timg
Minor Hero
Posts: 71
|
Post by timg on Sept 9, 2021 19:02:20 GMT -5
Take a look at SINAD vs. Measured Level from the RMC-1 test www.audiosciencereview.com/forum/index.php?threads/emotiva-rmc-1-av-processor-review.11673/page-46#post-344170Even the RMC 1 has SINAD close to 85 dB in some of Amir's testing. It looks like SINAD is impacted by input gain (Per Lonnie above, optimized for -20 dBFS input signal to match typical source content) and output gain. It's also heavily impacted by mode. I'm sure that Dolby, DTS, and Dirac also reduce SINAD / increase noise/distortion. I would love to see a much more thorough measuring test for pre-pros that includes a variety of input levels, variety of output levels, and variety of modes. The challenge is how to represent this and make this consistent across multiple brands. It would not surprise me if some brands have peak values far in excess of their tested values (like the XMC-2), and others have typical values far less than their optimized peak value. To me, some of the most interesting values would come when a Dolby or DTS signal is being decoded in a native mode (not upmixed). Tim |
|
|
|
Post by donh50 on Sept 9, 2021 19:03:40 GMT -5
Gentleman, Let me preface the following comments by stating I'm not attacking anyone here or on the other site.I'm going to try and state the facts as I see them in hopes that it will answer your questions. As I have said before, I can't answer anything in regards to Amir's test. You should ask him directly. We tried to set up our system to as closely match what he did as possible and that produced the -85db down harmonics but this is not a real world scenario and not how it should be tested but obviously a different test than what he did on the RMC-1. When tested to a real world scenario the results are as we publish and what I posted several post back. We have very tight QA standards and every single unit is tested and logged. We don't do a simple snap shot of a couple of metrics. The reason we don't post the QA reports is simple because it won't mean anything to the vast majority of people. Attached are two random QA test from production today. One is of an RMC-1 and the other is from the XMC-2. If you are a gluten for punishment, please feel free to go through them and you will see just how closely the two units match each other and just how well they perform. Lonnie View AttachmentView AttachmentSo I went through them, though did not compare every single spec line-by-line. Certainly the -20 dBFS THD+N, SNR, and other key parameters are closely matched. (Note: Some conditions list 100 mVrms, others -20 dBFS, are these are the same amplitude? Not sure what they reference dBFS to, is that 1 Vrms?) As Lonnie said earlier, the specs are at -20 dBFS rather than 0 dBFS. Nice to see THD+N and SNR are extremely close channel-to-channel! Level among channels also tracks closely to -70 dB, another stumbling point for some other components. I am curious about the THD ratio plot on page 12 of the XMC-2 report (it is virtually identical for the RMC-1). There is a significant rise as frequency drops, so 0.002% at 1 kHz is around 0.15%~0.2% at 5 Hz, curious why? It looks like resolution drops down there, which is probably related to the FFT record (automated tests don't usually wait for minutes to collect FFT points, and to produce an 88-page report takes some time!) Why the sharp cutoff around 12 kHz? And what is different to produce the similar plot on page 87? Thanks, Don |
|
Lonnie
Emo Staff 
admin
Pay no attention to the man behind the curtain
Posts: 6,999
|
Post by Lonnie on Sept 9, 2021 19:42:35 GMT -5
Hi Lonnie, As a very long time Emotiva customer, I appreciate all the work you and your team put into providing some of the best HiFi gear at very competive prices. As a devoted vinylphile, I am also grateful for the Balanced XLR input and Reference Stereo mode on my XMC-2. I think it would be cool and help others gain confidence if you or your team could take a look at a few of ASR's processor/AVR review tests and try to recreate them with the XMC-2. I know this would be a lot of work. But, you could then send the exact XMC-2 to Amir @asr and he could perform the exact tests. If you both achieve the same results, then you would be doing the Scientific method by independently replicating the