|
|
Post by garbulky on Sept 9, 2021 10:05:14 GMT -5
Measurements are only one criteria. Equally important, imo, are sonics, build quality, aesthetics, longevity, reliability, pride of ownership and resale value. And of course, all these things cost. In tube terms, measurements are not even a criteria  |
|
|
|
Post by audiobill on Sept 9, 2021 10:07:06 GMT -5
In your opinion.
|
|
cawgijoe
Emo VIPs 
"When you come to a fork in the road, take it." - Yogi Berra
Posts: 5,033
|
Post by cawgijoe on Sept 9, 2021 10:09:12 GMT -5
Not sure if Amir is really "agenda driven" but otherwise I agree with you. If you like a product then what difference do the numbers make? How many of us know all the specs for our cars? The likely answer would be not many, and to add to that, "I don't need to know the specs, I just like driving it." Or, "It does the job." So if that is the case with your car, then what is the deal with your audio gear? What if you have a car whose top speed is supposed to be 140 mph but some magazine or web site says it could only do 135 mph? Do you even drive that fast anyway?  If I spend the $$ to buy a TOTL product, then yes, I do want to know that a lot of time was spent on getting the distortion as low as possible. Or at least better than the competition. Otherwise, I would spend my money on a more inexpensive product as long as it had the features I wanted. Having said that, though these numbers aren't going to make anyone jump for joy, the distortion is likely inaudible. I certainly can't hear any distortion, even if it's there and I've been trying after reading this thread. However, at this point I'd like Emotiva to chime back in. |
|
|
|
Post by monkumonku on Sept 9, 2021 10:11:20 GMT -5
Not sure if Amir is really "agenda driven" but otherwise I agree with you. If you like a product then what difference do the numbers make? How many of us know all the specs for our cars? The likely answer would be not many, and to add to that, "I don't need to know the specs, I just like driving it." Or, "It does the job." So if that is the case with your car, then what is the deal with your audio gear? What if you have a car whose top speed is supposed to be 140 mph but some magazine or web site says it could only do 135 mph? Do you even drive that fast anyway? If I spend the $$ to buy a TOTL product, then yes, I do want to know that a lot of time was spent on getting the distortion as low as possible. Or at least better than the competition. Otherwise, I would spend my money on a more inexpensive product as long as it had the features I wanted. Having said that, though these numbers aren't going to make anyone jump for joy, the distortion is likely inaudible. "...distortion is likely inaudible." That is true. So the misgivings are psychological more than anything else. Pure sound-wise, not knowing any of the measurements but only having your ears to be the judge, what difference would a review/measurement make if it sounded good to you? I hear what you're saying and I would want to think the cost of the product is justified but then it kind of gets into something like, you have a product for which you paid list price and then you have a friend who tells you they got the same thing for 40% off. So does that change your opinion of your product? |
|
|
|
Post by monkumonku on Sept 9, 2021 10:13:10 GMT -5
True, but those who advocate solid state over tubes always point to the specs of both and how SS has much lower distortion, better damping factors, etc. But in the end, those who like tubes could care less because they like the way their tube equipment sounds. |
|
Lsc
Emo VIPs
Posts: 3,434
|
Post by Lsc on Sept 9, 2021 10:23:22 GMT -5
If I spend the $$ to buy a TOTL product, then yes, I do want to know that a lot of time was spent on getting the distortion as low as possible. Or at least better than the competition. Otherwise, I would spend my money on a more inexpensive product as long as it had the features I wanted. Having said that, though these numbers aren't going to make anyone jump for joy, the distortion is likely inaudible. I certainly can't hear any distortion, even if it's there and I've been trying after reading this thread. However, at this point I'd like Emotiva to chime back in. Yes for me the XMC2 still sounds excellent even after the ASR test. The voices still sound amazing and all the songs that I listen to sound the same which is nothing short of amazing since I got my dream speakers. I just want validation by Emotiva like you - they have the units and the tools. |
|
KeithL
Administrator 
Posts: 10,261
|
Post by KeithL on Sept 9, 2021 10:26:17 GMT -5
There are many characteristics that you could measure and report on audio equipment (this is true for most things).
However the commonly quoted specifications for audio gear are RMS power, THD, IMD, S/N, and THD+N.
