going to pick up all the subtle nuances associated with this instrument – the attack , the timbre, the decay, these subtle nuances that immediately allow the brain to identify the instrument are absolutely critical to an audiophile.

[audiobill]

Ah, different strokes.

Some of us judge sonics by looking at spec sheets and oscopes, some merely by listening....

[KeithL]

Digital audio data is often handled at 32 bits, or even 64 bits, inside digital audio editing software, and digital editing consoles.
When you're mixing multiple channels, making drastic level changes, and using plugins that apply a lot of numerical calculations to the audio, the extra resolution and dynamic range helps avoid rounding errors and similar issues.
(So you might see 32 bit files used to store audio clips involved in the editing process... but I don't recall ever seeing music DISTRIBUTED in a 32 bit format.)

Note that a 120 dB dynamic range is the range between the sound of your heartbeat in a really quiet room and the sound of a jet plane taking off.

Apr 18, 2018 15:10:10 GMT -5 Casey Leedom said:

Thanks KeithL . I was also looking at the product page for the Asahi Kasei Microdevices AK4490EN which I believe is what you're using in the RMC-1 (and probably for the front three channels of the XMC-1G3 and maybe even the DC-2). Similar to the ESS Sabre DAC example you offered, the AK4490EN is "32 bits" but its Data Sheet shows it as only having a Signal/Noise Ratio of 123dB in Mono Mode.

Also, I've never seen any digital audio file with 32-bit samples. That would e pretty crazy ...

Casey

[KeithL]

I absolutely agree......

HOWEVER, if the author of that article really felt that way, and believed that his customers feel that way as well...... then he SHOULD have said:

"R2R DACs are hard to make, and the specs aren't especially good, but I and many audiophile friends think they sound really great, and we think that's what counts."
"We hope you'll agree - and we encourage you to listen for yourself and judge for yourself. "

However, instead, he attempted to explain in technical terms why R2R DACs are superior - and - to my reckoning - did NOT succeed.
I have to admit that it just bugs me when someone spouts a bunch of technical, and only somewhat accurate, gobblydegook....
And then, after not providing any evidence to support his case, acts as if he expects you to believe that he had proven it.
It makes me feel as if he's just hoping that, if he can't find words to actually support his claim, he expects you to trust him on the volume of words alone....
Or, more likely, he hopes that you won't really understand the nuances of what he's saying, and so you'll just assume he must know what he's talking about.
He followed that up in several places by "of course, as an audiophile, you'll obviously want to do it this way"... which is simply a strategy right out of a first semester sales training manual.
("We make our bottled water from rain water, carefully siphoned from the west side of a mountain, by naked virgins, during a full moon..... so OF COURSE it tastes great".)

What I must admit that I find somewhat humorous is that there are legitimate arguments against Delta-Sigma technology.... which I don't personally find especially compelling.... but he missed THOSE entirely.

Apr 18, 2018 16:08:56 GMT -5 audiobill said:

Ah, different strokes.

Some of us judge sonics by looking at spec sheets and oscopes, some merely by listening....

[garbulky]

Apr 18, 2018 16:08:56 GMT -5 audiobill said:

Ah, different strokes.

Some of us judge sonics by looking at spec sheets and oscopes, some merely by listening....

For me, at the end of the day the listening is what's important. After finding out just how low of a dynamic range you actually need for regular content, I realized almost all DACs meet it, including the ones that sound not very god. So there's got to be something going on that isn't adequately described by the popular measurements, or at least properly interpreted.
To be clear, I believe that we can measure everything that's coming out of a signal. However, whether we are measuring the right thing or intepreting these measurements in a way that describe how we hear subtle differences is an entirely different question in my book.

People gloss over it. The human mind is too biased. So there's no point in going to the distance of quantifying our perceptions. It's way less accurate than a measurement tool (true, especially true in some cases).

But it has the ability to build a soundstage and describe it. It has the ability to create emotions and thoughts based on the quality and palpability of the reproduced sound. Some of which can sound surprisingly different in non level matched non double blind tests. It can build a remarkable 3 dimensional auditory model i nyour head and relate it to a visual picture. Instruments can't do that. So there's something missing here. Our tests and our instruments cannot quanitfy these sensations - at least not in a highly detailed manner. So there's more work that needs to be done here.

Because our day to day experience is not level matched and it's not double blind and neither is it done by looking at tones on an osciloscope. So at the end of the day science needs to figure out a way to build standards and transfer interpretations to subtle subjective experiences even if it's extremely complex and hard to do.

[garbulky]

A day in my life.
Me telling my wife: "I just got a new CD player. But it's actually from the late 80's."
Wife: ""What? Why?"

Me" "Well have you ever heard of a multibit dac?"

Wife: "No."

Me: Well my current dac is a delta sigma DAC. And these are usually 1 to 5 bit DACs which approximate a 16 bit signal from CD's and modulate it using all kinds of processing to produce the sound. But back then they did not have that technology so this muiltibit dac is a true 18 bit DAC. Which means that every bit of the CD signal is produced by itself using precision laser trimmed resistors. These types of DACs are no longer really being made except by some small companies.

