• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Review and Measurements of Emotiva XMC-1 Gen 2 Pre/Pro

DonH56

Master Contributor
Technical Expert
Forum Donor
Joined
Mar 15, 2016
Messages
7,836
Likes
16,502
Location
Monument, CO
I think you're missing my point.

Fidelity is all about the accuracy of the sound, relative to the recording. It is entirely possible for multi channel playback of a recording to be of lower fidelity than mono, and vice versa.

Provided the fidelity is excellent, authenticity of the experience can be better portrayed by multi channel systems. This is dependent on, among other things, the channel separation and loudspeaker placement. But multi channel has a definite advantage.

Define "excellent". Seriously, that is the rub for objectivists, defining what is "good enough" to not impact the listening experience. I agree with you, but as a listener I admit I am not sure how good SINAD needs to be before I can't hear any improvement. And of course it varies with material, loudness, my mood, phase of the moon... :)
 

Dimifoot

Addicted to Fun and Learning
Forum Donor
Joined
Mar 27, 2019
Messages
506
Likes
746
Location
Greece
It is entirely possible for multi channel playback of a recording to be of lower fidelity than mono

Multichannel reproduction is more accurate because it inherently overcomes room issues that are detrimental to accuracy, issues that plague stereo/mono reproduction.
I wouldn’t like to quote @Floyd Toole , since he is usually around, but he has analyzed it thoroughly in his recent book. If he reads that, he might contribute.
 

GrimSurfer

Major Contributor
Joined
May 25, 2019
Messages
1,238
Likes
1,484
Define "excellent". Seriously, that is the rub for objectivists, defining what is "good enough" to not impact the listening experience. I agree with you, but as a listener I admit I am not sure how good SINAD needs to be before I can't hear any improvement. And of course it varies with material, loudness, my mood, phase of the moon... :)

I'd define excellent as -118 dB SINAD or lower because, under the terms of existing medical research, humans can't hear lower than that.

Now one could make the case for something less stringent (say -110 dB) on the basis that masking would preclude noises of certain frequencies from being heard, but this is a conditional argument not an absolute one.

The other thing to consider is SINAD summing. Now I understand that peak noise and distortion on one device doesn't necessarily correspond to the FR where peak noise will occur on another. But noise and distortion do sum, so one should consider this as part of component matching.

All of this wouldn't be worth a nickel if the year was 1939 and we had bigger issues to address in audio (like perfecting mag tape). But -110 dB SINAD is eminently possible at very little cost in 2019. -118 dB takes a lot more effort, but companies developed reasonably affordable products capable of it a few years ago.

So why are we satisfied with anything less? Because it is cheap or readily available at Best Buy? Or because we have to hawk a kidney to afford it so as to claim we own a piece of audiophile grade electronics?
 
Last edited:

DonH56

Master Contributor
Technical Expert
Forum Donor
Joined
Mar 15, 2016
Messages
7,836
Likes
16,502
Location
Monument, CO
I'd define excellent as -118 dB SINAD or lower because, under the terms of existing medical research, humans can't hear lower than that.

Now one could make the case for something less stringent (say -110 dB) on the basis that masking would preclude noises of certain frequencies from being heard, but this is a conditional argument not an absolute one.

The other thing to consider is SINAD summing. Now I understand that peak noise and distortion on one device doesn't necessarily correspond to the FR where peak noise will occur on another. But noise and distortion do sum, so one should consider this as part of component matching.

All of this wouldn't be worth a nickel if the year was 1939 and we had bigger issues to address in audio (like perfecting mag tape). But -110 dB SINAD is eminently possible at very little cost in 2019. -118 dB takes a lot more effort, but companies developed reasonably affordable products capable of it a few years ago.

So why are we satisfied with anything less? Because it is cheap or readily available at Best Buy? Or because we have to hawk a kidney to afford it so as to claim we own a piece of audiophile grade electronics?

Thanks! Humans cannot hear lower than what? Trying to get an absolute reference; if we use 0 dB SPL as the min (or a little below) and 120 dB as the threshold of pain or so that pretty closely matches your goal. We can pull sounds from below the noise floor of the room but I am not sure by how much. My room is around 20~30 dB absolute SPL (I think) and my system can reach around 115~120 dB so I'm in the ballpark, given the XMC-1 is around 98 dB but we can hear a little below the noise floor. Even so, 118 dB through the entire chain is a lot of dynamic range.

Distortion and noise do combine through the signal chain, but even at the same frequency it is rare that distortion terms are in phase, so usually I RSS (root-sum-square) the terms and that matches measurements except in pathological cases.

