• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

How would you answer these questions? (Acoustics Insider Darko Interview)

Keith_W

Major Contributor
Joined
Jun 26, 2016
Messages
3,840
Likes
9,271
Location
Melbourne, Australia
Acoustics Insider is a pro audio channel that is usually science based. They decided to interview ... Johnny Darko. They got their membership base to put together some questions which were given to Darko. IMO the questions were really very good. Darko's answers ... not so much.

Since Darko failed so badly, I thought I would throw the questions open to ASR members. But first, here is the video (don't worry, the clicks go to Acoustics Insider and not Darko). I have typed the questions and paraphrased Darko's replies to keep it brief - hopefully I have preserved the meaning.


6:09: "What is one thing mixing engineers should know about audiophiles, and what is one thing audiophiles should know about mixing engineers?"
Darko: Not all audiophiles are the same. Audiophiles can be pretty painful, of a certain demographic. My passion for audio gear was started by my passion for music. Other audiophiles are obsessed with sound and listen to a limited range of music. Being an audiophile is a broad church. People like me are rare (i.e. "music first audiophiles"). Engineers think audiophiles are gear obsessed, they need to know that people like me exist. They also think of loudspeakers as a tool, and must think that audiophiles are nuts - they overspend on speakers and put them in rooms that sound like a bathroom.

14:50: "What characteristics of engineering that goes into a speaker that most affects the sound stage, and what do you look for in a speaker to get excellent imaging?"
Darko: Lower distortion and better time alignment. Coaxial speakers tend to give better imaging and soundstage. I don't understand why.
Comment: No, Darko. Imaging/soundstage is a function of ITD, ILD, and a few others, so it is the interaction of the speaker and the room that produces the soundstage. If we focus on the speaker only, it is its directivity characteristics, smoothness of off-axis response, and how close to a point source ideal we get. Also you could have said something about the importance of reflections which you are actively trying to kill with your acoustic treatment.

16:40: "As you appreciate the importance of both objective measurements and subjective impressions, how accurately do you think we can predict the sound of a loudspeaker based on measurements?"
Darko: I am fascinated by measurements. I read Stereophile's section on measurements and read how Atkinson correlates measurements to his subjective impression. There is no 100% objective analysis of a loudspeaker because you have to interpret the results, and that is by definition a subjective process. I don't buy into this dichotomy of subjectivity vs. objectivity. I don't think there is a 1:1 mapping of "if we measure this we hear this". It's not a solved problem. It can help speaker engineers understand if they've made a mistake, but every speaker engineer i've spoken to says that we measure up to a point then we sit down and listen.
Comment: how well you can predict the sound of a loudspeaker based on measurements depends on (1) the quality of the measurement and (2) how much experience you have with correlating measurements to sound quality. I am fairly confident that if I see x I will hear y, but only for the measurements I know I can correlate. Speaking for myself alone, there are still some subjective aspects to sound which I haven't correlated to measurements, and some measurements that I don't understand how they correlate to sound. This is a function of my own inexperience. For example, if someone on ASR says "a 10ms difference in subwoofer time alignment is enough to kill tight bass", I wouldn't know until I hear it myself.

19:00: "Do you feel there are aspects of a subjective experience that are not portrayed in measurements?"
Darko: Imaging. You could see a low distortion figure and think that will image well but that's not a guarantee. For example, I reviewed two speakers, one went down to 54Hz, the other to 64Hz, but the 64Hz speaker sounded like it had more bass. It's a loose guide for consumers, it's a shortcut. Life is more nuanced than measurements.
Comment: It's more likely that he doesn't understand why the 64Hz speaker sounded like it had more bass. All sorts of reasons abound, e.g. the missing fundamental effect, Darko may not actually know what low bass sounds like (lots of people who think they know what bass sounds like actually don't, most often they are mistaking midbass for low bass).

21:47: "Some music sounds good on most speakers, but other music will sound good on some speakers but terrible on others. Which audio recording / processing qualities ensure good sound on most speakers?"
Darko: Everything comes down to the recording and mastering quality. There is a whole genre of "audiophile music", artists nobody has ever heard of. These sell well because gear obsessed audiophiles are into the sound of those recordings. I'm not. I don't know what goes into it. But dynamic range compression definitely affects quality. It is important to maintain the dynamic range.

