I am baffled by this thread. Is it a complaint against capitalism? If so, it belongs elsewhere. Surely anyone is free to make and sell anything which is legal in their marketplace. Even illegality (e.g. fraudulent CE marking) does not stop some people selling stuff to the unsuspecting.
You certainly don't need a load of test equipment to sell audio. Some suppliers even claim lack of equipment as a plus. Different buyers want different supporting evidence: some want measurements, some want a good story, some want a pretty box with wood/chrome/LEDs/meters/valves, some want a suitably low/high price tag, some just want to be part of the 'in' crowd, others want to avoid the mainstream and require eccentricity.
So what is the problem? Has someone stolen someone else's lunch?
You certainly don't need a load of test equipment to sell audio. Some suppliers even claim lack of equipment as a plus. Different buyers want different supporting evidence: some want measurements, some want a good story, some want a pretty box with wood/chrome/LEDs/meters/valves, some want a suitably low/high price tag, some just want to be part of the 'in' crowd, others want to avoid the mainstream and require eccentricity.
So what is the problem? Has someone stolen someone else's lunch?
If it were a complaint against capitalism, I'd be entitled to make it here in the lounge.
I'm sorry that the argument escapes you, I'll try putting it a different way.
I don't like to see ivory poaching on the TV, but I understand that poachers are, in some cases, simply doing their best for their families. If you want to stop poaching you need to take a two-pronged approach. Enforce the illegality, but provide an alternate source of income for the poachers.
RocketScientist is doing a great job as an enforcer, but the way he has employed his prosperity, while undoubtedly making inroads into cleaning up a marketplace rife with malpractice, is at the same time reducing the diversity of choices available to customers by raising the cost of entry into the market for legitimate entrepreneur developers, in a way which could be avoided were he to take a different approach.
I'm sorry that the argument escapes you, I'll try putting it a different way.
I don't like to see ivory poaching on the TV, but I understand that poachers are, in some cases, simply doing their best for their families. If you want to stop poaching you need to take a two-pronged approach. Enforce the illegality, but provide an alternate source of income for the poachers.
RocketScientist is doing a great job as an enforcer, but the way he has employed his prosperity, while undoubtedly making inroads into cleaning up a marketplace rife with malpractice, is at the same time reducing the diversity of choices available to customers by raising the cost of entry into the market for legitimate entrepreneur developers, in a way which could be avoided were he to take a different approach.
I don't see at all that legitimate entrepreneurs have a raised cost of entry. If they want to be objectivist in approach that's their choice - there is also the choice of going the subjectivist route, which is ultimately the more 'legitimate' one as audio products are designed to be employed for listening not measuring.
Au contraire, if you want to stop poaching, you have to make the commodity being poached so inexpensive that it isn't worth the economic effort to poach. My apologies, but my micro-economics instructor was a grad student of George Stigler, back on the Midway, and micro-economics has never failed me in the interpretation of human events. This even explains how Moses turned his staff into a snake.
Is this a rough translation of what is going on? "I want to sell something and make some money, but someone else has produced a good product for that same market and is (almost) giving it away for free. He is also (unsurprisingly) criticising other products in that market. That's not fair!" Whoever said life was fair? This sounds like Microsoft whingeing about Linux.
"Someone's been eating my porridge", said Mummy Bear.
"Someone's been eating my porridge", said Mummy Bear.
My $350 sound card and $380 worth of software is very good. Better than hundreds of thousands of HP stuff when I was back in the lab in the 70's. Amazing as it is, it is still limited by "audio" frequency acquisition. I have used modern spectrum analysis hardware and there is no comparison. The real stuff is still orders of magnitude over PC based. Same problem with just a "basic" scope. To work an amp, you need at least a 100Meg bw. That requires external acquisition.
This is hard for me. I get that on the one hand some see what I'm doing as "ruining the fun of others". But I'm also just trying to share a lot of information that's already been out there but is mostly ignored or intentionally swept under the rug. I'm not making any of it up. I mostly base it all on credible articles by others and easily verified measurements and tests.
