This is not an appropriate interpretation of a null result in an ABX test. Strictly speaking, the objective point of view says that there is no apparent difference under the conditions of the test.
The ABX test asks the listener "Which device is it?". The sighted test asks the listener "Which device do you prefer?" These are two completely different questions. In this respect an ABX test can never replicate the conditions of a sighted test.
I suppose it is possible to design a DBT so that the question asked of the listener is "Which device do you prefer?" The issue is that such a test would be rather inconvenient to perform for all involved. The careful tester would need to devise a test such that the listener would not suspect his ability to discern was under test, only that he was simply being asked to state his preference between two different pieces of gear. Sometimes real science is hard to do.
My interest in this thread is as an owner of the Metrum Octave. I agree with others that it is an extremely good sounding DAC when fed with a low jitter signal.
This had been planned as a simple request . I had wanted to know what the original dacs were out of curiosity and what sort of application they had been made for. IC,s used in Audio as op-amps were sometimes not designed for audio use but had the right ingredients .
I understand that the sound of this Dac is excellent and there are a growing number of people who have heard the Dac who think it is a very special sounding unit indeed. The company appears to be a very small one man outfit who have been rather overwhelmed by the recent publicity and stocks of the units are very hard to come by.
I did not want to cause an outbreak of name calling and justifying NOS versus any other type of digital filtering.
I am not very well versed in the technical aspects of audio either digital or otherwise but in the case of digital outputs there does seems to be a trend that leads me to believe (rightly or wrongly I have not idea) that the actual chip conversion on most Dacs makes very little difference and while in the digital domain most well specified chips either expensive or not makes little difference. What really matters is when that digital signal is converted to an analogue output were it then matters a great deal about power supply , noise , and all the other things that always matter with audio.
If I read the comments on the Metrum correctly what it does is keep the signal as digital as simple as possible then convert the output . This may be technically flawed or technically brilliant but frankly who cares ? Does it sound better if does and it does so in your system with your ears then make your decsion either buy it or not. The price appears to be very sensible given the work and the construction of the unit there will be a profit element but it would appear to be a sensible trade rather than some of the huge mark ups with other brands such as (Linn , Naim , Wadia etc I am sure we can all name a few) . Good luck to Metrum I would love to hear their product but at present it is not that easy in the UK but I am sure it will happen .
One thing that has been through much of this thread which does make me wonder . If a product sounds fantastic the best you have ever heard but does not measure as good as another product which do you believe ? Your ears which tell you this is great music and have you running to your collection to dig out the next treasure that you want to hear again ? Or the measurement which tells you this product can not sound that good because it does not measure as well as b , c or d. I know which way I would go but I would not like to say for sure for some of the posters on this thread.
Remember Tom Waits is still called a singer , much of what he does is out of tune and frankly if analysed sounds very odd against a musical scale but it still does not stop me feeling that he is an artist that I would listen too for many high moments .
Thank you to all who contributed on this thread I am sadly still none the wiser about what chips were used but if I ever meet the make I will ask him just to see if he tells me.
I understand that the sound of this Dac is excellent and there are a growing number of people who have heard the Dac who think it is a very special sounding unit indeed. The company appears to be a very small one man outfit who have been rather overwhelmed by the recent publicity and stocks of the units are very hard to come by.
I did not want to cause an outbreak of name calling and justifying NOS versus any other type of digital filtering.
I am not very well versed in the technical aspects of audio either digital or otherwise but in the case of digital outputs there does seems to be a trend that leads me to believe (rightly or wrongly I have not idea) that the actual chip conversion on most Dacs makes very little difference and while in the digital domain most well specified chips either expensive or not makes little difference. What really matters is when that digital signal is converted to an analogue output were it then matters a great deal about power supply , noise , and all the other things that always matter with audio.
If I read the comments on the Metrum correctly what it does is keep the signal as digital as simple as possible then convert the output . This may be technically flawed or technically brilliant but frankly who cares ? Does it sound better if does and it does so in your system with your ears then make your decsion either buy it or not. The price appears to be very sensible given the work and the construction of the unit there will be a profit element but it would appear to be a sensible trade rather than some of the huge mark ups with other brands such as (Linn , Naim , Wadia etc I am sure we can all name a few) . Good luck to Metrum I would love to hear their product but at present it is not that easy in the UK but I am sure it will happen .
One thing that has been through much of this thread which does make me wonder . If a product sounds fantastic the best you have ever heard but does not measure as good as another product which do you believe ? Your ears which tell you this is great music and have you running to your collection to dig out the next treasure that you want to hear again ? Or the measurement which tells you this product can not sound that good because it does not measure as well as b , c or d. I know which way I would go but I would not like to say for sure for some of the posters on this thread.
