Piano World Home Page Forums Our Most Popular Forums Digital Pianos - Electronic Pianos - Synths & Keyboards Which headphones do you use/recommend?

None whatsoever - but I have a brain, and as a software engineer, I am pretty sure that when the a job of a given component is to

• take digital data (for example, read data from a cd/dvd),
• parse it (digitally), according to clearly specified algorithms (like decode FLAC or AAC audio), and
• output it in a digital format (like S/PDIF),

... then putting a f***ing piece of tape on the box (to "harmonize the vibrations") will. not. do. anything.

The output is either correct (bit-by-bit), or it's wrong. There is no other option, there is no way to "improve" it.

Kristof

ps. Sorry, I am always deeply pissed off when I see cheaters making money out of the ignorance of others. (Of course I don't mean EssBrace here; I am talking about the ones selling stuff like that.) Mentioning the topic stirred up my memories ... my anger is not directed at my fellow pianists.

That reminds me of this strip:


(If above link is broken, see here.)

Kristof

So the component you have described could be a CD transport. You can get a S/PDIF output from a very cheap CD player, or you can spend thousands. I mean, these things are just spinning a disc and extracting digital data aren't they? So would you say that the quality of sound you eventually get would be the same between a cheap transport and a high-end one? If you think they are the same, you would be very wrong...you couldn't be more wrong in fact. But this is what you have described (a component that takes digital data and outputs it in a recognised format).

Like I said before, in a general sense with almost everything - you get what you pay for. Simple as that.

I don't know what a "CD transport" does exactly. But whatever it does, obviously, there can be many ways to mess it up, and a "real cheap" device might do so, because it's trying to save cost. Also, it might not work reliably, or it might be inconvenient to handle, etc.

So no, I am not saying that all cheap devices are equal to all expensive devices.

What I am saying is this:

IF

• A) A given class of (digital) component has a clear, unambiguous specification (for input, transformation and output), and
• B) Two devices both implement the same specification correctly,

THEN

• talking about the "quality of sound" makes no sense. (Because the sound (at this stage of processing) is simply a stream of bits, which is either correct, or wrong. No third option.) Therefore, there in no point in "improving it" in any way.

I am not sure "CD transport" fulfills the A) condition. If it does not, then there is some room for competition. If it does, then I don't see how the sound quality of two correct implementations could differ - since the output must be match, bit-by-bit. (Of course there can be flaky devices, which don't fulfill condition B), I am not talking about those.)



I would be really thankful if you could explain me why.



Unfortunately, when lies and deception enter the picture, that is no longer true.
(And (again, unfortunately), whenever there is money to be made, lies and deception usually enter the picture.)

Best wishes:

Kristof

Some people really DO like their gold infused silver and their magneto-coupled phase inducers.
Show some respect!

Beyerdynamic DT880's are a real nice set of cans if you are looking for a neutral sound.

I use the Sennheiser HD 280 Pro ($100 USD) for isolation in a setting that has both plugged and unplugged players as well as more than a little conversation as a constant background. If you are squemish about being almost totally isolated then these are not for you. I would certainly check out active noise cancellation i.e. Bose Quiet Comfort 15($300 USD)if you plan on long term wearing comfort.

A CD transport spins a CD and contains the components (laser etc) that reads the disc. It then outputs a S/PDIF or optical digital output that is then fed into a DAC.

Only you will know if this meets the criteria you have set but to me it does. But any Hi-Fi enthusiast will know that the best transports cost loads and loads of money (thousands) - but it is just spitting out 1s and 0s. If you had experience of Hi-Fi you'd know how much better a good CD player or transport/DAC sounds.

I don't know what your area of expertise is but your dismissal (and the way others talk about this subject) really p*sses me off because by your own admission your opinion is based on absolute ignorance.

You might have some choice words of advice to someone who just landed on this forum and said "I've just seen a really nice looking Crapio digital piano in Costco. It's only $299. It's got 88 keys, it looks like a piano and it sounds like one to me. I don't see the point in spending more". You, like everyone else on this forum, would advise the person to play other pianos before making that choice and you might point out all sorts of ways that a more expensive piano would be better. That person's initial opinion is based purely on ignorance, because they know nothing about digital pianos.

