This just goes to show that different people take different things from these reviews (nothing wrong with that). Personally, these reviews come across as focusing on technical performance, not a “…complete picture what a piece of audio equipment will be like to live with and play music through…“.
Different strokes for different folks. 
As an owner of Mytek gear, I’m a bit disappointed in this analogy. Surely an experienced audio designer sees the difference between an instrument, a source of sound, and reproduction technology intended to faithfully reproduce a recorded instrument?
That puzzled me as well. It’s not even internally consistent: if you value the sound of the Stradivarius itself, surely you wouldn’t want to purposely change it while playing it back, would you?
I guess I don’t see any of your examples as “problems”.
I actually believe that symmetrical spikes around the main tone (even if inaudible) can nevertheless be “unfortunate”. And that linearity which is good enough for an AVR (even thought it’s MUCH worse than separates) merits a comment 
And Amir DOES reach out to manufacturers for comments - they sometimes don’t respond, however.
I’m not saying Amir is the greatest or that measurements are the only thing that is important (there’s always features, price vs. performance, even appearance), but I also think measurements are the MOST likely to provide objective information about how a system or component WILL perform IRL.
I just realized the thread got split. I’ll repost here. My discussions with Amir are not about if his measurements are valid or not but about Amir’s arbitrary assumption that good sound=minimum measurable THD. I challenged Amir on this with clear example of analog tape. Rather than having a discussion about why analog tape measures worse, but sounds better than a CD, he continues with his dogma of SINAD which states the opposite. My whole point here is that yes, these type of measurements provide valuable information but then that information has to be INTERPRETED and CORRELATED with what someone can hear. I don’t believe Amir offers any hypothesis like this in his reviews. Instead Amir arbitrarily says Good/Bad based on his own invented fundamentalist criteria. He says he listens to half(!) of products he tests (!). How can you be missing this crucial element of scientific testing when you claim you are testing the sound quality which implies that “sound quality” is what we are hearing?
That’s why this isn’t SCIENCE. Science is well defined process of hypothesis, testing and conclusions that can sustain peers reviews. Amir fails here for now.
I have now stated out my point and am bailing out from discussions with Amir until it’s a normal professional dialog without obvious holes in the scientific process that were pointed by me and many others. Amir needs to plug this holes first and present a rigorous scientific process that’s immune to scrutiny.
Again, all about perspective. To use the fruit analogy, compare apples to apples; to use automobiles, of course a Honda is likely to fall behind a Ferrari in some measurements.
I don’t believe I said there isn’t a value in publishing measurements in general. But add perspective. If something is admitted to be inaudible, or is perfectly normal and expected in a category, or at a certain price point, SAY IT. Otherwise you are creating at best a false choice, or at worst the impression that a product is somehow defective.
All too often ASR seems to be an exercise in making the perfect the enemy of the good.
I’m curious about a couple of things.
“Better” implies some scale on which the two are measured. What is that scale? How is the measurement performed?
Does anyone have any information on the physics of why this would be so?
To me, “tape is better than CD” is am example of a dogma that doesn’t have any theoretical or factual basis. Subjective opinions and sighted tests don’t count as evidence.
I can’t speak for Amir, but I don’t think that what he assumes. I agree a good THD is not a sufficient condition, but I believe it’s a necessary one.
I hope Amir will look at this as a valuable feedback. I don’t think his reviews are perfect either, and sometimes I disagree with his recommendations, but I always find his measurements valuable, and I can make my own educated opinion.
As an owner of Mytek gear, I’m a bit disappointed in this analogy. Surely an experienced audio >designer sees the difference between an instrument, a source of sound, and reproduction >technology intended to faithfully reproduce a recorded instrument?
This example was meant to show that the concept of “good harmonics” which are all around us in every sound is a part of our natural world, and that the ASR idea that all distortion (which includes these harmonics) are BAD is just a mistaken interpretation of reality. Because this is the foundation of ASR assessments, these assessments are in consequence erroneous. When I say “good distortion” I’m being ridiculed by ASR, which only proves how little understanding of sound science ASR members have.
Anyone who has produced music knows that each piece of equipment starting with microphones has its “sound”, ( meaning distortion) and understanding and managing these “sounds” or “distortion” is part of creative production process that in the end results in a recording that is full of all kinds of distortion. Of course very often, especially with acoustic or orchestral recordings transparency is desirable but it’s usually not accomplished by choosing equipment with the best SINAD but rather, as it’s a very complex process, it usually depends on recording engineer experience. Some of the best acoustic recordings may have been recorded with just two microphones, that happens.