These are also the measurements that are "at the top of the list" for most audio test gear... And they are the specs that are commonly quoted on spec sheets and in reviews of audio gear... I had personally never seen SINAD quoted in reference to audio test equipment anywhere else... until I read one of Amir's reviews.
So, regardless of whether you consider it a valid or significant metric, you will in fact have difficulty finding a similar measurement anywhere else to compare it to. Notice how we don't specify the top speed of a car in knots either... even though that is a standard metric for boats... and a perfectly valid metric for speed. But, if an automotive magazine were to start quoting their measurement results in knots, they would probably garner more than a few odd looks.
If you want more technical details on the relationships between some of these measurements you might want to check out this white paper:
For the record, from my memory of the IEEE Standards (I helped write and review them but that was long ago): - SINAD = signal to noise and distortion ratio = ratio of signal to everything else
- This is the same as THD+N
- SINAD was adopted by the IEEE and other standards bodies some time ago; I am not sure if THD+N persists outside audio but not something I track - THD = total harmonic distortion = ratio of fundamental to distortion terms using at least 10 harmonics (number varies a little bit among manufacturers)
- SNR = signal to noise ratio = ratio of signal to noise not including distortion terms
- IMD = intermodulation distortion, happens whenever two or more terms are distorted, creates non-harmonic tones as well as harmonic terms
- For a given harmonic distortion level, intermodulation distortion is higher for the same peak voltage level (e.g. IMD2 = HD2 + ~6 dB; IMD3 = HD3 +~9.5 dB) - To convert x% to dB in voltage use dBV = 20*log10(x/100)
HTH - Don |
|
Lonnie
Emo Staff
admin
Pay no attention to the man behind the curtain
Posts: 6,999
|
Post by Lonnie on Sept 9, 2021 10:34:09 GMT -5
Hello Gents, Sorry for the absence here, lots going on and I didn't want to pull the crew off of what they were doing so I just had to find the time to run it myself. The first graph below is from the RMC-1. This shows a full scale signal as well as the ground floor noise/ distortion products. The second graph is the exact same test under the exact same conditions for a XMC-2. Lonnie RMC-1 left center right.pdf (157.7 KB) XMC-2 left center right.pdf (158.37 KB) |
|
|
|
Post by derwin on Sept 9, 2021 10:42:56 GMT -5
Hello Gents, Sorry for the absence here, lots going on and I didn't want to pull the crew off of what they were doing so I just had to find the time to run it myself. The first graph below is from the RMC-1. This shows a full scale signal as well as the ground floor noise/ distortion products. The second graph is the exact same test under the exact same conditions for a XMC-2. Lonnie View AttachmentView AttachmentLonnie, thank you for sharing this! What changed between this measurement of the XMC and the previous one? We didn’t just turn the volume down did we? What is the output voltage for each of these two tests? |
|
|
|
Post by donh50 on Sept 9, 2021 10:45:42 GMT -5
There are many characteristics that you could measure and report on audio equipment (this is true for most things).
However the commonly quoted specifications for audio gear are RMS power, THD, IMD, S/N, and THD+N. These are also the measurements that are "at the top of the list" for most audio test gear... And they are the specs that are commonly quoted on spec sheets and in reviews of audio gear... I had personally never seen SINAD quoted in reference to audio test equipment anywhere else... until I read one of Amir's reviews. So, regardless of whether you consider it a valid or significant metric, you will in fact have difficulty finding a similar measurement anywhere else to compare it to. Notice how we don't specify the top speed of a car in knots either... even though that is a standard metric for boats... and a perfectly valid metric for speed. But, if an automotive magazine were to start quoting their measurement results in knots, they would probably garner more than a few odd looks. If you want more technical details on the relationships between some of these measurements you might want to check out this white paper:
For the record, from my memory of the IEEE Standards (I helped write and review them but that was long ago): - SINAD = signal to noise and distortion ratio = ratio of signal to everything else
- This is the same as THD+N
- SINAD was adopted by the IEEE and other standards bodies some time ago; I am not sure if THD+N persists outside audio but not something I track - THD = total harmonic distortion = ratio of fundamental to distortion terms using at least 10 harmonics (number varies a little bit among manufacturers)
