Will an external DAC sound better/different than my AVR DAC?

[knownothing]

In an AVR, DSP is (obviously) done in the digital domain. So if you input an analogue signal and want to do ANY DSP (Including Crossover to sub which is done by the DSP), then the analogue signal is converted to digital, DSPd, and then converted back to analogue using the AVR DAC.

LFE are part of a digital signal stream - so can only be extracted in the digital processing, and sent to a SUB via the AVR DAC.

There is no DSP done in the amp section of an AVR. DSP is DIGITAL signal processing.

So you are saying that all AVRs have ADCs that convert analogue inputs to digital, process the signal in two of multiple channels, and then convert that to analog using the internal DACs of the AVR?

kn

 

[antcollinet]

es, if AVRs actually have ADC capabilities for incoming analog signals, which may exist in some models, but I don’t think I’ve ever seen one advertised with that capability.

They all can do that. That capability is not advertised (because the user doesn't care) but it is required for some of the features they do care about, such as the ability to do DSP on analogue inputs.

EG For crossover functionality, or multi speaker stereo effects (upmixing), room correction, etc etc.

 

[knownothing]

Every Denon/Marantz AVR has this feature.

How do you use a sub with the AVR in direct mode? Doesn't that bypass all DSP including sub/bass management?

Yes. I use high level output from the speaker terminals in this mode. On my receiver, “Analogue Stereo” also outputs bass via high level outputs, but “Analogue Stereo + Dolby Pro Logic” or any of the other 5.1 surround settings outputs bass via the low level RCA connection.

 

[antcollinet]

So you are saying that all AVRs have ADCs that convert analogue inputs to digital, process the signal in two of multiple channels, and then convert that to analog using the internal DACs of the AVR?

kn

Yes - if they do any sort of DSP on the analogue inputs - which all of the ones I've ever looked at can do. In fact the WHOLE POINT of "pure direct" or whatever naming is used by the different brands - is to avoid the ADC DSP DAC pipeline.

 

[Nutul]

So you are saying that all AVRs have ADCs that convert analogue inputs to digital, process the signal in two of multiple channels, and then convert that to analog using the internal DACs of the AVR?

kn

If they do DSP and/or spatial audio enhancements (like your stereo -> 5.1) yes; they MUST have an ADC, as there is no other way to apply such changes to an analog audio signal.

 

[knownothing]

They all can do that. That capability is not advertiesed (because the user doesn't care) but it is required for some of the features they do care about, such as the ability to do DSP on analogue inputs.

EG For crossover functionality, or multi speaker stereo effects (upmixing), room correction, etc etc.

Thanks, see my reply to @SKBubba on my recent user experience. I also get very different results comparing two channel digital in and the internal DAC of AVR compared with analog in from external DACs, and I have clear preference for sound of external DACs. The difference is not subtle.

kn

 

[krabapple]

Yes, if AVRs actually have ADC capabilities for incoming analog signals, which may exist in some models, but I don’t think I’ve ever seen one advertised with that capability.

Modern AVRs generally can and do apply DSP to *all* inputs, including analog. Avoiding that is what the 'Direct'/"Pure Direct' options are for.

see discussion e.g.,

https://www.avsforum.com/threads/denon-x4400h-does-analog-input-gets-converted-to-digital.3286182/

 

[antcollinet]

Thanks, see my reply to @SKBubba on my recent user experience. I also get very different results comparing two channel digital in and the internal DAC of AVR compared with analog in from external DACs, and I have clear preference for sound of external DACs. The difference is not subtle.

kn

If you've not compared blind and accurately level matched you can't rely on your listening "impressions"

 

[tc2007]

Buy the cheapest Topping E30 II or equivalent SMSL etc. from Amazon and if you don't hear a difference or don't like the sound of internal vs external DAC (level matched), then return it.