Now, Brooklyn Bridge II include the “normal” (transparent) and “warm” (addtnl 12dB of even harmonics) mode of playback. This is a consumer device, meaning it’s designed to provide a good experience of listening. In our experience most users prefer the “warm” sound, they consider it more “musical”, they simply like it better. Since their satisfaction is the goal, this function makes sense in this case.
However, if a mastering engineer used the same DAC to asses their mastering results, I’d probably discourage them from turning on HAT in this case, as it will skew their mastering result. In other words with HAT on their master may end up sounding thinner than with HAT off. Make sense , right?
From 1990 until 2010 Mytek designed ADCs and DACs exclusively for profesional recording and mastering and we did well.
My design philosophy for these was, during the R&D process is to always A/B compare these two signal paths:
We listen to this, tweak ADC and DAC/ measure too, listen again/ tweak again and so on. Sometime we make circuit changes that don’t show in standard measurements but are audible.
This listening requires a very very hi res monitoring, speakers and amps and headphones like Staxes. The higher resolution of monitors the better you can tell the difference. And there always be some you want to get as close as you can.
Everything has a “sound”. Even a wire has a bit of its own character. Mytek has over the years defined “Mytek Sound” and BBridge has a version of this. When I design a DAC for this sound it’s I’m focusing on transparency, detail and transients, while trying to maintain “musicality” (opposite to clinical/boring which could be a result of approach like Topping). There is usually a sweet spot between the two - "Id make the consumer that probably a bit more musical vs a mastering DAC more precise.
Having said that our consumer DACs are used for mastering, for example Abbey Road has two of the Manhattan DAC IIs and so do severa other mastering studios.
You see, when thinking about “sound quality” you ought to be thinking about the final goal you are trying to accomplish. The goals are different for recording, different for mixing, yet different for mastering and completely different for playback. You ought to use your ears and common sense to get results in all these.
If measurements provided results in a recording studio, everybody would be using these. But none of studio recording sessions starts with measuring the recording chain for minimum distortion. Why? Because it wouldn’t provide the desired result.
Now on my point “Sometime we make circuit changes that don’t show in standard measurements but are audible”. This points to another problem in measurement only approach.
Nothing beats the ears- We can hear things that we cannot measure and this is another problem with Amirs approach. He misses all this that cannot be measured today. Rob Watts of Chord claim that he can hear filter difference down to almost -200dB actually surprisingly is true, but you can’t measure this with AP.
For Amir thing he did not “measure” just do not exists. This approach is equivalent to throwing out scientific data which is a straight road to scientific mistakes or even fraudulent conclusions…
Let’s say Amir measured an FFT of a tape with 60dB DR and saw the noise floor at -60dB. Burried in this noise would be more audio , as much as 40 dB+ down into the noise, yet you would not see this in the standard FFT because it’d be masked by noise. I’m pretty sure that Audio Precision is working on a new generation of intelligent algorithms that will eventually discern the useful signal from “noise” and I’d soon expect a slew of new DSP tools that has a way more sophisticated approach to audio measurements that includes brain science and goes beyond the old SINAD. But until these appear, I’ll reject SINAD as a benchmark for sound quality.
The real audio/sound science is way richer and way more interesting then the boring old simplistic mantra of SINAD at ASR forum. If someone wants audio science, goto aes.org , organization funded in 1948, become a member, start going to conventions, read papers, participate there.
ASR is in comparison an amateur attempt of usurp the sound science authority title by one person who can’t even pass the scientific rigor test. And the worse part is, it is deliberately setup as a hostile place, antithesis of what an open scientific environment needs to be.
I love that this debate is sitting in the Off Topic section, but in reality it’s fundamental to Roon’s promise to have Audiophile Grade Sound Quality. I very much appreciate that @Michal_Jurewicz1 and @Amir_Majidimehr are getting into the nuts and bolts of the great subjective vs. objective debate. For an especially eloquent description that I decided to re-read today and to help orient others who may not be familiar with it, see here: NwAvGuy: Subjective vs Objective Debate
I dunno, to me the thing that sticks out in this entire discussion is that dozens of people posted almost in tears, frustrated that they could get no response at all from this company. Once in a blue moon, someone from the company would drop into the thread to give some vague responses and promises, followed by dozens more owner posts saying none of the promises were kept.
