Do all DACs sound more or less the same?

[gmgraves]

I think you might be surprised at the prices of the chips in many audio DACs. Other than something quite specialized, like the chips in the Phasure, I think the delta-sigma chips found in most DACs are fairly close to that $5 figure you named, if my recollection of looking at price sheets a year or two ago is reasonably accurate. (I would expect the DAC chip in a smartphone to either be less than that, or if it is the same or more than the $5 you mentioned, it might be due to size or other specialty considerations.)

 

You are correct that most hi-fi audio quality DACs are around $5 or so in manufacturing quantities. There are exceptions, of course. I think some SabreDACs from ESS are more, and perhaps there are others, but in general, Integrated Circuit DACs are commodities. Now some companies may bin their DACs, testing them for linearity and sorting the results according to quality. There is always some variability in the performance of ICs in that some parameters might vary due to process from chip to chip and wafer to wafer. By binning the parts, the manufacturer can charge more for parts that achieve higher linearity. Most 24-bit DACs achieve 17-bit linearity. Occasionally some test-out at 18 and even 19-bit linearity. These can be sold to high-end manufacturers for more money than can the run of the mill yield.

 

But, I doubt seriously if the hi-fi audio DAC in a smart phone or an iPod is even a separate component, it is probably part of a much more dense Very Large Integrated Circuit (VLSI) chip that also contains a micro processor, logic functions, Firmware for the OS, the headphone/line driver amp, and the I/O for the memory, and possibly even the memory itself, etc. OTOH, in the telephone part of the smart phone, the DAC/ADC that encodes/decodes the voice for telecom communication is probably also part of a VLSI that handles all of the cell phone's operation; the dialing, the transmitter/receiver as well as the mic preamp and the speaker driver amp and the CoDec. They may even be that all of the functions mentioned above, are in the same chip, in fact I'd be much more surprised if they were not. Back in the 1980's, the semiconductor firm I worked for sold CoDec chips to the cell phone industry. These chips contained not only the 3 and 4-bit DAC/ADC but the compression/decompression hardware and firmware. We sold them in quantity for about a quarter of a dollar each. ICs are made pretty cheap by quantity and as the manufacturers get better at making them, they get more and more good chips per wafer which drives the cost down even more.

[Snowmonkey]

Do smartphones use the same DAC for telephony that they do for rendering music? The music from my phone isn't that bad.

[gmgraves]

Do smartphones use the same DAC for telephony that they do for rendering music? The music from my phone isn't that bad.

 

It's possible, I guess, but I would't think so The telecom and computer sections of a smart phone would, of course, share much of the same circuitry - for instance the microprocessor in a smart phone probably takes care of the dialing and storage/retrieval of phone numbers, but I suspect that the RF section of the phone would be separate. Telecom only needs a minimal (3 or 4 bits) DAC but it also needs an ADC, while a 16-bit DAC could be used for those processes, I don't think they do. One way to find out might be a simple test: does your phone allow you to listen to music and talk on the phone at the same time?

[Jud]

But, I doubt seriously if the hi-fi audio DAC in a smart phone or an iPod is even a separate component, it is probably part of a much more dense Very Large Integrated Circuit (VLSI) chip that also contains a micro processor, logic functions, Firmware for the OS, the headphone/line driver amp, and the I/O for the memory, and possibly even the memory itself, etc.

 

As I read this I thought of course that must be right - why didn't I think of that?

 

Doing a little research, it turns out the chip is perhaps somewhere in between a full system-on-chip and a discreet DAC chip. It's a Cirrus Logic chip that handles multiple audio functions - class D amplifier, DAC, etc. See

[James1776]

OK, I think you missed my point here. The telephone system does not use 16-bit/44.1 DACs. They use 3 or 4 bit DACs at about a 4 KHz sampling rate. So, comparing a telephone DAC with one in a pad or iPod or a CD player is like comparing the amp/speaker in a speaker phone with a pair of Krell mono-blocks with a pair of Wilson Alexandria XLF speakers. In other words it's not just an apples-to-oranges comparison, you're trying to make here, It's an apples-to-brussles sprouts comparison. Except for the fact that both DACs in question convert some form of digital audio to analog audio, they have nothing in common.

