Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
Audio Science Review of Expert 200.
@Confused

I'm not going to quote your long message but I will pick up on a couple of points.

In some ways one of the questions you're asking is about what measurements are relevant to what we hear and how relevant are they? Is every measurement equally relevant to what we hear. Does, for example, it matter whether an amp can deliver its full rated output on test tones at 19 and 20 kHz as one example of the relevance of a specific measurement.

So. taking the 19 and 20 kHz test tone question and looking at it from a practically objective viewpoint as opposed to a theoretically objective viewpoint, I'm going to say that it doesn't matter that an amp can't deliver full output at those frequencies. There are 2 reasons I have for saying that and they both relate to real world practicalities.

First, can most of us hear anything at those frequencies? The answer to that is no. While the human frequency range is quoted as extending from 20 Hz to 20 kHz, dig a bit deeper into the data on that and you find that we start losing the high frequencies very early in life and that the ability to hear 20 kHz is found only in the young with many being incapable of hearing 20 kHz at age 20 or so. We've got a poll here on the age of forum members and only 2.15% of respondents were under 34. Add the 3.82% aged between 25 and 30 and we/ve got less than 6% of respondents aged 30 or less. Over 50% are aged 45 or older. Most of us can't hear 20 kHz or anywhere near that.

Second, what content at 19 kHz and higher is present in music and at what level is it present? The highest fundamentals in music are between 7 and 8 kHz and everything above that is overtones which are at lower levels than the fundamentals. High frequencies are easily absorbed, even being absorbed by air, so the high frequency overtones we hear when listening to live music are down in level compared to the fundamentals anyway, simply because of distance and absorption in air. We listen to music at home from our speakers at much closer levels than we tend to listen to live music unless we play in a band or are sitting close to someone singing or playing so we can expect the level of the overtones at 19 kHz and higher to be higher in level relative to the fundamentals than they would be when we're listening to live music but even so the level of the high frequency overtones produced by voices and instruments is still lower in level than the fundamentals so the amplifier is never going to be called on to deliver enough power for any musical content at 19 kHz and higher to be delivered at the same sound pressure level as the fundamentals an octave and a half or more lower in frequency.

That leaves the question in relation to those 2 frequencies of intermodulation distortion, the reason for that particular test, because the IMD spectrum for those 2 frequencies is a distortion product at 1 kHz and at every 1 kHz above that. The simple fact, however, is that those distortion products are at much lower levels than the test tones and that means that they're also going to be at much lower levels than the actual content at those tones in music. In practice I suspect that the IMD products of those frequencies in music are going to be inaudible, especially given the IMD performance of Devialet's.

Put all of that together and the inability of the amp in ASR's test to deliver full rated output at 19 and 20 kHz is highly unlikely to affect our enjoyment of music because the amp is never going to have to deliver that output when playing music, even when it is delivering full output at frequencies where musical fundamentals are present, and because in practice most of us can't hear those frequencies anyway and the intermodulation distortion products of the frequencies in that range actually present in music are going to be inaudible in an amp with the Devialet's IMD performance

So what I'm saying is that not all measurements are equally relevant or important when assessing the performance of an amp in real life delivering music. Knowing what measurements are relevant and how important those measurements are is something most of us rely on the reviewer to tell us. You mentioned the test results of John Atkinson and Paul Miller and the fact that neither were overly concerned about similar measurements in their tests to those reported in the ASR review. In fact, my understanding is that Paul Miller actually uses a Devialet in his own system. Both Atkinson and Miller have extremely good reputations as reviewers, earned in part by their ability to relate their measurement results to what people are likely to hear when using a given product. Amir at ASR does not have such a reputation. I'm not questioning his measurements, the fact that both Atkinson and Miller measured similar things indicates that his measurements are probably in the ball park, but I would say that the fact that neither Atkinson nor Miller was overly concerned about their measurements is a lot more important than the actual measurement itself. Stereophile gave both the D-Premier and the 140 Pro an A rating and Miller apparently uses a Devialet himself. Actions speak louder than words as the saying goes, and those actions carry a lot more weight in my view than the ASR review's words.
Roon Nucleus+, Devilalet Expert 140 Pro CI, Focal Sopra 2, PS Audio P12, Keces P8 LPS, Uptone Audio EtherREGEN with optical fibre link to my router, Shunyata Alpha NR and Sigma NR power cables, Shunyata Sigma ethernet cables, Kimber Select speaker cables, Grand Prix Audio Monaco rack, RealTRAPS acoustic treatment.

Brisbane, Qld, Australia
Reply
Gerronwithit Wrote:
Confused Wrote:Overly long post redacted .....
Phew, that was a long one! If I decided to break down all the components that make up my BMW and subjected each one to varying subjective or objective tests would that colour my perception or otherwise of the whole car. The appreciation lies with the complete package.