Yet now all of a sudden, the company officers have all the time in the world to engage in one on one verbal jousts. Wouldn’t the time be better spent addressing the customer service issues reported by their users?
I considered a Mytek at one time and came close to pulling the trigger. I evidently dodged a bullet by going with a different manufacturer instead.
There’s a stunning level of misdirection here. On the one hand, you are adding distortion to give a sense of warmth, then you also appear to be declaring that there is a deep well of usable information in the noise that you are somehow tuning for? By ear because an AP is limping along with limitations? You cap it off with appeals to authority that are mostly irrelevant; if Abbey Road had good measurement data available, they might not have chosen your products.
Also, the debate over “science” in ASR is largely irrelevant. Call it Audio “Objective Measurement” Review if it is too irritating. Science has many flavors, including survey naturalism and theoretical cosmology that huddles at the border of falsifiability. Amir has surveyed an enormous range of equipment with the hypothesis that measurements can lead to determinations about the relative high fidelity of the gear. Along the way, there are wide-ranging discussions of Harman curves, audibility, echoic memory, statistical reliability, and so forth. It’s science and measurement-focused, so close enough.
I both believe this and absolutely do not believe it. We hear with both our minds and our ears/skin and other sensory apparatus. To the extent our minds and chemical balances and preconceptions and moods affect our hearing – yes, we hear things we can’t measure with instruments. And that’s always changing. You can’t “capture” a particular balance with hardware that’s relatively inflexible, certainly not with just two “normal” and “warm” settings.
But what we can measure, accurately, is whether or not a piece of equipment transparently reproduces the digital input signal given to us. That’s what Amir is measuring, as I see it. Do I wish he measured more facets of the equipment? Yes, I do. I’d love to see some torture tests of the various inputs to a DAC, for instance, to give us some information about how good the USB inputs are vs. the S/PDIF or AES inputs.
This whole “circuit changes that don’t show in standard measurements but are audible” is interesting. You say that we wouldn’t see buried audio in “the standard FFT”. How would we then hear it, if it’s buried in the noise? I’m not sure.
But what bothers me more about this whole response of yours is that you continually mix up source recording with source reproduction. Sure, there are various facets of recording and mastering that require human hearing, taste, and judgement. Introducing “musicality” is a critical part of the mastering process. But accurate reproduction of the recorded signal shouldn’t have any distortion, IMO. I’m trying to understand your argument here, and just not quite getting it.
I’m no expert on measurements, but I’m baffled how Amir can criticise the BB2, something he still hasn’t heard, and of all the posters who voted slaughtered it too I’ll bet less than 1% of them have actually heard it. Amir defends himself as to why he decided not to listen, but bottom line is he hasn’t listened to it. It’s like criticising a country or city because of bad reviews on Tripadvisor but not actually visiting yourself to experience it, or choosing not to eat in a restaurant because some random person didn’t enjoy the food. Experience things yourself first, it’s not rocket science (no pun)
I’m thinking… So what? Why would anyone care? Why would you trust what someone else thinks they hear? Especially someone you don’t seem to trust in the first place?
Because its the topic de jour Bill.
And why would you slag something you havent actually seen ,heard,smelled, touched. It’s bizarre.
And I’m not criticising what they “think they hear”. That is the point, they haven’t heard it.
why analog tape measures worse, but sounds better than a CD
“Better” implies some scale on which the two are measured. What is that scale? How is the >measurement performed?
To me, “tape is better than CD” is am example of a dogma that doesn’t have any theoretical or >factual basis. Subjective opinions and sighted tests don’t count as evidence.
Well, historically, what happened, when CD came to recording studios in late 1980s it was hailed as perfect sound, perfect copies, 96dB dynamic range vs the hissy tape etc etc. And it was new and welcomed until engineers started using it and then quickly discovered that it comes with its own other set of problems like a massive phase shift at high frequency or, fade out zipper noise from undithered ADCs and later jitter that was not really fully understood until about 1994. When I was at the Hit Factory Studios in 1989, the digital filters were the thing. For example Apogee Electronics (Bruce Jackson) made replacement antialiasing filters for digital Sony tape machines. We would install these in the tape machines- they sounded better (at that time Sony tape machine had NOS ADCs). That’s how I got into making the ADCs- there was an opportunity to make a better ADCs and I started making them right there in 1990 for various NYCity recording studios. Mark Levinson, the man, participated in this too. Then there were early converters from DCS, Prism, Apogee and also Mytek because studio needed better digital sound. Very early on we knew digital sounded different than analog, and for a good decade mixdown was done in parallel to digital and to the standard 1/2" 30 IPS tape just in case. Thanks God we also printed mixes to analog, as some record labels would transfer analog masters to digital and then throw tapes out. At about that time Michael Gerzon , my hero and one of the geniuses of audio science wrote a paper, shortly before his untimely death, “Don’t Destroy The Archives!” Re: [ARSCLIST] Analog: A Race Against Time
What the paper says is that the master is the master - keep it, because you don’t know the new digital medium good enough yet to fully trust it.