 

Now I am learning something. If the DAC in a phone is as limited as you note, then they are downsampling every file better than an MP3 before it gets to the earphone or audio out. That might explain why USB Audio Player pro outputting a native 24/96 file to an outboard DAC sounds so much better than the dross produced by iPod's and the like.

[gmgraves]

Now I am learning something. If the DAC in a phone is as limited as you note, then they are downsampling every file better than an MP3 before it gets to the earphone or audio out. That might explain why USB Audio Player pro outputting a native 24/96 file to an outboard DAC sounds so much better than the dross produced by iPod's and the like.

 

 

If I understand correctly, what you wrote, above, then you still aren't following me. The music DAC in a smart phone and the telephone DAC/ADC used to encode and decode the voice communications of said phone, are separate functions, and phone calls are most likely taken care of by a dedicated bit of circuitry, and not the same DAC that allows one to play music. If they do use the same DAC for both functions (possible, but I don't see the advantage), then that single DAC would be totally re-configured by the firmware and the microprocessor in the phone when a telephone call is made or received.

[James1776]

All I care about is the DAC in the phone used for music. To me they sound as good as any iPOD. My question is more basic. It THAT DAC a standard 16/44.1 DAC capable of CD quality sound (though the phones speakers are crap) or is all the music on the phone downsamped to something like MP3 quality? USB Audio Player Pro bypasses the phones DAC and rotes the file to an outboard DAC letting one store and play High Rez Music on the device.

[gmgraves]

As I read this I thought of course that must be right - why didn't I think of that?

 

Doing a little research, it turns out the chip is perhaps somewhere in between a full system-on-chip and a discreet DAC chip. It's a Cirrus Logic chip that handles multiple audio functions - class D amplifier, DAC, etc. See iPhone 5 Teardown - iFixit

​

 

Thanks for the reference. Looks like the bulk of the telephone functions are handled by a Qualcomm MDM9615M. Since It's a G4 phone and needs to simultaneously handle voice and data, this makes sense. Also, it looks as if Apple has chosen not to use the main system-on-a-chip VLSI with the ARM processor, etc., in it to house the audio CoDec, but instead uses a dedicated chip (Apple 338S1077 Cirrus audio codec) for those functions which include MP3 and ALC decompression as well as at least a 16-bit/44.1 KHz DAC. I've never owned a smart phone, so I don't pretend to know what it would use an ADC for other than for making phone calls (does it have an audio note-taking feature perhaps), so perhaps Apple does use the same audio chip for telephone voice encoding and decoding as well as music playback. I wish I could see a block diagram of the phone! I was very surprised by the number of components in this thing. I would have expected a much higher level of integration by now. You know, one big system-on-a-chip for everything except the telecom part and another chip for that function, and perhaps a separate RF module for the cell radio.

Very interesting.

[esldude]

​

 

Thanks for the reference. Looks like the bulk of the telephone functions are handled by a Qualcomm MDM9615M. Since It's a G4 phone and needs to simultaneously handle voice and data, this makes sense. Also, it looks as if Apple has chosen not to use the main system-on-a-chip VLSI with the ARM processor, etc., in it to house the audio CoDec, but instead uses a dedicated chip (Apple 338S1077 Cirrus audio codec) for those functions which include MP3 and ALC decompression as well as at least a 16-bit/44.1 KHz DAC. I've never owned a smart phone, so I don't pretend to know what it would use an ADC for other than for making phone calls (does it have an audio note-taking feature perhaps), so perhaps Apple does use the same audio chip for telephone voice encoding and decoding as well as music playback. I wish I could see a block diagram of the phone! I was very surprised by the number of components in this thing. I would have expected a much higher level of integration by now. You know, one big system-on-a-chip for everything except the telecom part and another chip for that function, and perhaps a separate RF module for the cell radio.

Very interesting.

 

Pretty sure they use the same ADC/DAC. My phone, a Galaxy S3, you can plug in headphones to listen to music. Quality is quite good. If the phone rings, you hear it over the headphones and can simply answer hearing audio over the phones. You also can get headphones with a mic. Sound quality is lots better that way too even with the cell network being simplex and limited to maybe 4500 hz for voice.