Many of the AES papers I reference are many years old. The great Audio Myths Workshop video has been around for years. Most of this stuff isn't new, but the ones who just politely point it out tend to be ignored or otherwise dismissed. So yeah, my style might be a bit more controversial. But I honestly don't know how else to get the message out there. Polite hasn't worked year after year.
I really believe the wine industry has some amazing parallels to high-end audio. The two industries have followed a similar path along a similar timeline. I discuss that more in my What We Hear article. For example, Robin Goldstein demonstrated over hundreds of tastings, when people know the price of wine they express an obvious overall preference for the more expensive wines. But when they taste blind the trend is exactly reversed, they prefer the cheaper wines. The same thing happens in blind audio testing except instead of "prefer" it's often "can't tell them apart".
Despite people like Robin Goldstein with their websites, blogs, books, podcasts, documentaries, etc. the high-end wine industry is alive and well. Robin has shouted from the rooftops you don't need to spend more than $10 for a really satisfying bottle--wines even the experts overwhelmingly prefer in blind tastings--yet there's still a booming market for much more expensive wines. Audio is, and will be, similar. I can only have a small influence on the big picture.
So I think it's a bit extreme to pretend my humble blog is going to totally change the rule book for all things audio and those who can't afford a five figure audio analyzer should just give it up. There will always be those who are in this hobby for non-objective reasons. They don't care about measurements or blind tests. And it doesn't matter what people like me publish, they'll just ignore it. And that's fine.
If the "Octave Thread" is referring to the Metrum DAC the thing is a train wreck in terms of objective measurements. It's really hard not to lump it in with other snake oil products like Brilliant Pebbles. More on that in my next article.
Abraxalito's attack on my transparency comment completely ignores all the research behind that comment. Does Abraxalito claim he can hear distortion and other anomalies that are -100 dB below the music? Can he prove he can? Plenty of people have demonstrated you can't. I've even provided links to test files made by Ethan Winer than demonstrate even the most dissonant obnoxious noise is readily masked by even soft acoustic music at -75 dB. Listen for yourself.
Ethan Winer defines and explains the required measurements for audio transparency in the video linked above. Then he provides many examples where you can listen for yourself. There are also links to supporting AES papers. What are Abraxalito's credentials? How can he credibly challenge all of the above? Where are his published peer reviewed AES papers?
Audio differencing (null testing) has shown that even using real music into real loads, amplifiers deliver performance that follows sine wave measurements. If null testing shows everything an amplifier is doing is below -75 dB can someone still credibly argue that amp somehow isn't audibly transparent? If so I'd love to see the credible evidence. I've presented numerous peer reviewed AES papers, a great video by a panel of experts complete with downloadable files you can listen to for yourself, other well documented blind tests, and much more.
I ask everyone: What credible objective data has Abraxalito, or any of my other critics, presented in support of their criticisms? I haven't seen anything. At some point, a reasonable person would agree you can define audible transparency under realistic listening conditions. That's all I've tried to do. And I've been conservative and gone way further than -75 dB in terms of the measurements in my definition.
It might also be worth pointing out those who believe a decent soundcard (or interface) and some free/cheap software is the equal of a real audio analyzer like and A.P. or dScope are mistaken. While it's true you can get sound hardware with a similar noise floor to some analyzers they're still limited in a great many ways in terms of what and how they can measure. For one, they're at the mercy of the operating system as they don't have any real-time measurement hardware on board. And they generally have no concept of absolute levels making many measurements either meaningless or only relative. They were not designed for making absolute measurements.
PC sound hardware lacks proper calibrated input conditioning circuitry and it's often difficult or impossible to accurately maximize the dynamic range of their A/D--at least without using other other external hardware. An analyzer has the ability to measure, and generate, signals in the digital domain while measuring/altering the digital signal parameters. A soundcard can't do any of that--like measure the jitter of an incoming S/PDIF signal. Analyzers can also make real time measurements including being able to listen to, and view, the residual distortion in real time. My RMAA article goes into a lot more detail.
