XPA2- Gen 3: Review after 4 months of usage

[hemanth]

Hi,

I have been using emotiva products from couple of years, but not active on the forum. My primary system back in India is DC-1 -> XSP pre-amp -> XPA -1L x 2.
I was looking for a secondary system in US which I can use it temporarily, finally decided to take a plunge on XPA2 Gen 3, since it will be easier to ship it to my home country(due to the low weight factor) and there weren't many decent amplifier in the 1K range(IMHO). I received the XPA2 Gen 3 in September.
This system configuration is as follows: DC1 -> XPA2 Gen3 ->Vintage Technic Speakers + Dayton 12" woofer.

This sound is awesome and I have been enjoying my music more than ever. The bass is tight and the sound is very clean. There is nice decay in the sound which I like a lot.I'm listen to music for long hours(during WFH) and there is no strain/harshness to the ears. There is a good soundstaging. This amplifier sounds way better than XPA -1L which I didn't expect and this will definitely replace my primary amplifier(XPA 1L).

My music source is 1. 200GB of FLACS/MP3s on a HDD connected to Dell Inspiron 3050 -> USB - DCC1
                                     2. HDD connected to Zotac zbox B1320 -> USB - DCC1
                                     3. Spotify Premium -> SPDIF - DC1

Couple of weeks back I bought a Raspberry pi 3 and HifiBerry Digi+ Transformer version, I installed runeaudio and the configuration looks as follows:
        HDD connected to RPI3 -> HifiBerry Digi+ Transformer -> SPDIF -> DC1

The above combination of the RPI3 + HifiBerry Digi+ beats all the above configuration. The music is more detailed and some of the background instruments are sounding more prominently which earlier I never thought existed. This will remain my primary configuration for a long time. I might just add a IFi power adaptor for the RPI3($50 adaptor for $35 RPI is not making sense for me yet).


Thanks,
Hemanth.

[ludi]

Related to the power supplies for the Pi, here is a video that is comparing a couple of them:

Dec 28, 2016 15:25:25 GMT -5 hemanth said:

... I might just add a IFi power adaptor for the RPI3($50 adaptor for $35 RPI is not making sense for me yet).

[garbulky]

Dec 28, 2016 15:25:25 GMT -5 hemanth said:

Hi,

I have been using emotiva products from couple of years, but not active on the forum. My primary system back in India is DC-1 -> XSP pre-amp -> XPA -1L x 2.
I was looking for a secondary system in US which I can use it temporarily, finally decided to take a plunge on XPA2 Gen 3, since it will be easier to ship it to my home country(due to the low weight factor) and there weren't many decent amplifier in the 1K range(IMHO). I received the XPA2 Gen 3 in September.
This system configuration is as follows: DC1 -> XPA2 Gen3 ->Vintage Technic Speakers + Dayton 12" woofer.

This sound is awesome and I have been enjoying my music more than ever. The bass is tight and the sound is very clean. There is nice decay in the sound which I like a lot.I'm listen to music for long hours(during WFH) and there is no strain/harshness to the ears. There is a good soundstaging. This amplifier sounds way better than XPA -1L which I didn't expect and this will definitely replace my primary amplifier(XPA 1L).

My music source is 1. 200GB of FLACS/MP3s on a HDD connected to Dell Inspiron 3050 -> USB - DCC1
                                     2. HDD connected to Zotac zbox B1320 -> USB - DCC1
                                     3. Spotify Premium -> SPDIF - DC1

Couple of weeks back I bought a Raspberry pi 3 and HifiBerry Digi+ Transformer version, I installed runeaudio and the configuration looks as follows:
        HDD connected to RPI3 -> HifiBerry Digi+ Transformer -> SPDIF -> DC1

The above combination of the RPI3 + HifiBerry Digi+ beats all the above configuration. The music is more detailed and some of the background instruments are sounding more prominently which earlier I never thought existed. This will remain my primary configuration for a long time. I might just add a IFi power adaptor for the RPI3($50 adaptor for $35 RPI is not making sense for me yet).


Thanks,
Hemanth.

I too found SPDIF to do better than the USB in the case of the DC-1.

[boomzilla]

Hi hemanth -

First, thank you for taking the time to write your review. I enjoyed it. I'm still considering the XPA-2, Gen. 3 myself, and appreciate your feedback.

Happy New Year!