 

[knownothing]

If you've not compared blind and accurately level matched you can't rely on your listening "impressions"

I have level matched. The difference is not subtle. Based on the measurements of AVRs here, its not surprising that external DACs connected to the AVR via analog inputs in Direct Stereo mode sounds better than using the AVRs internal DAC for 2-channel.

kn

 

[knownothing]

Buy the cheapest Topping E30 II or equivalent SMSL etc. from Amazon and if you don't hear a difference or don't like the sound of internal vs external DAC (level matched), then return it.

The AVR I have sounds fine to me for multiple channel sound for AV purposes when the speakers are set to small with a sub. I have issues using the AVRs internal DACs for careful two channel listening. My external DAC sounds great with the AVR for two channel. I will try one of the DACs you suggest and compare it to my current external DAC out of curiosity.

kn

 

[antcollinet]

I have level matched.

With test tones and a volt meter on the output of the DAC (or amp if you are using an AVR)? Level matching by ear, or using an SPL meter is not accurate enough to control for percptive effects.

I also note you didn't claim blind comparisons. You can do all the level matching in the world, but if you known which device you are listening to, you've not controlled for perceptive biases, and again you comparisons are not useful for any objective assessment. It doesn't matter how "unsubtle" the differences are.

Now of course - if you are using analogue input in direct mode - then there more than likely is a significant audible difference, since you are not doing any of the DSP in that case, compared with whatever DSP is going on with a digital input. However if that sounds worse, it suggests either incapable DSP, or you don't have it set up correctly (or you are using one of the soundfield modes (hall, jazz etc) which sound universally terrible.

If you are not in direct mode, then you are still using the AVR DAC (as well as it's ADC) and there is no benefit (in fact a detriment due the DAC - ADC - DAC pipline). In this case, if the ARV DAC is the bottleneck then more than likely the chain will measure - and probably sound - identical to a digital input.

 

[knownothing]

Warning: this post deals with user experience and not measurements. Reader beware.

@antcollinet the comparison is two channel digital source into my AVR from a disk drive via HDMI or toslink versus digital source into an external DAC (toslink), and analog out of the DAC into my AVR in Direct Stereo which turns off all digital circuits in my AVR and it functions as essentially an integrated amplifier. The AVR I am using is an Arcam AVR400, which is about 12 years old now. The external DACs I have used with it are the Audioquest Dragonfly v1.0, Arcam irDAC (v1), Chord QuteHD and Chord Qutest.

To my ears with extensive listening in two channel only (no processing tricks other than room correction for the internal DAC, not the external) at multiple volume levels and using hundreds of digital recordings at 16/44.1 resolution, there are very clear differences between the internal DAC of the AVR and the external DACs. When using analog in from the external DAC, the subwoofer operates with high level speaker terminal in - and room correction has been dialed in for this setting as best as possible using a sound meter and listening. Any sound leveling was done using a sound meter - not voltages, but these differences described here were consistent over a range of volumes and repeated listening sessions. I have blind tested other listeners with my system, and these perceived differences are consistent and not subtle.

My room is purpose designed with a floated ceiling and one side wall completely covered with books and record albums and a concrete floor with a large throw rug. It is a little dead with respect to high frequencies below 60 db, but above 65db average volume, very little coloration compared to other listening environments I have experienced.

The DAC in the Arcam in two channel sounds “brighter” with high frequencies emphasized to the point of sounding “brittle” and fatiguing at higher volumes. The bass is very clear and tight with no bloat in this setting, quite good. The soundstage is wide and a bit hollowed out in the center. I have run the room correction multiple times and while the AVR sounds terrific to my ears on multiple channel Bluray material, two channel redbook recordings never measured up.

In comparison, the first version of the AQ Dragonfly with a computer as source was not an improvement. The Arcam irDAC was different and better than the internal DAC in key areas. Tone, how “real” voices and instruments sound in the room were improved to my ears. Listening fatigue was reduced, especially at higher volume. The soundstage was collapsed in the center and did not extend at all beyond the front speakers compared to the internal DAC of the AVR. The bass was less precise, and this was the trade off of operating without room correction.