 

Why the ADC? Just as you said, notetaking, voice searches, and recording. Mine will do 48/16 and yes downloading the files they are 48. With some apps I can choose between 48, 44.1, 22.05, 16, 11.025, and 8 khz sample rates, and select 8 or 16 bit in each. I can do that to wav, mp3 and I think another odd codec .3gp which is audio for a video format.. I actually recorded a music concert with it awhile back. Was better than I expected, but mono. I have seen devices to attach to such phones with decent little mikes mounted something like portable recorders for stereo. Apparently even the mic in the phone, though limited by the cell network internally works pretty well to 8khz and has some response above that.

 

You can go to Archimago's site and see measurements of a Nexus 5 phone and Nexus 7 tablet. It looks like they limit it to 44/16. So it does vary from phone to phone.

 

Archimago's Musings: MEASUREMENTS: Google Nexus 5 and Nexus 7 (2013) audio quality...

[Jud]

All I care about is the DAC in the phone used for music. To me they sound as good as any iPOD. My question is more basic. It THAT DAC a standard 16/44.1 DAC capable of CD quality sound (though the phones speakers are crap) or is all the music on the phone downsamped to something like MP3 quality? USB Audio Player Pro bypasses the phones DAC and rotes the file to an outboard DAC letting one store and play High Rez Music on the device.

 

Best I can find thru University of Google at short notice is that the iPhone's internal DAC will do 24/48 resolution. Not sure at this point whether 44.1 is left alone or resampled to 48.

[gmgraves]

All I care about is the DAC in the phone used for music. To me they sound as good as any iPOD. My question is more basic. It THAT DAC a standard 16/44.1 DAC capable of CD quality sound (though the phones speakers are crap) or is all the music on the phone downsamped to something like MP3 quality? USB Audio Player Pro bypasses the phones DAC and rotes the file to an outboard DAC letting one store and play High Rez Music on the device.

 

Well, yes, of course the music DACs in iPhones and iPods the same quality. In fact, I suspect that the audio CoDec in both iPhones and iPods are the same chip. They might be packaged differently in different model iPods, but I suspect that they are essentially the same. Have you not fed an iPhone, iPod, iPad or similar devices into your stereo system for a listen? While not what I would call high-end audio, they are certainly capable of true 16-bit/44.1 KHz quality. I mean it would be unrealistic to expect a device like this (costing, in most cases less than $500) to have the same level of digital audio performance as a $1000+ stand-alone DAC. In fact, I'll guarantee that the $130 AudioQuest DragonFly DAC is a much better sounding device than any "i-type" music-playing device, regardless of manufacturer. (Exception to the rule - the Astell & Kern AK120. Of course, at $1300 for just a pocket-able music player, It had better have the "right stuff" and from my brief listen at a recent Hi-Fi show through a pair of Audeze headphones, it certainly seems to with its dual Wolfson 24-bit high-end DACs which will play hi-res audio as well as compressed (lossy and lossless) standard definition audio!)

[Snowmonkey]

As I understand it, the Samsung Galaxy S3, S4 and Note models incorporate Wolfson chips.

[gmgraves]

As I understand it, the Samsung Galaxy S3, S4 and Note models incorporate Wolfson chips.

 

I don't doubt it, but Wolfson, like Burr-Brown Division, ESS, Analog Devices, etc., Wolfson makes many different DACs for many different purposes and at all quality and price points. The Astell &Kern AK120 uses dual Wolfson WM8740 DACs which costs about $13 each in quantity and are sold as being "audiophile quality". Also, keep in mind that all of these DAC manufacturers have DAC models that are available as "library" parts for ASIC designs using design tools from Mentor Graphics, Synopsys, etc., as well. That means that a custom VLSIC these days can contain a microprocessor from, say, ARM or Intel, on-board memory from Samsung, an ESS SabreDAC, a high-end audio amplifier chip from TI, a USB receiver from another source, etc., all on the same piece of silicon. Not that these disparate chips are purchased from these individual manufacturers and then somehow "affixed" to the customer's substrate and then packaged all in the same encapsulating package. It doesn't work that way. What the ASIC designer is buying is the design for these chips, which are all pre-defined as a circuit element and optimized for the process and design tools being used to design the "system-on-a-chip" VLSI ASIC in question. These tools are so easy to use, that most hardware design companies can do it in-house. They lay-out the chip using the design tool package running on a regular PC, run simulation on the finished part to guarantee functionality and that the part meets performance expectations, then the design is sent electronically to a "foundry" company (who actually manufacture the chips) and when finished, the parts are shipped to the company that designed them.