And a lot of it is the software. Most PC soundcard-based software lacks the ability to do lots of critical things. It can't identify, flag, and and sum symmetrical jitter components. It can't apply all the required weightings and standards such as those from IEC, AES17, etc to make industry standard measurements. I've yet to see any such software that can even do a basic THD vs Output Level/Power sweep let alone some of the more advanced scripting the dScope and AP analyzers do.
Does someone have to measure all these things to have fun with audio? Absolutely not. But if someone wants to debate things like audible transparency, claim Product X is better than Product Y, etc. it can be a huge help. A huge benefit of making credible measurements, that conform with industry standard practices, is to allow others to verify your work. I've been at this for over a year, has anyone yet shown where I've been seriously wrong with any of my measurements or references? Instead my critics just like to stand back and throw rocks.
Finally, I like to use Bottlehead as a good example of how to have lots of fun without needing so much as a DMM. They've built a culture around their tube gear that's based on having fun rather than worrying about measurements. I'm fine with that. Unlike others, I don't see Bottlehead claiming their gear has higher fidelity, has 0.001% THD, is objectively better than Brand X, etc.
Many of the AES papers I reference are many years old. The great Audio Myths Workshop video has been around for years. Most of this stuff isn't new, but the ones who just politely point it out tend to be ignored or otherwise dismissed. So yeah, my style might be a bit more controversial. But I honestly don't know how else to get the message out there. Polite hasn't worked year after year.
I really believe the wine industry has some amazing parallels to high-end audio. The two industries have followed a similar path along a similar timeline. I discuss that more in my What We Hear article. For example, Robin Goldstein demonstrated over hundreds of tastings, when people know the price of wine they express an obvious overall preference for the more expensive wines. But when they taste blind the trend is exactly reversed, they prefer the cheaper wines. The same thing happens in blind audio testing except instead of "prefer" it's often "can't tell them apart".
Despite people like Robin Goldstein with their websites, blogs, books, podcasts, documentaries, etc. the high-end wine industry is alive and well. Robin has shouted from the rooftops you don't need to spend more than $10 for a really satisfying bottle--wines even the experts overwhelmingly prefer in blind tastings--yet there's still a booming market for much more expensive wines. Audio is, and will be, similar. I can only have a small influence on the big picture.
So I think it's a bit extreme to pretend my humble blog is going to totally change the rule book for all things audio and those who can't afford a five figure audio analyzer should just give it up. There will always be those who are in this hobby for non-objective reasons. They don't care about measurements or blind tests. And it doesn't matter what people like me publish, they'll just ignore it. And that's fine.
If the "Octave Thread" is referring to the Metrum DAC the thing is a train wreck in terms of objective measurements. It's really hard not to lump it in with other snake oil products like Brilliant Pebbles. More on that in my next article.
Abraxalito's attack on my transparency comment completely ignores all the research behind that comment. Does Abraxalito claim he can hear distortion and other anomalies that are -100 dB below the music? Can he prove he can? Plenty of people have demonstrated you can't. I've even provided links to test files made by Ethan Winer than demonstrate even the most dissonant obnoxious noise is readily masked by even soft acoustic music at -75 dB. Listen for yourself.
Ethan Winer defines and explains the required measurements for audio transparency in the video linked above. Then he provides many examples where you can listen for yourself. There are also links to supporting AES papers. What are Abraxalito's credentials? How can he credibly challenge all of the above? Where are his published peer reviewed AES papers?
Audio differencing (null testing) has shown that even using real music into real loads, amplifiers deliver performance that follows sine wave measurements. If null testing shows everything an amplifier is doing is below -75 dB can someone still credibly argue that amp somehow isn't audibly transparent? If so I'd love to see the credible evidence. I've presented numerous peer reviewed AES papers, a great video by a panel of experts complete with downloadable files you can listen to for yourself, other well documented blind tests, and much more.