For me, the reason is probably the price of a kidney is more than I want to pay... My grandfather lost one and I'm hoping to hang on to both of mine as long as possible. And again for me the dynamic range of my system is comfortably beyond what I can stand (noise floor low enough to hear the quiete stuff when the loud stuff is louder than I like).
 

GrimSurfer

Major Contributor
Joined
May 25, 2019
Messages
1,238
Likes
1,484
Agree. My room doesn't go as low as 20 dB... more like 30 dB average, maybe 28 if I hold my breath.

I'd start with 0.5 dB BTW, as that would be the limit between not being able to hear anything (0 dB meaning zero noise) and possibly being able to detect that something different has occurred (some psychoacousticians say that 0.3 increments can be heard but I haven't read anything definitive from them on this, whereas I have seen several studies over the years supporting 0.5 dB, as low as that may be).

118 dB is a lot of range, I agree. 110 less so (I would have been wiser to use multiples of 10, 6 or 3 dB) but I stand by that as a sensible figure in order to account for transients, overhead etc.

You asked me to define "excellent", but if the discussion moves to "the ballpark" then we're talking about something altogether different IMHO.

I hear you on the kidney thing. I've had a few kidney stones in my life (earned while on deployment in the ME) and they aren't fun when they get to be 1/2 inch and get stuck. The good thing is that the pain is only slightly worse than when they're 1/16th of an inch and pass thru.

Loosing a kidney would require a bit more recovery than I'd care to experience, so I'm not looking to trade one of mine for anything audio related either.
 
Last edited:

DonH56

Master Contributor
Technical Expert
Forum Donor
Joined
Mar 15, 2016
Messages
7,836
Likes
16,502
Location
Monument, CO
Agree. My room doesn't go as low as 20 dB... more like 30 dB average, maybe 28 if I hold my breath.

118 dB is a lot of range, I agree. 110 less so (I would have been wiser to use multiples of 10, 6 or 3 dB) but I stand by that figure.

You asked me to define "excellent", but if the discussion moves to "the ballpark" then we're talking about something altogether different IMHO.

I hear you on the kidney thing. I've had a few kidney stones in my life (earned while on deployment in the ME) and they aren't fun when they get to be 1/2 inch and get stuck. The good thing is that the pain is only slightly worse than when they're 1/16th of an inch and pass thru.

Loosing a kidney would require a bit more recovery than I'd care to experience, so I'm not looking to trade one of mine for anything audio related either.

No worries, you gave me exactly what I asked for (would you were one of my kids ;) ), and I do appreciate that. I can't do "excellent" so transitioned to "ballpark" (best I can do) on my own, my bad.

My room is isolated (floating walls/ceiling, concrete basement floor, mini-split HVAC to isolate form the house ducts) and heavily treated with absorbers so is pretty durn quiet. It was the "daddy tax" when we finished our basement some years back. :)
 

Blumlein 88

Grand Contributor
Forum Donor
Joined
Feb 23, 2016
Messages
20,524
Likes
37,057
Define "excellent". Seriously, that is the rub for objectivists, defining what is "good enough" to not impact the listening experience. I agree with you, but as a listener I admit I am not sure how good SINAD needs to be before I can't hear any improvement. And of course it varies with material, loudness, my mood, phase of the moon... :)

Yes this is the rub. When dealing with music or any meaningful sound (in other words excluding pure test tones or artificially generated difficult signals). You are into gray areas or inflexible absolutes that while good enough are somewhat unnecessary and therefore intolerable. Inflexible absolutes are fine and reduce debate. With something like music systems there is still good reason to venture into the gray.

I am pretty sure THD of -70 db or less is always inaudible with music. But I'm talking only THD, and not IMD, and not noise floors.

With IMD I feel about equally sure at -80 db.

With noise or dynamic range, 70 db is not enough. 80 db is iffy. 90 db gets close to matching with the above except in pathological conditions. I probably really need -100 db for noise levels to feel really as confident as I do for THD and IMD levels.

That is why I don't really like ranking by SINAD. You can't get anywhere close to as useful a ranking as my still highly simplified broad outline above. You have to get exceptional SINAD to be as sure. And while SINAD if good indicates good engineering, and confidently good results well........... excellent engineering doesn't over engineer things because it costs less not to over-engineer.

Now sure like in DACs, if $200 devices manage SINAD of -120 db you'd damn well better have a good reason your expensive device doesn't at least get within sniffing distance of it. But there are power amps that would meet my broad outlines above, and that likely don't impinge on the listener at all. While few power amps, especially affordable ones, get to SINAD of -120 db. Though Hypex and such designs are getting us closer.