[James1776]

Well, yes, of course the music DACs in iPhones and iPods the same quality. In fact, I suspect that the audio CoDec in both iPhones and iPods are the same chip. They might be packaged differently in different model iPods, but I suspect that they are essentially the same. Have you not fed an iPhone, iPod, iPad or similar devices into your stereo system for a listen? While not what I would call high-end audio, they are certainly capable of true 16-bit/44.1 KHz quality. I mean it would be unrealistic to expect a device like this (costing, in most cases less than $500) to have the same level of digital audio performance as a $1000+ stand-alone DAC. In fact, I'll guarantee that the $130 AudioQuest DragonFly DAC is a much better sounding device than any "i-type" music-playing device, regardless of manufacturer. (Exception to the rule - the Astell & Kern AK120. Of course, at $1300 for just a pocket-able music player, It had better have the "right stuff" and from my brief listen at a recent Hi-Fi show through a pair of Audeze headphones, it certainly seems to with its dual Wolfson 24-bit high-end DACs which will play hi-res audio as well as compressed (lossy and lossless) standard definition audio!)

 

I must not have been clear...not 'i' toys but the universe of android and windows phones. The IPods and iPhone are a tad behind the times in technology.

[Paul R]

I must not have been clear...not 'i' toys but the universe of android and windows phones. The IPods and iPhone are a tad behind the times in technology.

 

Perhaps, but they work and all of 'em can stream 24/96 digital outputs. Thy wound pretty good at CD resolution too.

[Jud]

 

Perhaps, but they work and all of 'em can stream 24/96 digital outputs. Thy wound pretty good at CD resolution too.

 

iPhone does 24/96 now?

[James1776]

Not that I am aware of....My phone will send a Hi-Rez file to an external DAC...I am not aware that an iPhone does the same. I know it does not for the same bucks...

[Snowmonkey]

@James1776: Wow, thank for the heads-up on USB Audio Player Pro. It works a treat on my Samsung Galaxy 2, coupled with Dragonfly DAC.

[esldude]

@James1776: Wow, thank for the heads-up on USB Audio Player Pro. It works a treat on my Samsung Galaxy 2, coupled with Dragonfly DAC.

 

Oh yeah, nice program to have. Works quite well.

[Mike Mcsweeney]

I mean it would be unrealistic to expect a device like this (costing, in most cases less than $500) to have the same level of digital audio performance as a $1000+ stand-alone DAC. In fact, I'll guarantee that the $130 AudioQuest DragonFly DAC is a much better sounding device than any "i-type" music-playing device, regardless of manufacturer.

 

Stereophile magazine in their review said that the Dragonfly DAC is as good or better than their reference exernal DAC costing $900. Wouldn't that suggest that there isn't any reason to buy an external DAC costing in the $200-$1000 range?

And then, when i compare my dragonfly playing the same song burned to a CD and play from my BDP-S790 Blu Ray player, the BDP-S790 sounds FAR BETTER than the Dragonfly. It seems to me, when you put both those facts together, there really isn't any need to buy any external DAC for $1000 or less??

[christopher3393]

Hi Mike,

 

I'm sure you've posted this elsewhere, but would you mind giving more details about your system? I have no doubt that this is what you're hearing, but my own opinion is that these things can be surprisingly system dependent, and perhaps not easily generalizable. Add to that what difference it might make if you had the second iteration of the Dragonfly, and maybe things get more interesting. What do you think?

[Jud]

 

Stereophile magazine in their review said that the Dragonfly DAC is as good or better than their reference exernal DAC costing $900. Wouldn't that suggest that there isn't any reason to buy an external DAC costing in the $200-$1000 range?