Remember Tom Waits is still called a singer , much of what he does is out of tune and frankly if analysed sounds very odd against a musical scale but it still does not stop me feeling that he is an artist that I would listen too for many high moments .
Thank you to all who contributed on this thread I am sadly still none the wiser about what chips were used but if I ever meet the make I will ask him just to see if he tells me.
When measurements do not seem to correlate with perceived sound quality, one might question which measurements are actually relevant with respect to sound quality, and in the end it might turn out that the relevant measurements for the better sounding unit are indeed better.
I mentioned before that a NOS dac has typically some - 2 dB at 20 kHz, whereas an OS dac is flat. However I (and I guess over 90% of all listeners) do not notice that very slight HF roll off of the NOS dac.
At the same time the absence of oversampling and digital filtering of the NOS dac should give better time domain characteristics, and personally I believe that our ears are more sensitive to this phenomenon (it's much like the way our ear-brain system works); more so than absolute flatness of frequency bandwidth.
Maybe the NOS dac measures better where it is important.
Ken Newton said:
no, the problem is that you will not accept that these 'significant differences' are distortions, instead preferring to call it more transparent and bringing you 'closer to the music' without becoming 'fatiguing' or 'rich in harmonic detail' a simple minded yet appropriate metaphor 'you cant have your cake and eat it too' quit making up theoretical and often somewhat radical effects to explain this preference that means you dont have to accept technical inferiority in order to have your preferred sound signature verified. distortions can be quite amazing and yes perhaps even missing in some recordings that should have it, but that doesnt make the playback more true to source, it just means the effect is convincing and enjoyable
this is the main problem we face in these disagreements. there is of course more than one type of distortion, but its these lesser known and somewhat esoteric mostly phase related distortions that are often claimed to be present in essentially perfect reproductions of limited audible bandwidth to claim an upper hand. In many of these cases, the component at hand is very much just idling while producing the audible bandwidth
you say its all about the sound, but the thing is, often more than other groups, it revolves around being special and or elite in some way, which doesnt gel all that well with measurements bettered by modern low level consumer gear and havent improved all that much (because they use much the same parts) since the late 80's-early 90's
there is nothing wrong with preferring a somewhat coloured sound signature, diversity is to be encouraged; people admit this with transducers and to a lesser extent amplifiers all the time and its not a big deal, but when it comes to source, people would rather sling mud and make stuff up than admitting they prefer a bit of colour, as if its somehow makes them less of an audiophile.
a slight roll off at 20khz is not where it ends according to measurements of this particular dac. thats not for me, but that doesnt mean its a flawed product overall, it just caters to a different preference.
Last edited:
Evidence please for this assertion?
If the differences being referred to here are ultrasonic (imaging distortion above 22k05Hz), then do you have a hypothesis for how are we able to hear them?
I concur. The last three DACs I designed and built had one purpose, to enable me to explore what, if anything, could be done about the annoying artifacts many of us hear when listening to generic CD playback. As it turns out, much can be done.
While D/A converter architecture, jitter suppression, I/V stage, and supply regulation all represent significant opportunities for sound improvement, my experiments have led me to conclude that, perhaps, the single MOST significant factor is the implementation of the nearly ubiquitous, half-band, digital sinc filter. Of all the improvements which I tried, the most significant subjective reduction in annoying artifacts came via both NOS (no digital filter), and so-called apodizing (digital filter with an in-band cut-off frequency).
BTW, I apologize to the thread participants for my part in the rather unseemly bit of name calling between Sonic and myself. He and I have a history of butting heads.
Last edited:
evidence that we are talking about distortions present only above that range please? perhaps i'm being overly general, but that is certainly not the case with a large number of passive or filterless output NOS dacs driving real amplifiers
Last edited:
That's a distraction because you were talking about distortions, not 'we'.
Your claim was that the significant differences being heard were distortions. You either have evidence supporting that assertion or you don't. Which is it?
you claim to be smart right? this is not much of a stretch for even a less than logical brain. very high level measured performance on one side, lower measured performance on the other. A 'significant difference' in sound is heard between them and perceived as an improvement. the difference between them can ONLY be a distortion of some kind (not necessarily THD) unless you are claiming an as yet unproven unmeasurable factor for which you have no proof even of existence. I think its a bit humorous actually how people dont even seem to have enough strength of character to just say yep it has distortions and I love the way it sounds, instead wishing to sweep it aside, yearning for that better number but torn by the enjoyment you feel.
for example in this particular thread we are talking about a dac that has 0.1% THD, as linked in a review off their own website. thats pretty **** in any language but yet apparently still sounds good
for example in this particular thread we are talking about a dac that has 0.1% THD, as linked in a review off their own website. thats pretty **** in any language but yet apparently still sounds good
Nope - what use is self-aggrandizement?