31:32: "At what level of treatment are we able to hear real differences in speakers? For example, do we need a certain RT60 value? What do we aim for?"
Darko: I didn't do it in an incremental way. I went from a room with 0.7-0.8 milliseconds (sic) to have flat, roughly 0.3. I don't know what the tipping point is. I'm guessing the tipping point may be 400-500ms? It's not only the value, but the stability of the value, it can't be 0.5, 0.3, 0.5, 0.3. It needs to be flat across the board.
Interviewer interjects: the more you treat the room and the closer you get to an anechoic chamber, the more of the speakers you will hear.
Darko: People in your world are different. You need to have a treated room. Audiophiles won't do it because they have to balance home life with sound.
Comment: Well, at least he measured and he's in the right ballpark with his recommended values. I would say look up the DIN 18041 standard because the target is different depending on the size of your room and your intention.

I skipped some of the Q&A because they were a bit boring.
 
I can’t correlate everything in measurements to the sound I hear. Must be the measurements’ fault.
 
To be somewhat fair: At the beginning when Darko said it was important for the viewers to know he's not a loudspeaker designer and "I can't answer technical questions about loudspeakers." He makes it clear that instead he tries to describe his experience listening to loudspeakers.

(Though I agree with some of the criticisms of his replies).
 
I can’t correlate everything in measurements to the sound I hear. Must be the measurements’ fault.
I know this was meant to be derisive but there absolutely are measurements that are relevant to audible phenomena that aren't regularly measured. The famous one of course was the focus on on-axis while neglecting off axis, but short term compression behavior is going to be pretty audible for how lively (or not) a speaker sounds. Part of why compression drivers sound the way they do is because at sane levels they have effectively zero compression.
 
I know this was meant to be derisive but there absolutely are measurements that are relevant to audible phenomena that aren't regularly measured. The famous one of course was the focus on on-axis while neglecting off axis, but short term compression behavior is going to be pretty audible for how lively (or not) a speaker sounds. Part of why compression drivers sound the way they do is because at sane levels they have effectively zero compression.
There are, of course, incomplete measurements. But one cannot simply assume that the failure to connect intuition (what you hear as you intepret it) to measurements is always a measurement problem.
 
What is one thing mixing engineers should know about audiophiles

Nothing. Mixing engineers should do their job as well as they possibly can, and not pander to any one particular segment of the population. If some particular segment wants an affected sound, they can get it with equipment that modifies the sound.

We have about 340 million people here in the USA, and much of the music produced here is sold in other countries around our world of 8 billion. Audiophiles are insignificant.

what is one thing audiophiles should know about mixing engineers?"

That an audiophile's personal opinions are not within the ability of the mixing engineer to predict. He's got his job to do, and he can't pander to you, or your buddy, or the guy down the street. It's just ... not ... possible.

However, it IS possible for him to satisfy his producer, Mr. Pursestrings. What Mr. Pursesrtrings wants comes first.

"Some music sounds good on most speakers, but other music will sound good on some speakers but terrible on others. Which audio recording / processing qualities ensure good sound on most speakers?"

The question is backwards. "If recording and mixing were done on neutral speakers of high quality capabilities, which audio recording/processing qualities would NOT ensure good sound?" Would there be any?

---------------------------------------------------------------------------------

I agree with Keith's comments on the other questions. :)

Jim
 
Last edited:
"What is one thing mixing engineers should know about audiophiles, and what is one thing audiophiles should know about mixing engineers?"
Both should admit to the other that they have spent a lot of time tweaking, a lot of time repeating received wisdom and not much time studying.
"What characteristics of engineering that goes into a speaker that most affects the sound stage, and what do you look for in a speaker to get excellent imaging?"
Directivity characteristics on and off-axis.
"As you appreciate the importance of both objective measurements and subjective impressions, how accurately do you think we can predict the sound of a loudspeaker based on measurements?"
"Predict the sound": let's understand that loosely, as meaning will what plays match what you expect to hear. So in that sense, very accurately with a full complement of high resolution measurements. For me, this means a two channel stereo context. For multichannel and immersive formats with complex signal matrixing, I don't know what to expect, lacking experience.