And then, when i compare my dragonfly playing the same song burned to a CD and play from my BDP-S790 Blu Ray player, the BDP-S790 sounds FAR BETTER than the Dragonfly. It seems to me, when you put both those facts together, there really isn't any need to buy any external DAC for $1000 or less??

 

Mike, it means you would not need to buy an external DAC, but for others the situation may be different. For me, a $450 external DAC (Schiit Bifrost) fed from a laptop quite handily beat the sound from my OPPO and from an older $2000 DAC (Theta) fed by the OPPO. So systems and individuals' experiences can and do vary widely.

[Allan F]

Stereophile magazine in their review said that the Dragonfly DAC is as good or better than their reference exernal DAC costing $900. Wouldn't that suggest that there isn't any reason to buy an external DAC costing in the $200-$1000 range?

And then, when i compare my dragonfly playing the same song burned to a CD and play from my BDP-S790 Blu Ray player, the BDP-S790 sounds FAR BETTER than the Dragonfly. It seems to me, when you put both those facts together, there really isn't any need to buy any external DAC for $1000 or less??

 

There may not be any need for you to buy an external DAC, but it's possible that you would obtain better sound quality with one. Not to be picky but, more precisely, Art Dudley wrote that the Dragonfly sounded as good or better than his $900 reference. That doesn't necessarily extrapolate to sounding better than all $200 to $1000 external DACs. There are a number of other DACs in that price range, many of which are newer designs than Dudley's reference, and some are more than likely to provide better sound quality. As others have pointed out, a lot depends on the associated equipment in your system.

 

The only conclusive way to determine the issue for yourself would be to listen to other DACs. OTOH, if you enjoy and are satisfied with the sound that your current system is providing, there may no reason for you to buy an external DAC at this time. One of the enjoyments many of us derive from this hobby is upgrading our systems over time as finances permit.

[Mike Mcsweeney]

Hi Mike,

 

I'm sure you've posted this elsewhere, but would you mind giving more details about your system? I have no doubt that this is what you're hearing, but my own opinion is that these things can be surprisingly system dependent, and perhaps not easily generalizable. Add to that what difference it might make if you had the second iteration of the Dragonfly, and maybe things get more interesting. What do you think?

 

My points are this:

 

1. If stereophile magazine (whom i understand is one the most trusted magazines and back up what they say with objective data from lab tests) said the first generation dragonfly is better than their $900 reference DAC, is there any real reason to buy an external DAC less than $1000 other than for "subjective" reasons?

 

2. To me, the DAC should just do the best job of reproducing the original analog waveform, and if one dac can't say they are more "accurate" than another, i don't want to pay for an algorithm that appeals to "subjective" reasoning...that is what i pick speakers for.

 

3. Granted, right now, i don't have high-end amp&speakers (old marantz & PSB image B5 speakers), so for my system, lets stick with these, to eliminate comparisons of brands of amps, speakers, etc, as being relevant.

 

a) DLNA->Sony BDP-S790 analog out->high end headphones

b) Lenovo I5 32GB PC -> USB-Dragonfly-monster cable -> High end headphones

 

 

and just for the record i have tried running dragonfly to analog in on about 5 different recievers (including a returned sony STR-DN1040), and 6 different sets of speakers. The external equipment didn't make a difference in the fact that the analog out of the BDP-S790 sounded MUCH better than the dragonfly, and stereophile says the dragonfly is as good or better than their reference $900 ext DAC. I would love for someone to provide objective data that would substantiate buying a DAC for less than today's $1000 DAC, when you can find a comparable or better DAC in a mid-level bluray. Try better speakers instead?

[gmgraves]

I must not have been clear...not 'i' toys but the universe of android and windows phones. The IPods and iPhone are a tad behind the times in technology.

 

Doesn't matter. The main difference between the Apple "iDevices" and those similar devices (to the iPhone) made to run Android and Microsoft OS's is the OS itself. Apple iOS is MUCH better than than Android (Google just does not get software design, IMHO) and it goes without saying that iOS is better than Windows - anything would be - even Android (again, IMHO). Hardware and function-wise, all of these phones and pads have similar capabilities and differ only in the details. But to your main point. The DACs in all of these devices are probably quite similar because their published audio specs are quite similar.