Then there are the recordings. I've not found any musical recording that wasn't a pure studio creation, which has a noise floor less than -65 dbFS. If you look with a curve similar to human hearing say ERB, in our most sensitive regions they are maybe -90-95 dbFS.

So my guidelines:
-70 db THD or less.
-80 db IMD or less.
SNR or Dynamic range of 100 db or better.
Frequency response +/- .1 db 20-20khz.

Anything which meets or exceeds all 4 criteria is effectively equivalent in sound for music.

Now I can think of pathological situations that isn't true. But they are just that pathological or highly, highly unusual. And I'm speaking of the resulting signal as delivered to the speaker terminals. Yet someone else might not agree without moving one or more criteria a decibel or three. It is a gray area. But one I think is not far from a good enough mark. And you know what? The thing that still trips this up more often than not is plain old frequency response. Which Dirac can help with.
 
Last edited:

GrimSurfer

Major Contributor
Joined
May 25, 2019
Messages
1,238
Likes
1,484
No worries, you gave me exactly what I asked for (would you were one of my kids ;) ), and I do appreciate that. I can't do "excellent" so transitioned to "ballpark" (best I can do) on my own, my bad.

My room is isolated (floating walls/ceiling, concrete basement floor, mini-split HVAC to isolate form the house ducts) and heavily treated with absorbers so is pretty durn quiet. It was the "daddy tax" when we finished our basement some years back. :)

Ha ha. The best part about all of that is that it is invisible. Therefore no WAF thresholds to cross. Well done to you, Sir.

I don't have an isolated listening room but the neighborhood I live in is quiet (I'm in my late 50s and am considered one of the young ones) with large lots. Roxul Safe-n-sound in the walls and ceilings. Not too many green and leafy broadband jammers near the house (mostly cedar, which doesn't rustle). Zero cicadas (unlike my place in Aus) and no rugrats (other than when the grandkids visit, a recent change that's reminding me how nice it is to be past the parent trap stage of life).

My HVAC (basement) is pretty good due to lots of silver tape, MLV on the main ducts, and Safe-n-Sound in the ceiling. That said, I'd never see 28-30 dB when it is on high cycle. Perhaps 32-33 dB.

WRT excellent vs ballpark, the difference is likely one of wiggle room. If the components come in at "x" dB but have XLR and RCA connectors, it's probably not a bad idea to assume that half of people will go with RCA out of cheapness, ignorance, or convenience. So say goodbye to a bit of SNR right there.

I guess this reflects my conservatism when it comes to specs. Ballpark + makes sense to me. Much as tailoring trousers with a big lunch or dinner in mind, or tires with slightly higher load and speed ratings, has never left me in discomfort or worried behind the wheel.
 

GrimSurfer

Major Contributor
Joined
May 25, 2019
Messages
1,238
Likes
1,484
So my guidelines:
-70 db THD or less.
-80 db IMD or less.
SNR or Dynamic range of 100 db or better.
Frequency response +/- .1 db 20-20khz.

I guess I have to ask (mostly because I respect your views way more than my own in most cases), are the above values the threshold you use to define "excellent"?

Agree with you on needing to see IMD come in lower than THD. Your FR tolerances are, if anything, tighter than mine (0.3 dB would be fine in my book). But the values for THD and IMD are numerically lower than I would use to describe as excellent.

(I place excellence as synonymous with inaudible by anyone on two legs. Beyond that is just being pathological.)
 
Last edited:

Blumlein 88

Grand Contributor
Forum Donor
Joined
Feb 23, 2016
Messages
20,524
Likes
37,057
I guess I have to ask (mostly because I respect your views way more than my own in most cases), are the above values the threshold you use to define "excellent"?

Agree with you on needing to see IMD come in lower than THD. Your FR tolerances are, if anything, tighter than mine (0.3 dB would be fine in my book). But the values for THD and IMD are numerically lower than I would use to describe as excellent.

(I place excellence as synonymous with inaudible by anyone on two legs. Beyond that is just being pathological.)

Humans can hear harmonic distortion down to about .1% (-60 db) using test tones. The level will have to be higher with music. One could possibly find no difficulty setting the bar at -60 db THD. So choosing -70 db for THD is somewhat conservative. You have a margin of safety. Plus most gear has lower harmonic distortion as the signal level is lowered.

More or less the same thinking on IMD. You need to be more careful of it as some of the distortion products are not harmonic and show up in places where the main signal won't mask it (though with music other signals in the part of the frequency band IMD shows will be there to mask it). I don't think anyone will hear an amp that measures THD and IMD at -120 db will hear any difference vs one with the numbers I give as a guideline.