I ask everyone: What credible objective data has Abraxalito, or any of my other critics, presented in support of their criticisms? I haven't seen anything. At some point, a reasonable person would agree you can define audible transparency under realistic listening conditions. That's all I've tried to do. And I've been conservative and gone way further than -75 dB in terms of the measurements in my definition.
It might also be worth pointing out those who believe a decent soundcard (or interface) and some free/cheap software is the equal of a real audio analyzer like and A.P. or dScope are mistaken. While it's true you can get sound hardware with a similar noise floor to some analyzers they're still limited in a great many ways in terms of what and how they can measure. For one, they're at the mercy of the operating system as they don't have any real-time measurement hardware on board. And they generally have no concept of absolute levels making many measurements either meaningless or only relative. They were not designed for making absolute measurements.
PC sound hardware lacks proper calibrated input conditioning circuitry and it's often difficult or impossible to accurately maximize the dynamic range of their A/D--at least without using other other external hardware. An analyzer has the ability to measure, and generate, signals in the digital domain while measuring/altering the digital signal parameters. A soundcard can't do any of that--like measure the jitter of an incoming S/PDIF signal. Analyzers can also make real time measurements including being able to listen to, and view, the residual distortion in real time. My RMAA article goes into a lot more detail.
And a lot of it is the software. Most PC soundcard-based software lacks the ability to do lots of critical things. It can't identify, flag, and and sum symmetrical jitter components. It can't apply all the required weightings and standards such as those from IEC, AES17, etc to make industry standard measurements. I've yet to see any such software that can even do a basic THD vs Output Level/Power sweep let alone some of the more advanced scripting the dScope and AP analyzers do.
Does someone have to measure all these things to have fun with audio? Absolutely not. But if someone wants to debate things like audible transparency, claim Product X is better than Product Y, etc. it can be a huge help. A huge benefit of making credible measurements, that conform with industry standard practices, is to allow others to verify your work. I've been at this for over a year, has anyone yet shown where I've been seriously wrong with any of my measurements or references? Instead my critics just like to stand back and throw rocks.
Finally, I like to use Bottlehead as a good example of how to have lots of fun without needing so much as a DMM. They've built a culture around their tube gear that's based on having fun rather than worrying about measurements. I'm fine with that. Unlike others, I don't see Bottlehead claiming their gear has higher fidelity, has 0.001% THD, is objectively better than Brand X, etc.
no, you see where it becomes a bit of an unfair thing, is because RS has all of the testing backbone of a fairly decent audio company (better in many cases), but has absolutely no concern about or a need for making it pay for itself, charges nothing for his time, or the designs, so the people who are selling it, have gotten their R&D done entirely free and owe no licensing on its production or distribution and neither does he owe anything for costs on said viral distribution and marketing.
still sound fair? notice i'm not even talking about objectively inferior products, even well engineered products, some would say ESPECIALLY well engineered products as they have higher costs to cover doing things properly. so this may well backfire, putting the well meaning start ups out of business whether they do their job well or not
sorry RS, i'm about to sleep, you and I have already talked about my position on this well before, and despite not being on entirely the same page, we at least knew what page each other is on; my position hasnt changed, I made attempts to communicate this to you in the past and i'm pretty sure we made ourselves understood. ive been meaning to PM you before this all got out of hand started from a comment that actually had slightly different connotations to what has been inferred. I also wanted to message you about the GB thread stuff before it all gets spun another way by your champions. i'll shoot you a PM tonight at some point, I dont have the time to reply properly to you or this thread just now either, its after 8am and i havent slept yet.
Last edited:
Obviously neither of us can predict the future, but it doesn't seem likely I'm going to put any "well meaning starts ups out of business." Any company who wants to market their products objectively, and has no way to properly measure them, can pay a few bucks to a consulting engineer and get them tested. To otherwise just make up specs isn't "well meaning", it's borderline fraud. And any company, like say Bottlehead or Schiit, that has a mostly or entirely subjective marketing pitch doesn't need a product that measures well.