I know that engineers and scientists have performed all sorts of tests and analysis on the materials that make up the components that make up the complete package, such as the engine, the gear box, the differential, the brakes etc. etc. and if a charlatan like Asrim, or whatever he is called, does some sort of pseudo scientific analysis on a failed machine I would absolutely disregard his so called critique. If Devialet do not deliver a reliable complete product we paid for they will pay for it in the market place.

I'd like to start with a thank you.  A couple of weeks ago I was not having a particularly good day.  Anyway, having sorted a few issues out, I checked into Devialet Chat for a bit of a distraction, and the first thing I read was a post from yourself.  I may have sounded overly positive about ASR in my post above, but even during a particularly challenging day your post (link below) made me laugh out loud, so many thanks for that!

https://devialetchat.com/Thread-Audio-Sc...5#pid92005

Regarding your more recent post above, this is a slightly unfortunate analogy for me.  The thing is, I have modified my BMW.  I really did not like some aspects of how it used to drive, specifically the transmission and gear change.  I ended up removing something called the clutch damper valve, which is a bit like the thing that dampens the tonearm drop on a turntable.  I also modified the differential mountings, taking out the voids in the void bushes.  I really like the car now, and I think these mods have resulted in me keeping it a lot longer than I otherwise would have done.

Curiously, I have felt zero desire to modify my Devialet.   The ultimate listening machine.  Angel
1000 Pro - KEF Blade - SOtM sMS-200Ultra Neo - tX-USBultra - Mutec REF10 - MC3+USB - Pro-Ject Signature 12
Reply
@David A - I agree with you, I am quite sure that the ability to run a high rated power output at high frequencies is utterly irrelevant to sound quality. This was part of the point I was making regarding the power of the words that surround the measurements, Stereophile write "a peak level of 20W into 4 ohms (the highest level the amplifier would deliver with this signal without the power supply collapsing)" and nobody notices or cares. Posters on ASR write something like "ha ha ha - a ten-thousand-dollar amp that collapses at 6W at 20kHz - French fail!" - Then suddenly we take notice. Does it matter to sound quality? No, I cannot see that it does. Plus, we still do not know why the ASR reviewed amp failed. I think it is best to revisit this point when we have more information. In terms of what really interests me, see my comments below.

@ssfas - I agree with your subjective observation regarding transient speed. This is one of the things that appealed to me about the Devialet during my very first audition of a D240 Expert, it just sounded more exciting and alive versus some similarly priced rivals. This is more in line with the question that interests me, what is it in the measurements that reveals this aspect of amplifier performance? In a similar way, I found the Expert to be maybe a bit smoother and more relaxed than my current Pro, whereas the Pro seams more detailed and revealing. What in the many available measurements reveals the reason for this?

Furthermore, I am pretty certain that if you listened to two amplifiers, one with a signal to noise ratio of 130dB, the other 115dB but otherwise identical in performance, it would be impossible to tell them apart. Yet on paper, the first example is clearly superior. The same thing with THD figures of 0.0001 versus 0.01 and similar.

As another example, there was one time I listened to a Chord SPM1200 back to back with a Bryston 4BSST, with no other changes made to the overall system. To me the Bryston has slightly more pronounced mid bass, whereas the Chord seemed to go deeper with better bass definition. I note that the Bryston has an output impedance of 0.06 ohms, the Chord slightly better at 0.04 ohms. Is this the reason for the subjective difference? Who knows? I could also say the Devialet Expert is the best of the lot for this aspect of bass depth and accuracy (which I subjectively believe to be true), and look, it has an output impedance of 0.01 ohms. So here we have full correlation between my subjective observations and the measurements, the lower the output impedance, the better the low bass accuracy. The trouble is that this does not prove causality, it might just be coincidental, or my ears and brain tricking me. I suspect there is something in this, but as I said before, who knows?

From what I have read to date, nobody has a clear answer to the question of exactly how all aspects of sound quality relate to specific measurements. Lots of ideas, speculation and opinions, but nothing definitive. In fact, I suspect that it would take a considerable amount of very expensive research to resolve this one

This is why I think sites like ASR would be a lot more interesting if there was a greater mix of the objective with the subjective, to try to improve understanding of how they correlate. Meanwhile, just stating amp A crushes amp B because it has a better SINAD number is a little misleading. OK – ASR measure a lot more than this, and some ASR posters appear to be very knowledgeable. Unfortunately, one or two ASR posters seam to fail to realise the full complexity of the subject, or at least do not wish to debate the subtleties, and this combined with an apparent joy for arguing and trashing the reputation of anything expensive, it can get a little ugly, and this shuts down the more nuanced debate. I think with a slight change of attitude and moderation, ASR could be excellent. However, this does not look likely, so we are left with the ASR we have, it is what it is.
1000 Pro - KEF Blade - SOtM sMS-200Ultra Neo - tX-USBultra - Mutec REF10 - MC3+USB - Pro-Ject Signature 12
Reply
You will kill a lot of tweeters giving them real 10 Watts at 20 kHz sinus signal without crossover, just my words. Devialet Watts are RMS - pink noise. All good! Best natural instrument for measurement are our ears.