The noise floor as I said almost could be set at only -90 db. Usually noise profiles are higher at lower frequencies and drop as frequencies ascend. So in our most sensitive region even an amp with a -90 db noise floor might only have noise of -100 db where we hear it. But the extra margins allow for imperfect gain matching in a system. Remember my guidelines are for the signal we feed to the speaker input terminals.
 
Last edited:

DonH56

Master Contributor
Technical Expert
Forum Donor
Joined
Mar 15, 2016
Messages
7,836
Likes
16,502
Location
Monument, CO
^^^ Nice summary. Note that is for voltage, not power, to get those dB numbers.

Aside: Two tones of equal amplitude 6 dB lower in voltage than a single tone (so peak voltage is the same for the two cases) will generate 3rd-order IMD products about 10 dB (9.542 dB) higher than the THD spurs, implying IMD spec 10 dB lower than THD is a good rule of thumb (albeit a little misleading). Third-order IMD spurs fall near the original tones, putting them right in the middle of the music (second-order spurs occur at low and much higher frequencies). IMD is also easier to hear since the spurs are atonal (not related to the key signature) instead of harmonic.

Aside 2: I find frequency response most tricky in real life. A relatively small (~1 dB), broad peak or null in the vocal band or just above seems to be fairly easy to detect in an AB test, whilst a deep (~10 dB) null that is very narrow is masked and not noticed. For years I felt phase and excellent time response to be critical, but in the past few years have been playing more with different speakers and EQ, and some of the best sound seems to be when I ignore step response and clean up the frequency response. My hypothesis is that real-world sounds (music, movies) is complex enough that ideal step response is just not a factor. Dr. Toole talks about it in his book; I need to read it again (I devoured the first edition, glanced at the second, and have barely read the third edition even though it is the best and most complete -- lazy, lack of time, etc.)
 
Last edited:

GrimSurfer

Major Contributor
Joined
May 25, 2019
Messages
1,238
Likes
1,484
Humans can hear harmonic distortion down to about .1% (-60 db) using test tones. The level will have to be higher with music. One could possibly find no difficulty setting the bar at -60 db THD. So choosing -70 db for THD is somewhat conservative. You have a margin of safety. Plus most gear has lower harmonic distortion as the signal level is lowered.

More or less the same thinking on IMD. You need to be more careful of it as some of the distortion products are not harmonic and show up in places where the main signal won't mask it (though with music other signals in the part of the frequency band IMD shows will be there to mask it). I don't think anyone will hear an amp that measures THD and IMD at -120 db will hear any difference vs one with the numbers I give as a guideline.

The noise floor as I said almost could be set at only -90 db. Usually noise profiles are higher at lower frequencies and drop as frequencies ascend. So in our most sensitive region even an amp with a -90 db noise floor might only have noise of -100 db where we hear it. But the extra margins allow for imperfect gain matching in a system. Remember my guidelines are for the signal we feed to the speaker input terminals.

You didn't answer the question of what you believe constitutes excellent but provided a very thorough response to what might very well be acceptable.

Many thanks for that... as if often the case with excellent posts, the rationale and principles behind the answer provide more valuable insight than the answer itself.
 

bigguyca

Senior Member
Joined
Jul 6, 2019
Messages
477
Likes
617
Agree. My room doesn't go as low as 20 dB... more like 30 dB average, maybe 28 if I hold my breath.

I'd start with 0.5 dB BTW, as that would be the limit between not being able to hear anything (0 dB meaning zero noise) and possibly being able to detect that something different has occurred (some psychoacousticians say that 0.3 increments can be heard but I haven't read anything definitive from them on this, whereas I have seen several studies over the years supporting 0.5 dB, as low as that may be).

118 dB is a lot of range, I agree. 110 less so (I would have been wiser to use multiples of 10, 6 or 3 dB) but I stand by that as a sensible figure in order to account for transients, overhead etc.

You asked me to define "excellent", but if the discussion moves to "the ballpark" then we're talking about something altogether different IMHO.

I hear you on the kidney thing. I've had a few kidney stones in my life (earned while on deployment in the ME) and they aren't fun when they get to be 1/2 inch and get stuck. The good thing is that the pain is only slightly worse than when they're 1/16th of an inch and pass thru.

Loosing a kidney would require a bit more recovery than I'd care to experience, so I'm not looking to trade one of mine for anything audio related either.

dB's are used to specify a ratio. 0dB means that the value of the ratio is one, x/y = 1, dB = 0. The measured value of noise at 0dB isn't zero.

https://en.wikipedia.org/wiki/Decibel
 

GrimSurfer

Major Contributor
Joined
May 25, 2019
Messages
1,238
Likes
1,484

DonH56

Master Contributor
Technical Expert
Forum Donor
Joined
Mar 15, 2016
Messages
7,836
Likes
16,502
Location
Monument, CO

lashto

Major Contributor
Forum Donor
Joined
Mar 8, 2019
Messages
1,045
Likes
535
@Lonnie
First of all welcome and thank you for the insightful input. Also kudos to Emotiva for sending a technical person and not the usual marketing trumpet.