results. The downside would be if the results continue to diverge. I do appreciate you posting the XMC-2 PDF AP plot which to my eyes look exceptionally good. Thanks. Amir measured the Okto DAC 8 Stereo (DAC) as did JA over at Stereophile. They both stated that “it measures superbly”. More importantly, the measurements by both were identical, IIIDENTICAL! JA knows what he is doing. Amir knows what he is doing. Doesn’t answer the question why the XMC measured differently from the RMC but adds some chops to Amir’s ability to take (consistent) measurements. Russ Please go back and read my previous post, I have answered this question multiple times. Thanks, |
|
|
|
Post by JKCashin on Sept 9, 2021 19:43:31 GMT -5
To put it quite bluntly I personally will be quite satisfied if we " merely have the best sounding processor out there"....  I basically agree with this approach... but... some of us use measurements as a way of comparing units, and certifying which one is "best". I initially chose the XMC-2 based on specifications, as there just wasn't much out there in the way of subjective reviews. Now there are plenty of reviews available, but I would still make the decision to purchase an XMC-2 were I to be making that decision today. The reason is ALL the reviews I have read/watched to date have indicated no issues with the sound (save one, see footnote 1), and in fact they all have high praise for the sound. There may be examples of negative subjective reviews around, but I have not seen/read any, and I have looked pretty far-and-wide for reviews of this (and other) units. Some of the subjective reviews available now do mention some non sound-related issues, but for me these issues aren't severe enough to offset the high value provided by the sound quality. In my case, the work-arounds, while annoying, don't impact my enjoyment of the XMC-2, because to me it sounds better than anything else I have heard. To close, wouldn't it be interesting to see how we all reacted in a world where the only measurements possible quantified only those things we can actually hear... my guess is there would be
1 So what is the exception I referred to? It's the bass bump/LFE issue. That's it. It's the only sound-related issue I am aware of. But the thing with this specific issue is it was not present until F/W 2.1. It's not a day-one issue, and it's not a H/W issue. It was introduced with F/W 2.1, so it is possible to fix the issue with some future F/W. Do I want it fixed now? Absolutely. Do I expect it to be fixed now? No, but I do hope Emotiva considers it a priority. I've written firmware, drivers, and embedded software for over 30 years, so I appreciate the level of frustration the folks over at HW must be experiencing. If this were easy to fix, it would have been fixed already. Yes, a company with more resources would likely have fixed it by now, but that's not reason enough for me to seriously 2 look elsewhere. 2 Recently my financial situation has changed, significantly, and for the better. As a result I have started looking at other options... starting with the processor. I'm looking at other options mostly because I can, and not entirely as a result of any dissatisfaction with the XMC-2. I have tempered my language... "mostly" but not "entirely"... to acknowledge there is a small component of wanting to move on due to the current situation. If I could find the same or better sound, 9.2.6 or better, differential LCR and full balanced ouputs, all tuned with Dirac Live, and the unit basically worked without the delays and glitches of the XMC-2, for a reasonable price (up to $US8K) I'd consider it. Before you suggest the HTP-1, aside from it looking like some poorly executed DIY project, it does not support any reference stereo mode, which is a non-starter for me. The XMC-2 has exactly the feature set I want, no more, with the exception of DTS-X:Pro and DLBC but my understanding is DTS-X:Pro is coming, and a MiniDSP 2x4 HD would provide me with mostly everything I would want DLBC for. |
|
|
|
Post by JKCashin on Sept 9, 2021 20:02:41 GMT -5