When I upgraded from the irDAC to the Chord QuteHD, the detail retrieval increased, the bass tightened up some (still not as good as using the internal DAC with room correction), and the soundstage opened up considerably. Unlike the internal DAC with room correction which pushed the sound to the left and right extremes, the QuteHD precisely positioned instruments and sounds across the soundstage. Impressive feat.

None of these external DACs described so far have been measured by ASR, nor has the AVR400, although the performance of some newer Arcam AVRs have been reviewed by Amir, and the weaknesses of their digital sections in two channel and the generally good performance of their amplifier sections have been reported on this site.

Switching to the Chord Qutest (which has been measured by Amir) there was a slight improvement in all aspects of reproduction over the Chord QuteHD. I noticed slightly more detail retrieval (on headphones at least), more precise articulation of soundstage and cleaner bass. The improvement to my ears between the Qutest and the QuteHD was smaller than the difference between the QuteHD and the irDAC. The Qutest measured well in Amir’s test and has the best overall reproduction of redbook quality recordings of any DAC I have used to date.

I have not yet tried in my system one of the inexpensive Chinese DACs that measure extraordinarily well and compared that to the Chord Qutest. I might try one or several out of curiosity. My conclusion from all of this, based on Amir’s testing and my personal experience I would not assume you are getting highest fidelity reproduction in two channels from digital sources using an AVR with the internal DACs. But YMMV.

kn

 

[antcollinet]

Warning: this post deals with user experience and not measurements. Reader beware.

@antcollinet the comparison is two channel digital source into my AVR from a disk drive via HDMI or toslink versus digital source into an external DAC (toslink), and analog out of the DAC into my AVR in Direct Stereo which turns off all digital circuits in my AVR and it functions as essentially an integrated amplifier. The AVR I am using is an Arcam AVR400, which is about 12 years old now. The external DACs I have used with it are the Audioquest Dragonfly v1.0, Arcam irDAC (v1), Chord QuteHD and Chord Qutest.

To my ears with extensive listening in two channel only (no processing tricks other than room correction for the internal DAC, not the external) at multiple volume levels and using hundreds of digital recordings at 16/44.1 resolution, there are very clear differences between the internal DAC of the AVR and the external DACs. When using analog in from the external DAC, the subwoofer operates with high level speaker terminal in - and room correction has been dialed in for this setting as best as possible using a sound meter and listening. Any sound leveling was done using a sound meter - not voltages, but these differences described here were consistent over a range of volumes and repeated listening sessions. I have blind tested other listeners with my system, and these perceived differences are consistent and not subtle.

My room is purpose designed with a floated ceiling and one side wall completely covered with books and record albums and a concrete floor with a large throw rug. It is a little dead with respect to high frequencies below 60 db, but above 65db average volume, very little coloration compared to other listening environments I have experienced.

The DAC in the Arcam in two channel sounds “brighter” with high frequencies emphasized to the point of sounding “brittle” and fatiguing at higher volumes. The bass is very clear and tight with no bloat in this setting, quite good. The soundstage is wide and a bit hollowed out in the center. I have run the room correction multiple times and while the AVR sounds terrific to my ears on multiple channel Bluray material, two channel redbook recordings never measured up.

In comparison, the first version of the AQ Dragonfly with a computer as source was not an improvement. The Arcam irDAC was different and better than the internal DAC in key areas. Tone, how “real” voices and instruments sound in the room were improved to my ears. Listening fatigue was reduced, especially at higher volume. The soundstage was collapsed in the center and did not extend at all beyond the front speakers compared to the internal DAC of the AVR. The bass was less precise, and this was the trade off of operating without room correction.

When I upgraded from the irDAC to the Chord QuteHD, the detail retrieval increased, the bass tightened up some (still not as good as using the internal DAC with room correction), and the soundstage opened up considerably. Unlike the internal DAC with room correction which pushed the sound to the left and right extremes, the QuteHD precisely positioned instruments and sounds across the soundstage. Impressive feat.