You should take the trouble to listen to some decent Hi-Fi and make some comparisons before writing about this subject. You talk about lies and deception. You probably wouldn't say that the makers of the best cars, pianos, cameras, computers etc are just in the business of telling lies or deceiving people. Higher price=better product in the vast majority of cases. Why would you think for a moment that Hi-Fi is any different? That would be an illogical conclusion.

The great thing about fantastic claims for audio equipment, it's not that difficult to devise a double blind test.

The claims for Monster Cable never seem to be backed up by any real testing.

I'm sure there are measurable differences between CD players but I'm guessing, ... no, I'm betting, there will be no winners in a double blind test.

So says another completely ignorant individual who's just much too willing to share his opinion - which is based on absolutely no experience whatsoever of course. The conceit and arrogance of you people just beggars belief.

Why don't you just troll around on loads of forums spouting uninformed rubbish about topics of which you actually have no knowledge. I mean, don't stop here. Spread some joy onto a bike forum, or sign up to a mountaineering forum, Hi-Fi, cars, cameras and just feel free to slag products or ideas off - there's no need to actually know anything about the topic, go for it. Gosh, there's almost no limit to the number of people you can p*ss off with your TOTAL BLEEDING IGNORANCE.

So says another completely ignorant individual who's just much too willing to share his opinion - which is based on absolutely no experience whatsoever of course. The conceit and arrogance of you people just beggars belief.

Why don't you just troll around on loads of forums spouting uninformed rubbish about topics of which you actually have no knowledge. I mean, don't stop here. Spread some joy onto a bike forum, or sign up to a mountaineering forum, Hi-Fi, cars, cameras and just feel free to slag products or ideas off - there's no need to actually know anything about the topic, go for it. Gosh, there's almost no limit to the number of people you can p*ss off with your TOTAL BLEEDING IGNORANCE.

EssBrace, can you point us to double blind tests of CD players?

This would seem to be a relatively simple test to create (as long as all units are kept within a third of one dB of each other. Listeners very often choose the louder source as being the better so it's important to keep the playing field extremely level.)

What do you think would generate more heat? A high power, tubed amp, or this thread?

It seems to me that it does. There are two points that I am not sure about:

• Do we have to care about read errors (and errors correction) on the CD? I hope not. (If the data on the CD is correct, than there is only one possible way to read it.)
• Does the DAC use it's own timer when playing the audio, or does the timing come from the CD transport side? If the DAC has it's own timer, than the small inequalities in the timing of the data stream coming from the CD transport does not count. If it's using the timing of the data stream, then it does.)

These two points might mean but the "CD transport" device class does not meat the requirements I specified above, but I hope it does.

An interesting claim. How do you suppose that is possible? I can imagine three possible things you might be thinking:

• A) the output of the device is not just 0s and 1s, there is something else, too.
• B) the output is just a stream of 0s and 1s, but different devices produce different streams.
• C) The output is just a stream of 0s and 1s, and they are indeed identical, but somehow, when played by the DAC, the same bitstream coming from one device sounds better than the same bitstream coming from the other device.

Which one is it? (I might have left out some options, please feel free to correct me.)

Now,

• if you think A), then the device is not a purely digital device (as per my definition), so my claim does not say anything about it. (Thus, I was not wrong.)
• if you think B), then again, the device is not clearly specified enough (because if it would be, then could not be more than one possible correct output), so, again, my claim does not say anything about is. (Thus, I was not wrong.)
• if you think C), then I would be really interested in your thoughts about how would that be possible. The whole point of digital technology is that it's deterministic: the same algorithm, working from the same data, will always yield the same result. The exact same result. No difference. Voltage levels are irrelevant. Timing is irrelevant. Only 0 and 1, end of story. If this was not true, that would have deeply troubling implications, and really, none of our technology would work. For example, a current Intel core i7 CPU contains around 2,270,000,000 transistors. Something that complex can only work, because the small inequalities in voltage and timing are irrelevant, only 0s and 1s are transmitted. If that was not true, the errors would add up, the the system would collapse.
  