The DSPs are 32 bit but just like in the analog realm, there is a price to pay for every layer of processing. We spent more then a year just optimizing the system structure to minimize the negative effects and I believe we have done a great job of it. If you listen to one of our new processors, I believe you will be blown away.

That sounds wonderful but considering that a few posts ago you wrote an extended (and excellent) clarification of how DSP does negatively affect the performance of AVP/Rs, your "blown-away" statement is a bit confusing. I for one am not particularly blown-away by lost bits and dBs.

Fully agree with your point that DSP/Eq/etc are not "free".
Also fully agree that an AVP/R cannot perform/measure as good as a pure DAC and some losses are to be expected from the DSP part of the circuit (and many others).

However, the amount of losses looks quite disappointing currently. Losing 5-10dBs would be fine but according to the measurements even the latest/best crop of AVP/Rs do waste +20dBs and quite a few bits. The numbers are quite clear and quite bad: best DACs & amps are above -120dB SINAD and 20bits and *all* AVP/Rs are below -100dB and can barely resolve 16-17bits (actually, looking again at the AVR table, almost all are under -96dB SINAD). That is simply (way) too much waste.

I am no DSP expert but proper 32bit DSP should be able to do better than that. Even if I'm wrong and 32bit is indeed not enough, one can just use 64bit DSP code & hardware nowadays. Yes, it will probably be more expensive but you will have a truly HighRes device. Otherwise, all those HighRes stickers and marketing brochures filled with hi-fi, hi-performance, hi-quality are just disingenuous blabla. -96dB/16bits is simply not HighRes. Neither is DSP at 48kHz like most AVP/Rs still do.
Yes, if well done -96dB/16bits/48kHz can be good enough and you may even call it "blown-away quality". Moreover, anything above might just be inaudible. But that is by definition standard-res and calling it HighRes is a pure and simple lie. And the whole AVP/R industry is plastered with it.

All the above is not just theory or wishful thinking, it is 100% possible with the current technology. And already available in the sub-$5000 price area (at least according to published specs). From https://www.audiotec-fischer.de/en/brax/processors/dsp :
DSP resolution 64 Bit
DSP power 3 x 295 MHz (3.6 Mrd. MAC operations/sec)
Sampling rate 192 kHz
Signal-to-noise ratio digital input 122 dB (A-weighted)
Total harmonic distortion (THD+N) digital input < 0.0004 %
IM distortion (IMD) digital input < 0.0013 %
Crosstalk > 110 dB (1 kHz)
And yes that is an in-car processor but apart from video it does everything an AVP does without losing huge amounts of bits and dBs. Just replace its car-audio DSP with room-audio DSP and you have an AVP worthy of the HighRes sticker.

@amirm would be interesting to see some measurements for that Brax device. And a brand new measurements section for car-audio would also be highly welcome.
 
Last edited:

digicidal

Major Contributor
Joined
Jul 6, 2019
Messages
1,981
Likes
4,838
Location
Sin City, NV
@lashto Holy crap that thing is an absolute monster! Although I know there are many incredible auto setups out there, it's been many years since I bothered keeping up with that market. I can't imagine what kind of rig you'd have to have to truly justify dropping that in the middle of it... but I'd love to see it regardless.

I've always considered car audio to be a bit of a odd pursuit. Don't get me wrong, I listen to a lot of music in my car... and I wish my ML system was better than it is - but considering I'm driving at the time (usually at very high speeds) the noise floor is probably 60dB minimum. I get it for competitions however. Very interesting piece of gear regardless.

The only problem I see is that if you put DSP of that power, into an AVP/AVR with everything else - you definitely are not going to have a $5K device when you're done... but you would have a competition destroying ~$10K one I'd imagine. ;)
 

amjosh

Member
Joined
Aug 3, 2018
Messages
28
Likes
14
This is a review and detailed measurements of the Emotiva XMC-1 Generation 2 Home Theater Processor. It is on kind loan from a member and that is looking to upgrade it to the XMC-2 (they apparently have an attractive trade up program). The XMC-1 came out in 2015 I think. Not sure when the Gen 2 version came out. The XMC-1 costs US $2,499 so not cheap.

I like the display on the XMC-1 as far as the wealth of information it shows:

This is a review and detailed measurements of the Emotiva XMC-1 Generation 2 Home Theater Processor. It is on kind loan from a member and that is looking to upgrade it to the XMC-2 (they apparently have an attractive trade up program). The XMC-1 came out in 2015 I think. Not sure when the Gen 2 version came out. The XMC-1 costs US $2,499 so not cheap.