That's a long-winded no then - no evidence. Fair enough.
haha yeah right, not all that convincing that quip
i'm sorry but there is no evidence required for a simple difference equation, i'm not the one claiming something exists that creates a 'significant difference' that cannot be measured. we have evidence that shows it measures worse (they use it to promote) and this one pretty significantly worse in some aspects (although the lack of ringing is quite impressive) so any improvement that is claimed is either something you cannot tell me, or a distortion and there is a correlated difference in distortion..so......
they are the only options, one has pretty good backup, the other has none. I would say it is you who has to either accept that you dont hold any cards and simply enjoy something that doesnt quite measure up, using a metric that you claim not to care for this shouldnt be much of a stretch right? but no, you only want to take the bits of engineering that make you and your preferences look good, but the ones that dont are not important.
I dont have the time to look more into it, but in my 2 minute search on this topic I also found conversation about a 4khz AM modulation in the audio band
i'm sorry but there is no evidence required for a simple difference equation, i'm not the one claiming something exists that creates a 'significant difference' that cannot be measured. we have evidence that shows it measures worse (they use it to promote) and this one pretty significantly worse in some aspects (although the lack of ringing is quite impressive) so any improvement that is claimed is either something you cannot tell me, or a distortion and there is a correlated difference in distortion..so......
they are the only options, one has pretty good backup, the other has none. I would say it is you who has to either accept that you dont hold any cards and simply enjoy something that doesnt quite measure up, using a metric that you claim not to care for this shouldnt be much of a stretch right? but no, you only want to take the bits of engineering that make you and your preferences look good, but the ones that dont are not important.
I dont have the time to look more into it, but in my 2 minute search on this topic I also found conversation about a 4khz AM modulation in the audio band
Last edited:
Ah but ironically, that's precisely what you are claiming. You are claiming distortions in the NOS case which have not been measured.
When you have the measurements which back up your alleged 'distortions', progress can be made.
what? you seem to be having trouble seeing that the word distortions has been used to relate to a claim for significant audible differences from a measured response that is many times the audio bandwidth, yet there is nothing to show for it but WORSE measurements
I forgot how much you love circular argument
bye
I forgot how much you love circular argument
bye
If it sounds good it is good.
No discussion can make up for that. Maybe we don't fully understand how ears and brains work otherwise the NOS people would not be happy with their NOS DACs. Let's agree that we should be happy that we don't have to listen to measurement equipment ....
No discussion can make up for that. Maybe we don't fully understand how ears and brains work otherwise the NOS people would not be happy with their NOS DACs. Let's agree that we should be happy that we don't have to listen to measurement equipment ....
Last edited:
Actually, I wonder whether you recognize that you are implicitly declaring all subjective differences to be some form of added distortion or inaccuracy? I've never claimed to know whether subjective differences are related to some particular distortion or not. I simply claim to hear such differences.
Let me ask you a question, if the information encoded on a CD is time-domain sensitive (i.e., music), does the digital sinc filter's time-domain response ("ringing") distort that information? Be careful in your answer, recall that you've already said that the system having the higher distortion is definitionally the more colored.
The allure of being among the audio elite may hold attraction for some, but not for me. I simply want to maximize the enjoyment of my home music listening experiences. Just because there may be those among us who aspire to golden ear or guru social status, doesn't mean that there aren't those who genuinely and honestly hear significant differences. Again, this sort of gets back to the implication that any subjective evaluation which appears to run counter to currently accepted measured parameters, is automatically wrong.
This paragraph is too laced with smug condecesion to address.
Last edited:
Some people are happy to eat toilet paper and call it "good food". It does not mean that the rest of the world have to accept their "rational".
Some people used to say that they see the world flat and therefore... it is flat. We needed measurements to prove the truth. Audio comunity is the only place where you see questioned the capability of measure a certain fenomen.
Some people used to say that they see the world flat and therefore... it is flat. We needed measurements to prove the truth. Audio comunity is the only place where you see questioned the capability of measure a certain fenomen.
Last edited:
Wow, this discussion is going at a brisk pace, so I needed some time to catch up.