Now, that I have to put this 1/2" 30 ips tape in this historical context, why does this studio tape sounds better than the CD? The fact is that a lot of engineers, producers and musicians over the last 40+ year said that it does sound better. There are experienced enough and there is enough of them, that most people either agree or accept it if they can’t hear it. There are no single measurements like the ASR Sinad to prove it. I’m sorry. It’s only these people ears. You are free to have have a different opinion, but it pretty much is a consensus in the recording industry.
Here is my hypothesis, why analog tape sounds better than a 16 bit/44.1 digital (higher res digital, especially 192k+ and DSD can be superior) with factors in order of importance:
Analog 1/2" tape has more of finer resolution. Despite noise, it does seem to resolve the detail well below the 96dB threshold of 16 bit audio. Low level details of sound (ambience, air are better on tape than on even the best dithered 16 bit audio)
Analog tape being analog has a natural harmonic structure meaning tape distortion are correlated with music thus giving it harmonic character. Digital on other hand is a modulation (PCM=Pulse Code Modulation) where distortion are often not directly correlated to sound. Take example of a steep 100dB FIR filter at lets say 20k which Amir was advocating as necessary for producing better DAC measurements. This thing produces a so called preecho, meaning you hear an echo, the actual sound and then the echo again. Following the thread of evolutionary science: There is no preecho in nature - this type of distortion is completely unnatural and did not exists before digital audio.
Linear Phase ; About FIR and IIR Filters - AKIRU MUSIC WORKSHOP
Our brain doesn’t know what to do with it and unlike with a regular harmonic which is just a coloration of main sound, our brains deems it extremely annoying - this is one of the examples what “digital sound” sounds like and why it’s often annoying despite artifacts being below 100dBs. I could go on and on about other examples of other non-harmonic byproducts of bad digital.
The total sonic effect is a combination of 1 and 2 where analog tape has more resolution and more distortion but they are natural harmonics and less annoying and the digital which has much less distortion but they are much more annoying as they are not natural.
This is yet another example to prove how nonsensical is a THD measurement in this context. Tape distortion are way better than digital distortion and should be treated accordingly in listening evaluations (btw a similar modulation effect is also true for Class D amps).
Now, if one understands all this, with high sampling rates (192k+) , high resolution (32 but quantization and 64 bit processing), proper algorithms (that don’t truncate), using proper filtering, digital can be mitigated and achieve spectacular results and today we are definitely at a stage where good digital beats analog tape. But it has to be properly executed hi-res digital. Most todays studio production is going full digital because it’s easier, cheaper and can sound very good (if you know what you are doing).
Happy Reading , Michal at Mytek
Thanks for explaining it, and for taking all this time. It’s great to hear someone explain some of these things. Of course, I can’t help thinking that over 40 years, things have changed, as you say below. To often people make a decision early in the history of a technology, and then forget to update as time winds on and things change.
So you’d say it’s still impossible to make good Redbook CDs (as good as good tape), as the resolution just isn’t there in the format?
This is a basic fact of tape recording. Human brain discerns the correlated music from the background noise while AP FFT does not. 40dB+ deep into the noise. The evolutionary theories give an explanation of our ancestors sitting next to a waterfall with its white noise background but being able to hear a predator noises that are lower than the background.
However we came to this, it does work this way.
Brain is an amazing machine; think about it: we feed it a wavefrom through our two ears, and suddenly the brain discerns it into a beautiful 3D soundstage where we can hear instruments closer and further, point where they are, how far, separate them. Our brain will plug in dropouts from a memory. All from this one waveform. When you compare what brain can do vs what AP can do today there is no contest at the moment. There will be a long way before computers can do this. There is a lot of cool popular science videos on the subject of brain sound and vision, like PBS NOVA, highly recommended.
It truly is amazing.