Compared to a processor a DAC is a relatively simple device. I am blind in my left eye. A few years back I was looking for a "really good" pair of binoculars, only to be shocked by the price. The units I was looking at that met my needs were $400+, but my budget was $200 Then a friend suggested I look into getting a monocular. Initially I thought this would result in me getting a device with optics equivalent to a binocular variant, at half the price. Turned out I got something equivalent optically for about $150 (I actually paid $99, sale, but that's not my point). My point is that without the need for a complex mechanism for simultaneously focusing both left and right units, and keeping both units perfectly aligned, my money went into just the optics. My monocular is to a binocular almost the same as a DAC is to processor. The XMC-2 is not a DAC with a few extra bits. It's a DAC with a whole metric poop ton of other stuff so it should not come as any surprise that there might be some DACs that outperform the XMC-2 even though they are cheaper. But they don't do all the other things the XMC-2 does, just like my monocular doesn't do all the things a binocular can do. |
|
Lonnie
Emo Staff
admin
Pay no attention to the man behind the curtain
Posts: 6,999
|
Post by Lonnie on Sept 9, 2021 20:06:51 GMT -5
Take a look at SINAD vs. Measured Level from the RMC-1 test www.audiosciencereview.com/forum/index.php?threads/emotiva-rmc-1-av-processor-review.11673/page-46#post-344170Even the RMC 1 has SINAD close to 85 dB in some of Amir's testing. It looks like SINAD is impacted by input gain (Per Lonnie above, optimized for -20 dBFS input signal to match typical source content) and output gain. It's also heavily impacted by mode. I'm sure that Dolby, DTS, and Dirac also reduce SINAD / increase noise/distortion. I would love to see a much more thorough measuring test for pre-pros that includes a variety of input levels, variety of output levels, and variety of modes. The challenge is how to represent this and make this consistent across multiple brands. It would not surprise me if some brands have peak values far in excess of their tested values (like the XMC-2), and others have typical values far less than their optimized peak value. To me, some of the most interesting values would come when a Dolby or DTS signal is being decoded in a native mode (not upmixed). Tim Well Tim, wish granted.  On page 10 of this thread I posted our standard QA test for both the RMC and XMC. In that test you will find specs for all inputs, at varying input levels and signal types including Atmos. Lonnie |
|
|
|
Post by louron on Sept 9, 2021 20:07:10 GMT -5
So I would like to clarify something here. Amir is testing home theatre processors as if they were some simple two channel DAC and this I believe is the root to the discussion. DACs are very simple products, you have an input MUX going staight to the DAC chips, an I/V converter and then onto a volume control if it has an analog volume control, other wise the volume is done within the DAC chip itself. That’s it, pretty simple and very straightforward. By contrast a Home Theatre processor is a very complex device. Aside from all the different inputs and how they get handled, you also have to deal with Dolby and DTS decoders within the DSP, time delays, crossovers, mixers, EQs and much more. All of which affect how the signals are handled and manipulated. Thus, the testing methodologies need to suit the product under test. Thus we have a fundamental difference in how things are tested. When I test our own DACs, we look for very similar things, although our test are quite a bit deeper, but when we test our HT processors, the test are setup to suit the piece of gear under test. One test does not fit all and this in my opinion is the root of this whole discussion. Lonnie Hi Lonnie, What is the difference between the first test results you published and the latest ones? I would like to understand. I can’t adjust the input level of my equipment and I do listen at higher volume than -20 dB because it is just insufficient with my amp gain and speakers efficiency specially after Dirac Live calibration. I often find myself around -12dB to -8 dB and my ears are not bleeding. I just want a level field! Does your initial tests results are the results at 0 dB and the new ones at -20dB ? yes or no? thank you. |
|
|
|
Post by donh50 on Sept 9, 2021 20:14:51 GMT -5
Curious: When the generator is at -20 dBFS on e.g. page 8 of the XMC-1 report, what are the input and output voltages for the THD+N test?