The idea is to raise the bar by showing what's possible, what really matters objectively, and if companies make objective claims to hold them accountable for those claims. Why should audio be different than say Secrets of Home Theater exposing all the DVD players that flunked lots of objective tests? People buy a DVD (or Blu Ray) player expecting it to accurately reproduce what's on the disc not tint it green or turn diagonal lines into coarse jagged ones.
If the gear isn't accurate, or the company has no idea, a well meaning company shouldn't make claims with no basis in fact. This is just one example but the concept holds true for most objective claims. So I don't see how I'm causing harm to well meaning honest companies. Only the ones trying to deceive people with false objective claims need worry and even most of those, just like the expensive wine makers, will still find plenty of buyers.
Finally, the O2 and ODAC are not being sold especially cheaply. They could be mass produced for a small fraction of what they're currently selling for. And, for the O2, by anyone. So the hypothetical company that you argue "can't complete" can just start selling the O2 if that's really their biggest hurdle. I also hope to have a fully open source DAC someday. But, honestly, anyone who knows which end of a soldering iron to hang onto can probably reverse engineer the existing ODAC when it hits the street. The can go sign the right paperwork with Tenor and ESS and make their own. China is famous for doing exactly that. To me they're a much bigger threat to boutique audio start ups than I am.
Last edited:
sorry you missed the point again. no matter what they do or what claims they make. lets say they have a normal operating budget they do all the right things, dont make any claims they cant actually do (which takes more than throwing a few dollars at the problem, sorry thats ridiculous, youve outlined in your blog how much work went in to choosing each part to meet the criteria and you want to say a few dollars testing?). to imply that throwing a few dollars at a testing firm somehow makes a difference to the fact that your dac is being offered without anyone paying much for advertising, you do that for them as does the viral marketing team, they dont have to pay the designer, you did that for them, they dont have to cover multiple generations of prototype, you did that for them, they dont have to pay for documentation or manual, you did that for them, they dont have to pay any licensing, you gave it away for free. thats the majority of the costs of bringing a product to market, the cost of parts, though significant, fades in comparison to these factors, which have been taken out of the equation.
but not for the honest competitor, you have raised the bar for what their costs are, which can be argued should be the case and i agree for the most part, but you have then given them an impossible sales target to match and an impossible budget, given a real production and design workflow with actual costs and actual paid labor
but not for the honest competitor, you have raised the bar for what their costs are, which can be argued should be the case and i agree for the most part, but you have then given them an impossible sales target to match and an impossible budget, given a real production and design workflow with actual costs and actual paid labor
Last edited:
I can't argue with some of that, but fundamentally how is the O2 that different than Owen's The Wire project here? He used an expensive Audio Precision analyzer to refine his design. I don't see anyone questioning his motives. What's so horrible about sharing a design for free?
The "Microsoft whining about Linux" analogy has a lot of validity. The people who put lots of free time into the development of Linux were not doing it with the expectation of putting Microsoft out of business (however much some may have disliked MS). They were doing it because they wanted to provide an option and show what else was possible.
Competition is nearly always a good thing for consumers in the long run. But, somehow, you seem to be saying it's bad in this case if I'm understanding you correctly?
The "Microsoft whining about Linux" analogy has a lot of validity. The people who put lots of free time into the development of Linux were not doing it with the expectation of putting Microsoft out of business (however much some may have disliked MS). They were doing it because they wanted to provide an option and show what else was possible.
Competition is nearly always a good thing for consumers in the long run. But, somehow, you seem to be saying it's bad in this case if I'm understanding you correctly?
I've been over to your blog and read 'What we Hear' and it was a very long article, with a fair few red herrings. So its possible I missed where you posted up the results and experimental methods of rigorous audibility trials. If so just help me out here and I will retract my criticism. In science its the person who claims who must support and you claim 'audible transparency'.