I like the display on the XMC-1 as far as the wealth of information it shows:


Otherwise the look and feel is ordinary.

The menu system drove me crazy. Every button press takes 2 seconds or more to take hold. I don't know how such a slower processor would have passed any kind of review in a high-end processor.

The back panel shows excellent connectivity in the form of both XLR Output and (one) analog input:

View attachment 34938

Routing my PC desktop through it resulted in UHD resolution but at just 30 Hz. Older HDMI chipset takes the blame.

To test, I connected my cables to far right connectors marked "R" and "L." Could not get any output from them or the RCA jacks above them. Tried everything. Reset to factory, etc. Nothing worked. I then just moved the cable until I got a signal. Only then did read the fine print that the connectors on the right are for subwoofer Right and Left! Sometimes it is hell being a reviewer!

DAC Audio Measurements
Here is our usual dashboard using HDMI In and XLR out:

View attachment 34939

We have an accurate and proper 4 volt output at 0 dB volume. I think this is the first AVR or processor to get this right. Note that the test mode is Reference Stereo which is supposed to turn off all processing including bass management. SINAD performance is good for an AVR/Processor:
View attachment 34940

But not compared to any decent desktop DAC:
View attachment 34941

We have yet to have enough escape velocity in any home theater product to break into our green bucket let alone blue.

Here is the jitter performance over HDMI:
View attachment 34942

It looks reasonably clean. Noise floor is high though and is hiding jitter components. Fortunately the levels are well below audibility.

From here on, I had to switch to using Coax input. For some reason, using ASIO interface with my graphics card would result in 7.1 output instead of stereo. And the XMC-1 would proceed to perform some processing on that. Fortunately, coax performance is essentially identical to HDMI so results should hold for both inputs:
View attachment 34943

Notice how we are missing the spec by fair bit in channel 1. Channel 2 is better but still shy of the specification. The differential in channel performance tells me poor routing of signals or power to the DAC.

Dynamic range was not that great for a processor at this price:
View attachment 34944

We are barely clearing 16 bits.

Here is our intermodulation distortion+noise relative to level:
View attachment 34945

I am showing both output types (XLR and RCA) for those of you interested in RCA measurements. XLR is just a bit quieter but otherwise the same.

THD+N versus frequency is nothing to write home about:

View attachment 34946

The red line is a $99 DAC board.

Frequency response is flat as we would expect:

View attachment 34947

I know. I know. You are wondering why the decapitated tiger is adorning the XMC-1 in the review picture. This is why:
View attachment 34948

I think this is the most broken linearity test we have seen! The output is muted for any signal below -90 dB! I thought something was broken when the test started from -120 dB with no output from the unit. Then all of a sudden it output the -90 dB signal. To show that clearly, see these graphs:
View attachment 34949

How the heck did you break this Emotiva? It is not like the output gets inaccurate. There is just no output below -90 dB. Once in a while it would do something as indicated by the partial sine wave but otherwise nothing.

Analog Input Performance
I took advantage of the nice XLR input to run some analog pre-amp tests. Here is the pass-through performance at the same level in/out:
View attachment 34950

Hmmm. This is worse than the DAC performance. :(

Here is the frequency response both in Reference Stereo and Direct (which allows DSP):

View attachment 34951

We see that Direct mode puts the ADC in the loop causing a sharp cut off in bandwidth to tune of 22 kHz. Here is the performance in both mode versus level:

View attachment 34952

Performance without ADC is good but with ADC, it does down the drain although it still beats the much more expensive NAD M17 Processor I recently reviewed.

Crosstalk was good, beating the NAD M17 there too:

View attachment 34953

Conclusions
I had high hopes for Emotiva XMC-1 seeing how it is one of few home theater products with specifications. Alas, I was not able to meet those specs even though the measurements it did produce, put it high in the home theater department.

The worst failing was the linearity test showing that there is some serious signal processing error inside this unit, muting the output below -90 dB. Was this put in there to improve specs? Or downright bug and lack of testing to find the same?

Even without that problem, we still can't match performance of a $99 desktop DAC even though we paid $2,500. Yes, we got more channels than 2 and have room EQ but still, I want my good performance damn it!

The XMC-2 has no specifications but I have read that they plan to release Audio Precision test results. Let's keep our fingers crossed that they deliver there.

------------
As always, questions, comments, recommendations, etc. are welcome.