|
|
Lonnie
Emo Staff
admin
Pay no attention to the man behind the curtain
Posts: 6,999
|
Post by Lonnie on Sept 9, 2021 20:21:08 GMT -5
So I would like to clarify something here. Amir is testing home theatre processors as if they were some simple two channel DAC and this I believe is the root to the discussion. DACs are very simple products, you have an input MUX going staight to the DAC chips, an I/V converter and then onto a volume control if it has an analog volume control, other wise the volume is done within the DAC chip itself. That’s it, pretty simple and very straightforward. By contrast a Home Theatre processor is a very complex device. Aside from all the different inputs and how they get handled, you also have to deal with Dolby and DTS decoders within the DSP, time delays, crossovers, mixers, EQs and much more. All of which affect how the signals are handled and manipulated. Thus, the testing methodologies need to suit the product under test. Thus we have a fundamental difference in how things are tested. When I test our own DACs, we look for very similar things, although our test are quite a bit deeper, but when we test our HT processors, the test are setup to suit the piece of gear under test. One test does not fit all and this in my opinion is the root of this whole discussion. Lonnie Hi Lonnie, What is the difference between the first test results you published and the latest ones? I would like to understand. I can’t adjust the input level of my equipment and I do listen at higher volume than -20 dB because it is just insufficient with my amp gain and speakers efficiency specially after Dirac Live calibration. I often find myself around -12dB to -8 dB and my ears are not bleeding. I just want a level field! Does your initial tests results are the results at 0 dB and the new ones at -20dB ? yes or no? thank you. The output level of your sources (your other gear) is set to a level consistent to -20db, which is what the system is optimized for. The output level you play the processor at makes no difference. Turn it as loud as you want. Lonnie |
|
|
|
Post by donh50 on Sept 9, 2021 20:44:30 GMT -5
-20 dBVrms (100 mVrms)?
|
|
Lonnie
Emo Staff
admin
Pay no attention to the man behind the curtain
Posts: 6,999
|
Post by Lonnie on Sept 9, 2021 21:01:14 GMT -5
-20dbfs pertains to a digital signal level whereas 100mv pertains to a analog level. So they are two different things and used to measure in different ways. It all depends on context. Hope this helps. Lonnie |
|
Lsc
Emo VIPs 
Posts: 3,434
|
Post by Lsc on Sept 9, 2021 21:09:10 GMT -5
-20dbfs pertains to a digital signal level whereas 100mv pertains to a analog level. So they are two different things and used to measure in different ways. It all depends on context. Hope this helps. Lonnie We have so much to learn Lonnie. Maybe you can have a lunch end learn session you know….given all that free time you have 😬😂 |
|
Lonnie
Emo Staff
admin
Pay no attention to the man behind the curtain
Posts: 6,999
|
Post by Lonnie on Sept 9, 2021 21:15:08 GMT -5
-20dbfs pertains to a digital signal level whereas 100mv pertains to a analog level. So they are two different things and used to measure in different ways. It all depends on context. Hope this helps. Lonnie We have so much to learn Lonnie. Maybe you can have a lunch end learn session you know….given all that free time you have 😬😂 Would love to. Seriously, what topics would everyone like to know more about? Nick and I were talking about doing some videos, but I would like to know what everyone would like to see. Lonnie |
|
|
|
Post by garbulky on Sept 9, 2021 21:20:26 GMT -5
We have so much to learn Lonnie. Maybe you can have a lunch end learn session you know….given all that free time you have 😬😂 Would love to. Seriously, what topics would everyone like to know more about? Nick and I were talking about doing some videos, but I would like to know what everyone would like to see. Lonnie Hi Lonnie. I'm trying to understand the -20db limitation. Why would sources output only a -20db and not a 0db or close to that digital signal? Are you counting lossless or music sources in this? I have numerous music sources that easily go past -20 db peaks digitally. |
|
Lonnie
Emo Staff
admin
Pay no attention to the man behind the curtain
Posts: 6,999
|
Post by Lonnie on Sept 9, 2021 21:27:48 GMT -5
Would love to. Seriously, what topics would everyone like to know more about? Nick and I were talking about doing some videos, but I would like to know what everyone would like to see. Lonnie Hi Lonnie. I'm trying to understand the -20db limitation. Why would sources output only a -20db and not a 0db or close to that digital signal? Are you counting lossless or music sources in this? I have numerous music sources that easily go past -20 db peaks digitally. Because you need headroom. A nominal output level of -20db gives you 20db of headroom before everything runs out of steam. Digital parts cannot go above 0, they can only go negative. If your source was set to 0db then you have no headroom and everything would be clipped and distorted. Hope this helps. Lonnie |