No, I make no such claims. And have said that I don't in public (on DIYA) at least once. But this is irrelevant - I'm not making the claim that I can. If I were then you'd be reasonable to ask me to support that, just as I'm asking you to provide support for your claim of 'audible transparency'. Instead of providing support, you engage in red herrings. Strange that.
Deflection.
I've followed Ethan Winer for quite some time, nowhere have I seen him produce experimental results based on listening to the Benchmark DAC1 which demonstrate its 'audible transparency'. But if I've missed them just give me the link and I'll shut up.
More deflection
If you've done null testing on your ODAC that would indeed be a great start. Can you provide links showing your DAC produces better than a -75dB null? - I'd be impressed if you can.
Blind testing on DAC1 or ODAC would be interesting. Haven't seen published experimental details so far. Should I hold my breath?
None is required because I'm making no claims about my DAC so far. You're claiming and yet not supporting, which is unscientific.
I have the feeling, no matter how much detail I documented of the blind test, you would find ways to insist it was invalid. I believe you've even taken shots at Meyer & Moran with their 600 trial year long effort. Without you wearing the headphones yourself, it seems you're impossible to satisfy when you want to be on the other side of the fence--as clearly you do here.
All you would have to do is insist both listeners in the ODAC vs DAC1 test were predisposed to hearing no difference. So until we get some genuine critics who expect to hear a difference wearing the headphones, it's pointless for me to document the test further. I've made it clear it wasn't a highly rigorous test.
Well if you actually watch the video, and look at the power point slides, you'll see Ethan presents his criteria for transparency. I point it out very clearly in my article. He clearly states gear meeting those specs will sound just like every other gear meeting those specs because it's transparent and its contribution to the signal path is inaudible. The ODAC and DAC1 meet those specs (see all the AP measurements in the back of the DAC1 manual, the Stereophile measurements, or any number of other reviews as you apparently don't like my numbers). So Ta Da! By his definition the ODAC is transparent.
So there's the criteria for my claim. Will you please shut up now? (but we all know you won't. And, sorry to disappoint you, but you've proven many times over you're not worth the time at the keyboard to debate blow by blow. Others here have called you a troll multiple times including some of the moderators.)
Last edited:
That's peer review for you - you've chosen the 'objective' route in terms of marketing your designs and science is very unforgiving of lack of rigour in the longer term. If you've done the science I'll be impressed. But somehow I do get the feeling that with all the deflections in your earlier post, it just won't have been done. In which case most certainly I won't be going away any time soon.
I haven't studied Meyer & Moran in depth but I do note a lack of rigour, that's true. Its not science as I know it. But why do you make this about me RS? Its just science. Science is by nature rigorous - if you don't like (more likely too lazy for) rigor you could have chosen the subjectivist route, yet you frequently ridicule that. Seems you want to have your cake and eat it.
I must have missed something vital - why has this become about me now? I'm merely asking for you to support your claim, which has long been the tradition in science.
I have seen the criteria (set out on your blog, not yet watched the video) but they are not rigorous. For example, one thing stands out - the noise is assumed not to be correlated with the signal and no indication of the time scale for measuring the noise is given. Yet S-D converters of the kind you have chosen are fairly well known for exhibiting noise modulation.
So let me get this clear - its an argument from Ethan Winer's authority? Because Winer states something, it enters the annals of scientific truth never to be questioned? Is that what you're saying?
I see - so to you, Ethan Winer is a guru - you just take his word for it. End of.
Indeed all correct, based on individual's subjective feelings. But just because a mod calls me a troll, does that mean the claim is scientifically a valid one? I've yet to see a falsifiable definition of 'troll'.
I also referenced additional information besides just Ethan. And yeah, compared to the complete lack of credentials I've seen for you, Ethan is a relative guru. I've never seen anything that indicates we should accept your word over his.
Curious - when has science been about accepting anyone's word, no matter how credentialed?
Rhino horn?
I think the best approach was when the park rangers mounted dual 50 caliber machine guns on their jeeps.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.