Gesendet von iPad mit Tapatalk
Reply
Maybe these measurements are also helpful https://www.stereo.de/hifi-test/produkt/...0-pro-1585
Reply
(22-Apr-2020, 23:26)markush Wrote: Maybe these measurements are also helpful https://www.stereo.de/hifi-test/produkt/...0-pro-1585

Thanks for posting @markush.  

I am finding myself fascinated by measurements at the moment.  I know some people might think this is a little sad, but at least it something interesting to study during "lockdown".

I actually recall seeing the above measurments on another thread, where a few of us at the time were obsessing about power outputs.  The stereo.de review was notable as one of the first to measure the 220 Pro, and it was particularly notable that it achieved 170w / 8 ohms, 343w / 4 ohms, which is arguably superior to the D-Premier, which managed 180w / 8 ohms 245.8w / 4 ohms, and certainly a big step up from the 120w / 8 ohms, 240w 4 ohms HFN measurements for the D170.  How much this matters is debatable, but it does hint at how the amp and power supply design has improved over time.

For those that might be interested, I notice that Stereo have also published measurements for the 210 Pro dual mono

https://www.stereo.de/hifi-test/produkt/...-dual-1807

There are a few things I noted in the Stereo measurements:

They only measured via the analogue inputs for the 220Pro, which as we have seen elsewhere do not measure quite as well as the digital inputs.

The 220 Pro measurements show S / N ratio Phono MC (0.5 mV / 10 ohms) at 5 watts: 74 dB (A).  This appears to be a worse result than anyone one of us achieved in the thread linked below:

https://devialetchat.com/Thread-Phono-st...asurements

This makes me wonder if there is some variation in the performance of the phono stage between units?  Or maybe just a quirk of how the measurements are reported?  I note that those who seemed most concerned about hiss never actually made a recording to send to Mathieu Pernot.  Maybe we are missing some interesting data here, that is the worst performing phono stages might not have not been measured yet?  Just idle speculation on my part.  (although I recall @Greg was waiting for a friend to perform the recording, and having a friend visit would not currently be legal or advisable during the UK's "lockdown")

The 210 Pro measurements do include a signal to noise ratio measurement via the digital input, 115dB, somewhat better than the analogue input measurements.

The 50 m/w signal to noise ratio for the 220 at 76db is fractionally superior to the 210 at 74dB (nothing in it really), but at 5w the 220 is even further ahead at 95dB versus the 210's 84.  I find this point interesting as it implies that the 220Pro is the superior amplifier, but I suspect the 210 may be the better sounding amp for other reasons.

Whist posting nonsense about measurements, I can advise that I have been having much fun this morning both listening to music and looking at the various amplifier measurements available on Paul Millers AVtech site.  I was thinking about the Devialet's quoted 130dB S/N ratio figure and the fact that Paul Miller measured 118dB for the D170, and I was wondering how rival products measure in this regard.  The interesting thing is that I could not find anything with a lower figure than 118dB, and nothing remotely close to 130dB.  As I mentioned in an earlier posts, you can match Devialet's measurements if you are very selective about parameters.  For example, Miller quotes a S/N ratio of -133.9dBV @ 40Hz, but quoting a S/N ratio of 130dB but only below 45Hz seems a bit tenuous.  The point is, the Miller measured S/N ratio at 118dB is excellent, and there are very many highly regarded products measured by Miller that are much worse in this regard.  So why quote figures like 130dB?  This does open Devialet to acusations of quoting meaninless numbers which cannot be backed up by measurements, which seems a bit daft to me when the actual figures backed up by recognised peramiters are themselves excellent by any satandards. 

The AVtech measuremnts are only available for reviews 2013 and earlier, so do not include the current ASR favoourite of the Benchmark AHB2, but the HiFi News review is available online, and reports a measured S/N ratio of at 0dBW / 100W = 106.0dB / 126.0dB.  So yes, the AHB2 does have better numbers, for noise at least, and is just about the only thing that I have seen that gets anywhwere near the S/N ratio numbers quoted by Devialet.  In fact, mindfull of all the other measuments of decent kit I have looked at this morning, the Benchmark numbers are pretty remarkable.  I wonder what it sounds like?