Driving up and down another $200 miles to pick up more home theater gear on Thursday. Need gas and insurance money folks. So donate generously using: https://www.audiosciencereview.com/forum/index.php?threads/how-to-support-audio-science-review.8150/

Great review @amirm as usual. I had this intuition, that AVRs are many generations behind the high end DAC / AMPS in audio department. I have upgraded to earlier version of Emotiva surround processor to improve the audio from the Denon's flagship receiver. The difference was immense, but still the sound fell short compared to the output from AudioGD NF 11.28. That is also proving to be relatively high on noise floor :). I am not sure, what this difference. I have another receiver from Harman Kardon, AVR 3700 7.2 receiver. This one is all digital amplifier and weighs less than 1/4th of the Denon. However the sound quality coming out of this is quite incredible compared to any other receiver I have heard. Please do a review of this if you can :) @amirm.

Otherwise the look and feel is ordinary.

The menu system drove me crazy. Every button press takes 2 seconds or more to take hold. I don't know how such a slower processor would have passed any kind of review in a high-end processor.

The back panel shows excellent connectivity in the form of both XLR Output and (one) analog input:

View attachment 34938

Routing my PC desktop through it resulted in UHD resolution but at just 30 Hz. Older HDMI chipset takes the blame.

To test, I connected my cables to far right connectors marked "R" and "L." Could not get any output from them or the RCA jacks above them. Tried everything. Reset to factory, etc. Nothing worked. I then just moved the cable until I got a signal. Only then did read the fine print that the connectors on the right are for subwoofer Right and Left! Sometimes it is hell being a reviewer!

DAC Audio Measurements
Here is our usual dashboard using HDMI In and XLR out:

View attachment 34939

We have an accurate and proper 4 volt output at 0 dB volume. I think this is the first AVR or processor to get this right. Note that the test mode is Reference Stereo which is supposed to turn off all processing including bass management. SINAD performance is good for an AVR/Processor:
View attachment 34940

But not compared to any decent desktop DAC:
View attachment 34941

We have yet to have enough escape velocity in any home theater product to break into our green bucket let alone blue.

Here is the jitter performance over HDMI:
View attachment 34942

It looks reasonably clean. Noise floor is high though and is hiding jitter components. Fortunately the levels are well below audibility.

From here on, I had to switch to using Coax input. For some reason, using ASIO interface with my graphics card would result in 7.1 output instead of stereo. And the XMC-1 would proceed to perform some processing on that. Fortunately, coax performance is essentially identical to HDMI so results should hold for both inputs:
View attachment 34943

Notice how we are missing the spec by fair bit in channel 1. Channel 2 is better but still shy of the specification. The differential in channel performance tells me poor routing of signals or power to the DAC.

Dynamic range was not that great for a processor at this price:
View attachment 34944

We are barely clearing 16 bits.

Here is our intermodulation distortion+noise relative to level:
View attachment 34945

I am showing both output types (XLR and RCA) for those of you interested in RCA measurements. XLR is just a bit quieter but otherwise the same.

THD+N versus frequency is nothing to write home about:

View attachment 34946

The red line is a $99 DAC board.

Frequency response is flat as we would expect:

View attachment 34947

I know. I know. You are wondering why the decapitated tiger is adorning the XMC-1 in the review picture. This is why:
View attachment 34948

I think this is the most broken linearity test we have seen! The output is muted for any signal below -90 dB! I thought something was broken when the test started from -120 dB with no output from the unit. Then all of a sudden it output the -90 dB signal. To show that clearly, see these graphs:
View attachment 34949

How the heck did you break this Emotiva? It is not like the output gets inaccurate. There is just no output below -90 dB. Once in a while it would do something as indicated by the partial sine wave but otherwise nothing.

Analog Input Performance
I took advantage of the nice XLR input to run some analog pre-amp tests. Here is the pass-through performance at the same level in/out:
View attachment 34950

Hmmm. This is worse than the DAC performance. :(

Here is the frequency response both in Reference Stereo and Direct (which allows DSP):

View attachment 34951

We see that Direct mode puts the ADC in the loop causing a sharp cut off in bandwidth to tune of 22 kHz. Here is the performance in both mode versus level:

View attachment 34952

Performance without ADC is good but with ADC, it does down the drain although it still beats the much more expensive NAD M17 Processor I recently reviewed.

Crosstalk was good, beating the NAD M17 there too:

View attachment 34953

Conclusions
I had high hopes for Emotiva XMC-1 seeing how it is one of few home theater products with specifications. Alas, I was not able to meet those specs even though the measurements it did produce, put it high in the home theater department.

The worst failing was the linearity test showing that there is some serious signal processing error inside this unit, muting the output below -90 dB. Was this put in there to improve specs? Or downright bug and lack of testing to find the same?