Quite a lot of what Ken and 5th element have written seems to be in line with my own experiences with audio.
Well, let me tell you about my experience with the Octave during auditioning it. To me, there were no siginificant differences between the 128x oversampling delta-sigma DAC (+OPA132 opamps) in the Primare D20 and the non oversampling Octave. That's why I didn't give a detailed description in a previous post. The differences were small and no doubt caused partly because of a slight difference in output level of the analogue outputs. Despite being both specified at 2 V, the Octave's is slightly higher, I believe the HiFi Critic review confirms this.
I don't have a very long attention span, so I drifted off quite a few times resulting in not paying that much attention to the music anymore. It was during one of those moments that a note from a brass instrument drew my attention. The rasp just sounded slightly more detailed and not as muffled through the Octave.
On one occasion when I was paying attention, I felt the Primare sounded a little "too easy" when I switched back from the Octave.
An hour or so later, I couldn't get myself to detect those differences anymore, so they must have been very small. A big difference would have remained easy to detect, IMO.
The only thing that always stood out was the better perception of dynamics with the Octave, and it's mainly why I bought it.
The Octave sounded no more or less distorted that the Primare.
Phase related distortions are caused in the time domain, right? It's my opinion that we're very sensitive to variations in the time domain as a result of evolution.
To be able to locate the source of sounds, we must inherently be very sensitive to the minute amount of time it takes sound to travel the distance between left and right ear. When sound is travelling at 340 m/s, that distance takes less than 0.6 ms, and yet we have no trouble telling the direction the sound came from. IMHO the hearring gets confused and fatigued by the smearring effect digital filters have on sound during longer listening sessions. Kusunoki visualizes this as sound coming from a row of speakers.
Audio measurements mainly focus on the frequency domain. I'd like to see the emergence of a measurement that visualizes how coherently a device performs in the time domain.
Quite a lot of what Ken and 5th element have written seems to be in line with my own experiences with audio.
Well, let me tell you about my experience with the Octave during auditioning it. To me, there were no siginificant differences between the 128x oversampling delta-sigma DAC (+OPA132 opamps) in the Primare D20 and the non oversampling Octave. That's why I didn't give a detailed description in a previous post. The differences were small and no doubt caused partly because of a slight difference in output level of the analogue outputs. Despite being both specified at 2 V, the Octave's is slightly higher, I believe the HiFi Critic review confirms this.
I don't have a very long attention span, so I drifted off quite a few times resulting in not paying that much attention to the music anymore. It was during one of those moments that a note from a brass instrument drew my attention. The rasp just sounded slightly more detailed and not as muffled through the Octave.
On one occasion when I was paying attention, I felt the Primare sounded a little "too easy" when I switched back from the Octave.
An hour or so later, I couldn't get myself to detect those differences anymore, so they must have been very small. A big difference would have remained easy to detect, IMO.
The only thing that always stood out was the better perception of dynamics with the Octave, and it's mainly why I bought it.
The Octave sounded no more or less distorted that the Primare.
Phase related distortions are caused in the time domain, right? It's my opinion that we're very sensitive to variations in the time domain as a result of evolution.
To be able to locate the source of sounds, we must inherently be very sensitive to the minute amount of time it takes sound to travel the distance between left and right ear. When sound is travelling at 340 m/s, that distance takes less than 0.6 ms, and yet we have no trouble telling the direction the sound came from. IMHO the hearring gets confused and fatigued by the smearring effect digital filters have on sound during longer listening sessions. Kusunoki visualizes this as sound coming from a row of speakers.
Audio measurements mainly focus on the frequency domain. I'd like to see the emergence of a measurement that visualizes how coherently a device performs in the time domain.
Last edited:
You use inches and call it good, we use centimeters and call it good...... Some call hamburgers good food, some think an entrecôte is good food. Some think that the rest of the world should think like what they think it is right.
Any way, if someone likes the sound and gets touched by what he/she hears that should be good enough.
You can debate as much as you want but I tend to think that we don't have the knowledge to know what to measure exactly what we define as good sounding. I have heard many fine measuring devices that sounded like ****. Tell me what the use is of a very good measuring device when you want to shut it down when you have heard it for just 5 minutes ? After a lot of years the only thing I am sure of is that the time domain is very important in digital audio (like member jitter" points out). The exact topology is less interesting as it is only a means to a goal. But we all know some think the means are more important than the goal...
Truth is very subjective. Your truth does not have to be my truth. I see a lot of truths that are not my truths. And now please adapt to the metric system because it is better !
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.