Boomzilla

[Jim]

Dec 29, 2016 21:40:26 GMT -5 boomzilla said:

Hi hemanth -

First, thank you for taking the time to write your review. I enjoyed it. I'm still considering the XPA-2, Gen. 3 myself, and appreciate your feedback.

Happy New Year!

Boomzilla

You should buy one and do a DBT against your current amp.

[yves]

Dec 29, 2016 22:09:25 GMT -5 Jim said:

Dec 29, 2016 21:40:26 GMT -5 boomzilla said:

Hi hemanth -

First, thank you for taking the time to write your review. I enjoyed it. I'm still considering the XPA-2, Gen. 3 myself, and appreciate your feedback.

Happy New Year!

Boomzilla

You should buy one and do a DBT against your current amp.

Do you mean a double-bollocks test?

[dregoric]

Dec 30, 2016 4:37:50 GMT -5 yves said:

Dec 29, 2016 22:09:25 GMT -5 Jim said:

You should buy one and do a DBT against your current amp.

Do you mean a double-bollocks test?

DBT - double blind test. 

[rbk123]

Feb 21, 2017 5:57:28 GMT -5 dregoric said:

Dec 30, 2016 4:37:50 GMT -5 yves said:

Do you mean a double-bollocks test?

DBT - do***e b**** t**t. 

Doh! Never speak those words in the presence of Boom.

[Casey Leedom]

Dec 28, 2016 17:14:02 GMT -5 garbulky said:

Dec 28, 2016 15:25:25 GMT -5 hemanth said:

...
Couple of weeks back I bought a Raspberry pi 3 and HifiBerry Digi+ Transformer version, I installed runeaudio and the configuration looks as follows:
HDD connected to RPI3 -> HifiBerry Digi+ Transformer -> SPDIF -> DC1

The above combination of the RPI3 + HifiBerry Digi+ beats all the above configuration.
...

I too found SPDIF to do better than the USB in the case of the DC-1.

Huh. That's weird. I would have thought that the asynchronous buffering for both digital inputs would have been essentially identical.

Casey

[garbulky]

Feb 21, 2017 11:37:34 GMT -5 Casey Leedom said:

Dec 28, 2016 17:14:02 GMT -5 garbulky said:

I too found SPDIF to do better than the USB in the case of the DC-1.

Huh. That's weird. I would have thought that the asynchronous buffering for both digital inputs would have been essentially identical.

Casey

It should, I guess. I don't know why. I don't see a strong reason for them being different other than maybe that USB goes through the computers motherboard through what's probably not a USB part of high quality.

[Casey Leedom]

Yeah, but going through a bad USB Transmitter shouldn't affect the sound unless it's completely failing to supply data fast enough which should show up as an audio drop out. The job of "making things right" is really the USB Receiver's. It spools data into a local buffer and then clocks it out at the rate that the DAC circuitry needs. This is the "Asynchronous" part of the USB Receiver. The same thing happens with S/PIDIF input (optical or coax). This is why fancy "audiophile grade" USB Cables, etc. are unnecessary.

I'm sure that KeithL could describe this better.

Casey

[garbulky]

Feb 21, 2017 13:08:12 GMT -5 Casey Leedom said:

Yeah, but going through a bad USB Transmitter shouldn't affect the sound unless it's completely failing to supply data fast enough which should show up as an audio drop out. The job of "making things right" is really the USB Receiver's. It spools data into a local buffer and then clocks it out at the rate that the DAC circuitry needs. This is the "Asynchronous" part of the USB Receiver. The same thing happens with S/PIDIF input (optical or coax). This is why fancy "audiophile grade" USB Cables, etc. are unnecessary.

I'm sure that KeithL could describe this better.

Casey

I get the concept - bits is bits basically. Like I said, I don't know...

[KeithL]

Once you get past asynchronous USB, the differences you're talking about are VERY slight.

On any asynchronous USB input, the data is essentially being clocked by the receiving device - the DAC.
And, on the DC-1, if you have the ASRC enabled, all of the inputs are being re-clocked by the ASRC.
However, in both cases, the absolute limit on jitter performance is the clocks themselves - and the circuits they go with.
(The clocks in the DC-1 are really good - and beyond our ability to measure - and doing so would require very specialized equipment.)

With S/PDIF there are different limiting factors.
Most S/PDIF outputs are what I would characterize as "pretty good" - but, especially on computers, they are still somewhat limited.
Likewise, the S/PDIF INPUT circuitry has a certain minimum jitter spec.... and that may vary depending on things like the noise level and even the cable used.