  So, my point is: it's theoretically impossible that the same stream of 0s and 1s will sound better, coming from one device than other.

(MSc, Computer Science)I am not dismissing anything; I have not denied that there are flaky devices; I have not denied that it's hard to get analog devices right, and there are different levels of quality there. I have only stated that that same is not true for (a subset of) digital devices, and that there are fake products on the market, marketed with dubious statements.
The feeling is mutual. (But of course it's not directed against you.)
No, that's not what I have said. What I said that my opinion is based on pure (but strong) theoretical knowledge. Quite simply, if the theory (which happens to be the base of all current IT technology) is true, than some of the claims made by Hi-Fi marketing (along the lines of point C) above) simply can not be true in any possible way. As long as the validity of the theory is not questioned by anyone, there is no way out of this; it's as sure as 2+2=4.

Now of course it's possible that

• The devices you are talking about are not purely digital (that's why I have tried to give a full definition to exclude these cases), or
• The theory might be wrong. That would be very interesting (and troubling at the same time), but it's always a possibility. However, in this case, those vendors who stumble on results that indicate the theory might be wrong (ie. the same digital data can behave differently, in different circumstances), ought to do research to uncover this fact, and then come forward with the claim that they have disrupted the whole current scientific paradigm. However, they do not seem to be doing this. (They just sell expansive stuff instead.) To me, this suggest even they don't really believe that they have achieved a scientific break-through; they just like to rip people off.

The difference is that in digital pianos, there are many, many things that can be done in different ways.
However, with clearly specified (again, input, transformation, output) digital devices, this variety does not exist. The output is either correct, or it's wrong.


I leave that task up to others. Please note that I have not written anything specific about any specific devices; I have just made some general, theoretical claims - the validity of which is based on theoretical consideration. (Which, in turn, are based on, and validated by the experience of many, many other people.)

And you doubt that these thins are a prominent in today's society?
No, I would not; not just telling lies. But, in addition to making great stuff, they do tend to tell smaller or bigger lies to their customers, again and again. (That's what they call marketing, among other things.)



I tell you why: because in Hi-Fi, it's very, very hard to objectively measure the quality. Thus, there is a big room for making false claims that are impossible to sufficiently refute, because they just can't be measured properly. And, if there is a room to make false claims (and make money), there are always some people who are willing to do so. (And that's what pissing me off.)

Anyway, this is not limited to Hi-Fi in any way, there are many other areas where it's hard to verify whether you are indeed getting what you are paying for. Think about anti-allergy products, water purifying technology, a whole lot of "alternative medicine", etc.

In these cases, it's not "you get what you pay for", it's mostly "you pay for what you think you get (which might or might not be similar to what you get)".

It's sad, but that's the current state of affairs. (Not that I am content with accepting this...)

* * *

If you are pissed of at me, please take that anger, turn it into energy, and try to explain me why am I wrong. I am listening.

Best wishes:

Kristof

One small clarification here: this only holds if the input (where the data is going to) is really purely digital.

If the DAC is designed really poorly, than the small voltage inequalities (in the digital stream) could theoretically influence it's analog output. However, with a properly designed DAC, that should be impossible (just apply some proper decoupling to the input, problem solved), and even if this problem exists, the size of the distortion should always b at least one magnitude lower than the the value of the least significant bit - thus, with 24-bit audio data, it would be less than 1 / 2^25 = 0,0000000298 on the [0-1] scale. (I don't think anybody could hear that, but if someone can show me a test to prove otherwise, I am willing to reconsider.)

Kristof

Jitter can be a problem in the digital domain.

Greg.