Even without that problem, we still can't match performance of a $99 desktop DAC even though we paid $2,500. Yes, we got more channels than 2 and have room EQ but still, I want my good performance damn it!

The XMC-2 has no specifications but I have read that they plan to release Audio Precision test results. Let's keep our fingers crossed that they deliver there.

------------
As always, questions, comments, recommendations, etc. are welcome.

Driving up and down another $200 miles to pick up more home theater gear on Thursday. Need gas and insurance money folks. So donate generously using: https://www.audiosciencereview.com/forum/index.php?threads/how-to-support-audio-science-review.8150/

Great review @amirm as usual. I have always wondered about the low audio quality from these high end surround processors and receivers. I have the earlier version from Emotiva. It improves the sound quality by a generation comapared to the default one from Denon's flagship receiver. But still, its a far cry from AudioGD NFB 11.28, that I compared to. That one also seems to have high levels of noise compared to other flag ship ones. However, I also have another receiver from Harman Kardon, AVR 3700 7.2 channel. This is featherweight compared to Denon, for being all digital amplifier. However, the sound quality coming out from this is quite incredible and gives run for money to Emotiva. Would love to see this receiver going through your rigorous tests.
 

Tonygeno

Active Member
Forum Donor
Joined
Sep 11, 2019
Messages
192
Likes
248
Location
Massachusetts
Sorry if this has been addressed, but when I owned the Gen 1, Dirac muted between tracks. So when listening to an opera recording that flowed from track to track, there was a second of silence as the next track started. And it only occurred with Dirac. Needless to say, this was a non-starter given my musical preferences, so I sent it back and took the $99 loss on Dirac that I had paid for.
 

Flak

Senior Member
Industry Insider
Joined
Nov 18, 2018
Messages
385
Likes
593
@Lonnie
First of all welcome and thank you for the insightful input. Also kudos to Emotiva for sending a technical person and not the usual marketing trumpet.



That sounds wonderful but considering that a few posts ago you wrote an extended (and excellent) clarification of how DSP does negatively affect the performance of AVP/Rs, your "blown-away" statement is a bit confusing. I for one am not particularly blown-away by lost bits and dBs.

Fully agree with your point that DSP/Eq/etc are not "free".
Also fully agree that an AVP/R cannot perform/measure as good as a pure DAC and some losses are to be expected from the DSP part of the circuit (and many others).

However, the amount of losses looks quite disappointing currently. Losing 5-10dBs would be fine but according to the measurements even the latest/best crop of AVP/Rs do waste +20dBs and quite a few bits. The numbers are quite clear and quite bad: best DACs & amps are above -120dB SINAD and 20bits and *all* AVP/Rs are below -100dB and can barely resolve 16-17bits (actually, looking again at the AVR table, almost all are under -96dB SINAD). That is simply (way) too much waste.

I am no DSP expert but proper 32bit DSP should be able to do better than that. Even if I'm wrong and 32bit is indeed not enough, one can just use 64bit DSP code & hardware nowadays. Yes, it will probably be more expensive but you will have a truly HighRes device. Otherwise, all those HighRes stickers and marketing brochures filled with hi-fi, hi-performance, hi-quality are just disingenuous blabla. -96dB/16bits is simply not HighRes. Neither is DSP at 48kHz like most AVP/Rs still do.
Yes, if well done -96dB/16bits/48kHz can be good enough and you may even call it "blown-away quality". Moreover, anything above might just be inaudible. But that is by definition standard-res and calling it HighRes is a pure and simple lie. And the whole AVP/R industry is plastered with it.

All the above is not just theory or wishful thinking, it is 100% possible with the current technology. And already available in the sub-$5000 price area (at least according to published specs). From https://www.audiotec-fischer.de/en/brax/processors/dsp :
DSP resolution 64 Bit
DSP power 3 x 295 MHz (3.6 Mrd. MAC operations/sec)
Sampling rate 192 kHz
Signal-to-noise ratio digital input 122 dB (A-weighted)
Total harmonic distortion (THD+N) digital input < 0.0004 %
IM distortion (IMD) digital input < 0.0013 %
Crosstalk > 110 dB (1 kHz)
And yes that is an in-car processor but apart from video it does everything an AVP does without losing huge amounts of bits and dBs. Just replace its car-audio DSP with room-audio DSP and you have an AVP worthy of the HighRes sticker.

@amirm would be interesting to see some measurements for that Brax device. And a brand new measurements section for car-audio would also be highly welcome.

Nice unit indeed :)

not a cheap in-car processor though at €4.900 and I'm not sure that one can take advantage of its great specs because of the noise in cars (especially with traffic)
 
Top Bottom