- SNR = signal to noise ratio = ratio of signal to noise not including distortion terms
- IMD = intermodulation distortion, happens whenever two or more terms are distorted, creates non-harmonic tones as well as harmonic terms
- For a given harmonic distortion level, intermodulation distortion is higher for the same peak voltage level (e.g. IMD2 = HD2 + ~6 dB; IMD3 = HD3 +~9.5 dB) - To convert x% to dB in voltage use dBV = 20*log10(x/100)
HTH - Don I am familiar with the technical details, and Walt (nice guy), thank you. I was just defining the terms. Note RMS power is not normally quoted for processors, and in fact it is average power that is the correct term for the product of RMS current and RMS voltage; RMS * RMS is not RMS but RMS squared, a definable but meaningless unit. As one of my early professors was fond of saying: "Follow the units!" This is a quote from the white paper you referenced (emphasis mine): "For a given input frequency and amplitude, SINAD is equal to THD + N, provided the bandwidth for the noise measurement is the same for both (the Nyquist bandwidth)." I don't care what you call it, I am curious about why the reported performance differs, and if the LF noise shelf effect is from bass management as seems likely. |
|
Lonnie
Emo Staff
admin
Pay no attention to the man behind the curtain
Posts: 6,999
|
Post by Lonnie on Sept 9, 2021 10:51:26 GMT -5
Hello Gents, Sorry for the absence here, lots going on and I didn't want to pull the crew off of what they were doing so I just had to find the time to run it myself. The first graph below is from the RMC-1. This shows a full scale signal as well as the ground floor noise/ distortion products. The second graph is the exact same test under the exact same conditions for a XMC-2. Lonnie View AttachmentView AttachmentLonnie, thank you for sharing this! What changed between this measurement of the XMC and the previous one? We didn’t just turn the volume down did we? What is the output voltage for each of these two tests? Output voltage is 4Volts. This was setup as per our standardized QA test, I didn't have the time to try and duplicate whatever Amir did. Lonnie |
|
|
|
Post by derwin on Sept 9, 2021 10:52:28 GMT -5
Lonnie, thank you for sharing this! What changed between this measurement of the XMC and the previous one? We didn’t just turn the volume down did we? What is the output voltage for each of these two tests? Output voltage is 4Volts. This was setup as per our standardized QA test, I didn't have the time to try and duplicate whatever Amir did. Lonnie What varied from the original XMC2 test you guys ran showing -85 dB 2nd harmonic? |
|
|
|
Post by dj7675 on Sept 9, 2021 10:53:02 GMT -5
As noted by several here already, Sinad is just THD+N. Easy to convert them. Also, both SINAD and THD+N are shown in the reviews, so both numbers are always there on Amir's and Lonnie's test results. But the bigger picture is really what is important... Both Amir's and Lonnie's measurements show THD+N of .005/.006 and the RMC1 tested at .001 at 4Volt. It is just finding out why.. Broken unit, Firmware Issue etc.
|
|
Lonnie
Emo Staff
admin
Pay no attention to the man behind the curtain
Posts: 6,999
|
Post by Lonnie on Sept 9, 2021 10:56:59 GMT -5
Output voltage is 4Volts. This was setup as per our standardized QA test, I didn't have the time to try and duplicate whatever Amir did. Lonnie What varied from the original XMC2 test you guys ran showing -85 dB 2nd harmonic? The difference is in the input signal level. The system is optimized for -20dbfs which is what your cable box, APT, Streamer, Blueray and everything else in the world is set to. Amir was running it at 0dbfs, which I can see that if you are testing a simple device like a DAC and if my memory serves me correctly, -20dbfs is what he originally tested the RMC-1 at. Lonnie |
|
|
|
Post by tabbycph on Sept 9, 2021 11:01:34 GMT -5
The difference is in the input signal level. The system is optimized for -20dbfs which is what your cable box, APT, Streamer, Blueray and everything else in the world is set to. Amir was running it at 0dbfs, which I can see that if you are testing a simple device like a DAC and if my memory serves me correctly, -20dbfs is what he originally tested the RMC-1 at. Lonnie In the original test, amir testet at 0 db and the other test hi did, hi didn't specify. |
|
|
|
Post by derwin on Sept 9, 2021 11:01:42 GMT -5
What varied from the original XMC2 test you guys ran showing -85 dB 2nd harmonic? The difference is in the input signal level. The system is optimized for -20dbfs which is what your cable box, APT, Streamer, Blueray and everything else in the world is set to. Amir was running it at 0dbfs, which I can see that if you are testing a simple device like a DAC and if my memory serves me correctly, -20dbfs is what he originally tested the RMC-1 at. Lonnie Thank you guys for taking the time to do this. |
|
Lonnie
Emo Staff
admin