Great article. My conclusions, relations to the current discussions:

Jitter can cause two kinds of problems:

• When the receiving end reconstructs the digital data (arriving on the cable), it might get some bits wrong, because the timing of the data-stream is not perfect. I consider this an error, and devices that to this (either devices sending bogus data, or devices receiving the data wrong) simply flaky. This is not really an issue of quality; either the transmission is correct, or it's faulty. To avoid these problems, it would be best to use a protocol designed with built-in error detection and correction (like TCP/IP), but for now, we have to run with what we have.
• When a DAC is playing the bit-stream directly (ie. using the timing of the digital stream), the timing inequalities in the data-stream might sneak into the analog signal. The easy way to solve this is to introduce a really small buffer before the DAC, and feed the data (from the buffer to the DAC) using the DAC's built-in, precise clock. That way, the problem can be avoided completely. (I don't know if current DAC's do this.) So, again, if the DAC is designed properly, the small timing inequalities in the input should not count.

BTW, my favorite part in the linked article:

That's what I am saying. Since these things are hard to measure and verify, some people like to make up things. (For fun and profits.)

Said, but true.

Kristof

UPDATE: After a bit of research, it seems that the problems coming from the unstable timing on the sender side (ie. jitter) can be problem with S/PDIF. (Which, btw, means that according to my original definition, the S/PDIF signal is not purely digital, since the clock signal (which is an integral part of the transmitted data), is an analog quantity.) However, good DAC desing can solve this. (See references at the and of the above link.)

Well, even digital stuff has to live on the analog world at some point. You transfer digital information over the internet yet you need mechanisms to cope with the numerous errors induced by the physical, analog transmission medium you use.

I don't know what exactly is a CD "transport", but in the end, the laser is an analog device reading a digital information encoded on an analog medium in an imperfect manner. So the existence of more robust, durable, failproof electronics is not something that "cannot" exist just because in the end the information is digital. Digital information is a construct of the mind and the technology, not a reality.

That said, I'm not defending the HiFi world because some do tend to be overzealous. Not unlike a piano, digital or acoustic, regardless of what someone has to say, if you find that 200€ headphones do the trick for you, I'd say that you don't need to spend more

Yes, and those mechanisms are there (called TCP), so the data I transmit over the Internet is exactly the same as the original.



It's a weird feeling to hear this, as I am someone someone who works with digital information. Believe me, it's very real.

Sure, even digital information is stored in the analog world; and that's why the safety margins (and, in some cases, redundancy) are there. In the case of the CD, the analog medium and the analog laser should not be able to introduce any error; if they do, it's detected and corrected. If this does not happen, then device in question is flaky, the implementation is not correct.

The thing is, if you read in the data from a CD (an undamaged CD, that is), with a digital CD drive (for example one like used in computers), the data will be identical, bit my bit, every single time, every single CD drive. There is no variation there. That's the magic of the digital technology. This is not an unobtainable ideal; this is how digital stuff actually works.

Of course there can be flaky devices, which make errors, but that's not a question of "sound quality", it's simply a bug that should be fixed - like your car not starting, your computer crashing, etc.

* * *

Now of course it's getting way more complicated when one needs to actually move the digital information back to the analog world; that's why I am not saying anything about what happens after the DAC.

Kristof

I have Sennheiser HD560 Ovation II and I like them. Open, transparent sound with the correct amount of bass and no annoying problem spots. Fits nice on the head too.

I heard some more mixed comments on the newer HD's of Sennheiser. I can't comment on those as I never tried them.

Reconstruction of an analog signal by any real-world DAC is subject to real-world errors. These may be in the time, phase and/or frequency domains. They may be correlated or uncorrelated.

Key points: 1) These errors are not identical each time the real-world DAC is presented the same bit stream. 2) These errors are not identical between any two real-world DAC examples. 3) The existence of random and systematic errors in the reconstructed analog signal in no way implies that such errors are audible.

Here's a hint. If you want to debate someone claiming to hear what you do not believe he can actually hear, do not frame the debate in terms of the imaginary error-free perfection of some Platonic ideal of the digital to analog conversion process. Stick to discussing what the person can or can not hear. Specious claims as to the perfection of real-world analog signals reconstructed by a real-world DAC are just as silly as claims about having Golden Ears that can hear unimaginably subtle differences among various real-world bits of gear.