https://www.scvdistribution.co.uk/files/...nt_1_3.pdf

https://www.stereophile.com/content/benc...-amplifier
1000 Pro - KEF Blade - SOtM sMS-200Ultra Neo - tX-USBultra - Mutec REF10 - MC3+USB - Pro-Ject Signature 12
Reply
(20-Apr-2020, 08:12)Confused Wrote: @David A - This is why I think sites like ASR would be a lot more interesting if there was a greater mix of the objective with the subjective, to try to improve understanding of how they correlate.  Meanwhile, just stating amp A crushes amp B because it has a better SINAD number is a little misleading.  OK – ASR measure a lot more than this, and some ASR posters appear to be very knowledgeable.  Unfortunately, one or two ASR posters seam to fail to realise the full complexity of the subject, or at least do not wish to debate the subtleties, and this combined with an apparent joy for arguing and trashing the reputation of anything expensive, it can get a little ugly, and this shuts down the more nuanced debate.  I think with a slight change of attitude and moderation, ASR could be excellent.  However, this does not look likely, so we are left with the ASR we have, it is what it is.

I popped in to ASR the other day - no reports on Devialet returning the Expert unit that they blew up.

The site seems to be dominated by a few people who are measurement fundamentalists and within a few posts, if not one post, the quite pointed view is basically that everyone should have an RME DAC, Benchmark amplifier and Genelec or Neumann speakers. 

My view is that measurements are the beginning, not the end. There are enough products around to ignore the ones that measure bad and choose from the ones that measure well. It doesn't mean we should only buy that which measures best. We buy audio because we like to listen to music and if we like how something sounds subjectively, it is the height of arrogance for anyone else to say otherwise - which goes on a lot at ASR - in fact take that attitude and you will quickly get banned. They don't take much provoking. 

There is also the touchy issue of money. OK, pro audio is very price competitive. It's a hard-nosed global business. Consumer spending is far more discretionary and if I or anyone else wants to spend a chunk of money on home audio no one else can tell me not to, or deny that may give me pleasure. The ASR view seems to be that most consumer audio is over-priced. Well, fine for them, they don't have to buy it.

Finally, the Benchmark AHB2 that they rhapsodise over is not cheap - £3,500 in the UK - and provides 100w into 8 ohms, barely enough for my speakers to perform at their best. The Devialet 140 is slightly more powerful and contains a huge amount of additional electronics (streamer, DAC, phono etc.). It costs £4,500 - so in my book is far better value for money.
250 Pro CI; Innuos Zen Mk3; Claro dual turntable (Koetsu + Ortofon Mono); Harbeth SHL5plus 40th Anniversary; Townshend Isolda EDCT speaker cables; Shunyata UK6 + Delta NR
Reply
It's not the tests Amir does that is the problem, it's the hardline followers of ASR and Amir's negative attitude towards anything that is "subjective" by his own definition. I like measurements myself and have always checked measurements in Stereophile reviews. Now also at ASR. They both use Audio Precision equipment for tests, but their interpretations differ. John Atkinson has a lot of experience and most of the time I agree with his views.
*
Devialetless!
Roon, Core on NUC with Debian/Audiolense XO/Music on NAS/EtherRegen/Pi4+RoPieeeXL/USPCB/ISORegen/USPCB/Sound Devices USBPre2/Tannoy GOLD 8
250 Pro CI, MicroRendu(1.4), Mutec MC-3+USB
Reply
(13-May-2020, 14:37)ogs Wrote: It's not the tests Amir does that is the problem, it's the hardline followers of ASR and Amir's negative attitude towards anything that is "subjective" by his own definition. I like measurements myself and have always checked measurements in Stereophile reviews. Now also at ASR. They both use Audio Precision equipment for tests, but their interpretations differ. John Atkinson has a lot of experience and most of the time I agree with his views.

They got the unit back. It was blown up under unofficial testing, it was not under warranty and was repaired free of charge with doo-to-door collection in 3 weeks during a pandemic. 

And people are still complaining, calling it a disaster. 

Some people will never be happy. But I am, as I have a 250 CI and I and my family enjoy it tremendously!
250 Pro CI; Innuos Zen Mk3; Claro dual turntable (Koetsu + Ortofon Mono); Harbeth SHL5plus 40th Anniversary; Townshend Isolda EDCT speaker cables; Shunyata UK6 + Delta NR
Reply
(14-May-2020, 13:37)ssfas Wrote: They got the unit back. It was blown up under unofficial testing, it was not under warranty and was repaired free of charge with doo-to-door collection in 3 weeks during a pandemic. 

And people are still complaining, calling it a disaster. 

Some people will never be happy. But I am, as I have a 250 CI and I and my family enjoy it tremendously!

Seems they got the unit back with zero communication from Devialet.
No comment about why it had failed.
No comment about the measurements that seem to make a lie of Devialets claimed specs.
Why?

Poor form on Deviaet's part. But I guess we've all become inured to their lack of communication.
Reply


Forum Jump:


Users browsing this thread: 1 Guest(s)