Back in the day these multibit units were very expensive to make because the resistors had to precisely matched. And this one is matched using lasers! The current delta sigma units are much cheaper to make and produce greater accuracy for a couple of bucks. But back then multibit dacs were very expensive. The difference is that the process of producing the signal was more simple and each bit just gets produced. So some say the sound might be better! How cool is that?

Wife: "Please stop."

[KeithL]

Let me just clarify a few things here... just for the rest of the crowd.

In the early days ALL DACS were "multi-bit DACs" - it was the original DAC technology and all the alternatives came along later.
Back then things like 18 bit DACs were a big deal because the ones commonly used in many CD players were barely able to handle 16 bits accurately (many were only really "good" to 14 or 15 bits).

Commercial multi-bit DAC chips are in fact still made; they're just not used in audio equipment much any more.
They are more commonly used in measurement equipment these days - including medical and telemetry applications - where high speed is more important than accuracy.
Likewise, as far as I know, all video DACs are multi-bit types.
That's why you'll see companies like Schiit using DAC chips intended for other purposes.... and then expending a lot of effort tweaking them to work properly for audio.

Incidentally, LASER trimming is quite common these days; it's just that, much like "fuel injection", it doesn't impress people as much as it used to.

And, finally, I'm going to point out something about marketing.
It's funny how many people whose hearing seems to be governed more by what they read than what they hear.

Marketing literature for multi-bit DACs insists on reminding us that "Delta-Sigma DACs reconstruct an approximation of the signal" while "multi-bit DACs just reconstruct the signal directly".
They try to give you an IMPRESSION..... that "Delta-Sigma DACs only give you an approximation of what they should - while multi-bit DACs reproduce it exactly using precision adjusted parts".
While this is technically true in one sense it is also quite misleading.
Unless one of your ancestors was R2D2, neither of us listens to that digital signal directly; we're all hearing "an analog approximation of what the numbers say".
In fact, Delta-Sigma DACs replaced multi-bit DACs in the first place because they did a BETTER job of accurately "approximating" the original signal.
Again, in technical terms, Delta-Sigma DACs deliver ridiculously better resolution and level accuracy - while they do in fact tend to sacrifice a little bit in terms of "time domain accuracy".
HOWEVER, you need to put that last statement in its proper perspective......
Assuming we had a DAC which "added five samples of time smear to the signal coming from a CD" - we would be talking about a little over one tenth of one millisecond.
This is FAR less "time error" than even the best microphones, or rooms, or speakers can even hope to approach.
And those pictures showing all those cycles of ringing on a super-short duration impulse.....
Before they worry you too much, you need to go back and see what the output of a SPEAKER looks like if you feed it an impulse, or a phono cartridge.....

And, as for the argument that "multi-bit DACs are very simple" while Delta-Sigma DACs "subject your audio to all sorts of complicated processing"....
That is absolutely true.... however, since all you get to listen to is the analog that comes out the end, it doesn't MATTER how complicated the stuff is that goes on in-between.
Audiophiles have a habit of assuming that "simpler is always better".
In fact, simpler is SOMETIMES better, or perhaps simpler is OFTEN better, but not always.
Anyone who doesn't believe me should try doing their taxes on an abacus.
Likewise, I suspect that no halfway intelligent horse would let you accidentally "drive" it over a cliff... which makes them "smarter" than the most advanced smart car we have so far.
(But most of us still drive cars.... because the benefits FAR outweigh the limitations.)

I give Garbulky enormous credit for stating the case fairly.... and not leaving us trying to assume that "simpler is better" or "all modern technology is inferior to the old stuff".

I should also mention one other thing.... for certain fans of non-oversampling DACs.
The SINGLE BIGGEST drawback of non-oversampling DACs is the compromise they're forced to make in their reconstruction filter.
(It's almost impossible to make a filter that's sharp enough to eliminate aliasing without rolling off the high end.)
However, this is only really a problem with 44.1k files..... and much less so at higher sample rates....
So non-oversampling DACs tend to sound much better with 24/96k and higher sample rate files.

Apr 18, 2018 21:56:24 GMT -5 garbulky said:

A day in my life.
Me telling my wife: "I just got a new CD player. But it's actually from the late 80's."
Wife: ""What? Why?"

Me" "Well have you ever heard of a multibit dac?"

Wife: "No."

Me: Well my current dac is a delta sigma DAC. And these are usually 1 to 5 bit DACs which approximate a 16 bit signal from CD's and modulate it using all kinds of processing to produce the sound. But back then they did not have that technology so this muiltibit dac is a true 18 bit DAC. Which means that every bit of the CD signal is produced by itself using precision laser trimmed resistors. These types of DACs are no longer really being made except by some small companies.

Back in the day these multibit units were very expensive to make because the resistors had to precisely matched. And this one is matched using lasers! The current delta sigma units are much cheaper to make and produce greater accuracy for a couple of bucks. But back then multibit dacs were very expensive. The difference is that the process of producing the signal was more simple and each bit just gets produced. So some say the sound might be better! How cool is that?

Wife: "Please stop."