None of these external DACs described so far have been measured by ASR, nor has the AVR400, although the performance of some newer Arcam AVRs have been reviewed by Amir, and the weaknesses of their digital sections in two channel and the generally good performance of their amplifier sections have been reported on this site.

Switching to the Chord Qutest (which has been measured by Amir) there was a slight improvement in all aspects of reproduction over the Chord QuteHD. I noticed slightly more detail retrieval (on headphones at least), more precise articulation of soundstage and cleaner bass. The improvement to my ears between the Qutest and the QuteHD was smaller than the difference between the QuteHD and the irDAC. The Qutest measured well in Amir’s test and has the best overall reproduction of redbook quality recordings of any DAC I have used to date.

I have not yet tried in my system one of the inexpensive Chinese DACs that measure extraordinarily well and compared that to the Chord Qutest. I might try one or several out of curiosity. My conclusion from all of this, based on Amir’s testing and my personal experience I would not assume you are getting highest fidelity reproduction in two channels from digital sources using an AVR with the internal DACs. But YMMV.

kn

Thanks for the detailed explanation.

So you are comparing the internal DAC with room correction - against an external DAC without.

It is not hugely surprising you can hear a difference, room correction is supposed to make an audible difference else what is the point.

Your AVR you say is 12 years old. Built in room correction has come a long way in that time. It may well be that the software is optimised for movie viewing and not so good for music, may not be set up perfectly - or it may just not be very good.

In any case it is more likely the difference you are hearing is from DSP vs No DSP, rather than DAC performance differences. Even though AVR's typically don't measure as well as good standalone dacs, it is also typical for that poor measurement still to be inaudible for real world (especially speaker) listening.

EDIT - I would add - all your descriptions of the different sound characteristcs of various DACS (even though they've not been measured) don't really stack up with what we understand of the science and engineering - which tell us nearly all decent dacs are so good they are transparent to human hearing, and hence won't exhibit differences in properly controlled listening. Further it is generally accepted here that the types of differences you are describing, in particular/for example sound stage - are not of the type that can be influenced by a DAC in any case.

It is almost certain that most of what you are describing as differences between DACs is going to be from perception/cognitive biases, and not from any inherent characteristics of the device.

 

[knownothing]

Thanks for the detailed explanation.

So you are comparing the internal DAC with room correction - against an external DAC without.

It is not hugely surprising you can hear a difference, room correction is supposed to make an audible difference else what is the point.

Your AVR you say is 12 years old. Built in room correction has come a long way in that time. It may well be that the software is optimised for movie viewing and not so good for music, may not be set up perfectly - or it may just not be very good.

In any case it is more likely the difference you are hearing is from DSP vs No DSP, rather than DAC performance differences. Even though AVR's typically don't measure as well as good standalone dacs, it is also typical for that poor measurement still to be inaudible for real world (especially speaker) listening.

EDIT - I would add - all your descriptions of the different sound characteristcs of various DACS (even though they've not been measured) don't really stack up with what we understand of the science and engineering - which tell us nearly all decent dacs are so good they are transparent to human hearing, and hence won't exhibit differences in properly controlled listening. Further it is generally accepted here that the types of differences you are describing, in particular/for example sound stage - are not of the type that can be influenced by a DAC in any case.

It is almost certain that most of what you are describing as differences between DACs is going to be from perception/cognitive biases, and not from any inherent characteristics of the device.

Edited reply:

Thanks for those thoughts on the DSP set up or capabilities being to blame for the poor 2 channel performance of the internal DAC of the AVR, that makes sense.

Perhaps the sonic differences are due to implementation of the analog circuits in DACs. Perhaps as with the AVR signal processing, the DACs have come a long ways in 11 years. Both the Arcam irDAC and the Chord QuteHD were top of class for their respective price categories in 2013. The Qutest is a newer design. Perhaps something was “broken” in older Brown Burr DAC or its implementation in the irDAC. Or perhaps differences in soundstage as reported by listeners in blind tests I’ve run were from perception/congnitive biases demonstrated repeatedly and reliably by subjects.

kn