If you want to compare the two, the jitter on the asynchronous USB input should basically be determined by the clock - and nothing else should matter much.
The jitter on the S/PDIF input will be limited by the source, and by the inherent jitter in the input circuitry itself, which varies.
And, when it's enabled, the ASRC removes most of the jitter from either, and the limiting factor becomes the ASRC itself.

So, if you want to figure out what's LIMITING the performance...... on the DC-1......
With an asynch USB input, and the ASRC disabled, it will be the clock on the USB input.
With an asynch USB input, and the ASRC enabled, it will be the clock on the ASRC.
With a S/PDIF input, and the ASRC enabled, it will be the clock on the ASRC.
With a S/PDIF input, and the ASRC disabled, it will be either the jitter of your source, or the inherent jitter noise floor on the input circuitry.

Another way of looking at that is....
With the ASRC enabled, it should mostly remove everything else from the picture, and make the various inputs sound mostly the same.
With the ASRC disabled, the asynchronous USB input will probably sound the best....
However, the S/PDIF input WILL sound different depending on your source (it's more likely to sound worse than better - but both are possible).
(So, if the ASRC is off, you may well hear differences between S/PDIF and USB, and between different S/PDIF sources).

As I mentioned, measuring this level of jitter performance directly is beyond the capabilities of the test equipment we have.
And, measuring it indirectly, by analyzing the jitter spectra in the audio, the best we can do is to confirm that the variations are unmeasurable with the equipment we have.

You should also be aware that very few of these things are absolutes.....
For example, the ASRC does not remove ALL jitter.... but it reduces the jitter that is present by 40 dB or more.
However, before even thinking too hard about it, note that we're talking about imperfections hundreds of times lower than those present in the equipment that created those recordings to begin with.

Also note that the ANALOG circuitry in the DAC also has  a major effect on sound quality.
And especially the power supply (which affects both the digital and analog performance.)
Almost certainly a far greater effect that the other stuff we're talking about here.
(And the DC-1 has a really nice power supply and analog section.)

And, to address the original point, "audiophile" USB cables are entirely pointless - at least with asynch USB DACs like the DC-1.
Any reasonable quality USB cable, with proper shielding, and that meets USB spec, should work equally well (think Amazon Basics for $10).
There might be a tiny difference between Coax S/PDIF cables, if you have the ASRC turned off (although I've never heard one).
HOWEVER, even if some may sound a tiny bit different, most of the stuff they're selling is purest snake oil, so there's not much reason to expect a more expensive one to sound better.
(Really good Coax cable costs a dollar or two a foot at most..... and the more expensive stuff is unlikely to actually perform any better even if the improvement would be audible.)

Feb 21, 2017 13:19:28 GMT -5 garbulky said:

Feb 21, 2017 13:08:12 GMT -5 Casey Leedom said:

Yeah, but going through a bad USB Transmitter shouldn't affect the sound unless it's completely failing to supply data fast enough which should show up as an audio drop out. The job of "making things right" is really the USB Receiver's. It spools data into a local buffer and then clocks it out at the rate that the DAC circuitry needs. This is the "Asynchronous" part of the USB Receiver. The same thing happens with S/PIDIF input (optical or coax). This is why fancy "audiophile grade" USB Cables, etc. are unnecessary.

I'm sure that KeithL could describe this better.

Casey

I get the concept - bits is bits basically. Like I said, I don't know...

[Casey Leedom]

KeithL, you used an acronym "ASRC" with respect to the DC-1. Is that some setting on the DC-1? Just curious. It sounds like it would be extremely counterproductive to disable it. Or is that to allow for the use of an external clock? Again, pure and simple curiosity.

Casey

[garbulky]

Feb 22, 2017 14:11:21 GMT -5 Casey Leedom said:

KeithL, you used an acronym "ASRC" with respect to the DC-1. Is that some setting on the DC-1? Just curious. It sounds like it would be extremely counterproductive to disable it. Or is that to allow for the use of an external clock? Again, pure and simple curiosity.

Casey

It is an asynchronous thingamijig that reduces jitter. Basically it reclocks the stream. Now the nice thing about this one is that it is more advanced than the one on the XDA-2. The XDA-2's ASRC unit not only reclocked it...it unfortunately resampled it to a specific sample rate thereby making the stream no longer bit perfect. The DC-1's ASRC does not resample - a good thing, so the stream is bit perfect.
I have tried enabling it and disabling it. I didn't find a signficant difference so I left it on.