|
|
|
Post by autocrat on Sept 9, 2021 21:56:10 GMT -5
One of my favourite reviews on ASR is of a PS Audio DAC. www.audiosciencereview.com/forum/index.php?threads/review-and-measurements-of-ps-audio-perfectwave-directstream-dac.9100/ It's interesting because: a. the DAC is very expensive. Way more than I would ever consider paying for a DAC, and I am not alone there. b. it performs very very poorly in the tests. Outstandingly badly in fact. It performs poorly because Amir tests to an engineering standard that assumes higher scores equal better engineering. As far as it goes that may be true, but the PS Audio units are engineered to perform exactly as they do, because regardless of the test results people are paying the big bucks because they absolutely love the sound the get from these DACs. I mean absolutely. So IMO the cause of the test failure is not bad engineering but good engineering. ASR and PS Audio have very different objectives, it just so happens that the equipment is the same. Same here. |
|
|
|
Post by garbulky on Sept 9, 2021 22:02:01 GMT -5
Hi Lonnie. I'm trying to understand the -20db limitation. Why would sources output only a -20db and not a 0db or close to that digital signal? Are you counting lossless or music sources in this? I have numerous music sources that easily go past -20 db peaks digitally. Because you need headroom. A nominal output level of -20db gives you 20db of headroom before everything runs out of steam. Digital parts cannot go above 0, they can only go negative. If your source was set to 0db then you have no headroom and everything would be clipped and distorted. Hope this helps. Lonnie Hi Lonnie. Appreciate you participating in this thread. Are you saying that the devices the XMC-1 is connected to send a max digital signal of -20db? Which devices are these? All of them? How is this different from say what a two channel dac receives from a CD player or music files? |
|
|
|
Post by garbulky on Sept 9, 2021 22:42:57 GMT -5
One of my favourite reviews on ASR is of a PS Audio DAC. www.audiosciencereview.com/forum/index.php?threads/review-and-measurements-of-ps-audio-perfectwave-directstream-dac.9100/ It's interesting because: a. the DAC is very expensive. Way more than I would ever consider paying for a DAC, and I am not alone there. b. it performs very very poorly in the tests. Outstandingly badly in fact. It performs poorly because Amir tests to an engineering standard that assumes higher scores equal better engineering. As far as it goes that may be true, but the PS Audio units are engineered to perform exactly as they do, because regardless of the test results people are paying the big bucks because they absolutely love the sound the get from these DACs. I mean absolutely. So IMO the cause of the test failure is not bad engineering but good engineering. ASR and PS Audio have very different objectives, it just so happens that the equipment is the same. Same here. The sound of...distortion? From a DAC? I saw their measurements and the highest harmonic is still -80db so that is probably inaudible. |
|
|
|
Post by autocrat on Sept 10, 2021 1:50:03 GMT -5
One of my favourite reviews on ASR is of a PS Audio DAC. www.audiosciencereview.com/forum/index.php?threads/review-and-measurements-of-ps-audio-perfectwave-directstream-dac.9100/ It's interesting because: a. the DAC is very expensive. Way more than I would ever consider paying for a DAC, and I am not alone there. b. it performs very very poorly in the tests. Outstandingly badly in fact. It performs poorly because Amir tests to an engineering standard that assumes higher scores equal better engineering. As far as it goes that may be true, but the PS Audio units are engineered to perform exactly as they do, because regardless of the test results people are paying the big bucks because they absolutely love the sound the get from these DACs. I mean absolutely. So IMO the cause of the test failure is not bad engineering but good engineering. ASR and PS Audio have very different objectives, it just so happens that the equipment is the same. Same here. The sound of...distortion? From a DAC? I saw their measurements and the highest harmonic is still -80db so that is probably inaudible. Yes I know. But the rationale for using SINAD as a comparative measure is it enables you to conclude that the better the number the better the engineering. And conversely. My point is that in this case the engineering is labelled bad by default when in fact from the point of view of both the manufacturer and the users the bad score is a result of deliberate design decisions that improve the product. I could be overthinking it of course. I'll shut my cakehole now. |
|
ttocs
Global Moderator
I always have a wonderful time, wherever I am, whomever I'm with. (Elwood P Dowd)
Posts: 8,154
|
Post by ttocs on Sept 10, 2021 6:41:47 GMT -5
mmmmmm . . . . . . . . cake.
|
|