Pay no attention to the man behind the curtain
Posts: 6,999
|
Post by Lonnie on Sept 9, 2021 11:02:55 GMT -5
The difference is in the input signal level. The system is optimized for -20dbfs which is what your cable box, APT, Streamer, Blueray and everything else in the world is set to. Amir was running it at 0dbfs, which I can see that if you are testing a simple device like a DAC and if my memory serves me correctly, -20dbfs is what he originally tested the RMC-1 at. Lonnie Thank you guys for taking the time to do this. Sorry it took some time to do it. Lonnie |
|
|
|
Post by brutiarti on Sept 9, 2021 11:06:54 GMT -5
Not sure if Amir is really "agenda driven" but otherwise I agree with you. If you like a product then what difference do the numbers make? How many of us know all the specs for our cars? The likely answer would be not many, and to add to that, "I don't need to know the specs, I just like driving it." Or, "It does the job." So if that is the case with your car, then what is the deal with your audio gear? What if you have a car whose top speed is supposed to be 140 mph but some magazine or web site says it could only do 135 mph? Do you even drive that fast anyway? If I spend the $$ to buy a TOTL product, then yes, I do want to know that a lot of time was spent on getting the distortion as low as possible. Or at least better than the competition. Otherwise, I would spend my money on a more inexpensive product as long as it had the features I wanted. Having said that, though these numbers aren't going to make anyone jump for joy, the distortion is likely inaudible. Usually the people the spend the big bucks don’t care about those things. They care about how they make them feel. In the other hand, the people that don’t have the coin to get those things, they need to find the reasons why they would never get those. |
|
|
|
Post by louron on Sept 9, 2021 11:23:06 GMT -5
If I spend the $$ to buy a TOTL product, then yes, I do want to know that a lot of time was spent on getting the distortion as low as possible. Or at least better than the competition. Otherwise, I would spend my money on a more inexpensive product as long as it had the features I wanted. Having said that, though these numbers aren't going to make anyone jump for joy, the distortion is likely inaudible. Usually the people the spend the big bucks don’t care about those things. They care about how they make them feel. In the other hand, the people that don’t have the coin to get those things, they need to find the reasons why they would never get those.  I got this product…. I can also afford others….I care about them. I am not showing off. I am not the only one. Most people here are. What is the goal of such a comment? |
|
|
|
Post by AudioHTIT on Sept 9, 2021 11:23:18 GMT -5
Measurements are only one criteria. Equally important, imo, are sonics, build quality, aesthetics, longevity, reliability, pride of ownership and resale value. And of course, all these things cost. I agree that those are all important, though I wouldn’t say equal, and each of us will value these aspects differently. They might also vary in importance based on the type of device, for instance resale value on a processor is much more volatile than on an amplifier, whereas we all might value aesthetics the same regardless of device (though different from each other). I personally value sonics, aesthetics, build quality, and reliability over resale value and pride of ownership; but don’t expect others to feel the same. Measurements however, are something that helped take this industry from a hobby built on kitchen tables and in workshops, to the multi-billion dollar force it is now. But they are evolving as well, and while most manufacturers still provide basic specs on amps (standardized by the FTC), other components often have less data available, and processors (although the most complex) typically have virtually none, even McIntosh only list one distortion spec, and it only for the analog path. Emotiva published very thorough specs for the XMC-1, and were rightly proud of them; I think many of us believed we would eventually see the same from the G3P, but felt getting the product’s full feature set working properly was more important. But that doesn’t reduce the value of measurements, they confirm that the design criteria, implementation, and manufacturing all allowed the original vision to make it successfully to our racks and shelfs. Now we have a case where two different models based on the same design and architecture measure differently; one is near the top of the class on one particular scale, the other (while still good) is much lower, and the question is why? Edit: I see much went on as I wrote this (lots of interruptions), I’m happy to see the new tests and that they look much better. We might still try to figure out why ASR’s results were different, but this is a solid step forward. |
|