Going a bit further, this means mandatory EQ for frequencies below 1kHz, where the room has the most influence.
"Do you feel there are aspects of a subjective experience that are not portrayed in measurements?"
There's a lot to say here.

We do not have any comprehensive models of hearing as a whole to investigate auditory phenomena generally. We have increasingly great physical measurements and decent methods of interpreting those thanks to a lot of research efforts. Measurement is getting more precise, and the current issues are not to do with the inaccuracy of measurement equipment (for the most part), but to do with creating new batteries of tests that attempt to coordinate multiple physical measurements meaningfully. For example, the coarse spatial sampling in multipoint measurements in correction systems like Dirac's. Or the precise spatial sampling of Klippel's NFS.

We do not have any models that explain, plainly, small differences between different speakers and room setups. We do not have a clear approach that help us use what is known of normal auditory adaptation and how to apply it to room design. We have good guidelines but no good models nonlinear distortion (which is a very complex topic outside of THD/IMD, as when dealing with codecs, for example).

In room acoustics particularly, there is a disparity in advice and approach to what is considered optimum, and nothing that allows precise comparison. You have books and papers that contradict each other (more reflections! less reflections!).

Despite all that, for the most part, good sound can be had cheaply. Understanding precise tradeoffs is the main thing allowed by measurements, and, really, where there are lapses in what measurements can present meaningfully, ambiguity sets in. There is lots of good advice to be had, but in the end some things can only be understood by listening, which is actually quite difficult to do in current circumstances, with audio dealers closing, and many models not available for easy comparison. This is in the context of speakers and headphones. For electronics, I believe these questions are settled.
"Some music sounds good on most speakers, but other music will sound good on some speakers but terrible on others. Which audio recording / processing qualities ensure good sound on most speakers?"
This is the problem of translation. It is very complicated and not currently solvable. Some amount of manipulation through EQ at home solves most things. I don't want to get started on the rest.
31:32: "At what level of treatment are we able to hear real differences in speakers? For example, do we need a certain RT60 value? What do we aim for?
See the answer above on room acoustics. Targets are unclear. Applying treatment is very audible, so much so that the result is hard to interpret.

I generally think most treatment is unnecessary, many will say otherwise. Where I find treatment helpful is in and around the room's transition region, where SBIR is strong, and unfortunately there are few practical methods for dealing with that that are not bulky.
 
There's a lot to say here.

We do not have any comprehensive models of hearing as a whole to investigate auditory phenomena generally. We have increasingly great physical measurements and decent methods of interpreting those thanks to a lot of research efforts. Measurement is getting more precise, and the current issues are not to do with the inaccuracy of measurement equipment (for the most part), but to do with creating new batteries of tests that attempt to coordinate multiple physical measurements meaningfully. For example, the coarse spatial sampling in multipoint measurements in correction systems like Dirac's. Or the precise spatial sampling of Klippel's NFS.

We do not have any models that explain, plainly, small differences between different speakers and room setups. We do not have a clear approach that help us use what is known of normal auditory adaptation and how to apply it to room design. We have good guidelines but no good models nonlinear distortion (which is a very complex topic outside of THD/IMD, as when dealing with codecs, for example).

In room acoustics particularly, there is a disparity in advice and approach to what is considered optimum, and nothing that allows precise comparison. You have books and papers that contradict each other (more reflections! less reflections!).

Despite all that, for the most part, good sound can be had cheaply. Understanding precise tradeoffs is the main thing allowed by measurements, and, really, where there are lapses in what measurements can present meaningfully, ambiguity sets in. There is lots of good advice to be had, but in the end some things can only be understood by listening, which is actually quite difficult to do in current circumstances, with audio dealers closing, and many models not available for easy comparison. This is in the context of speakers and headphones. For electronics, I believe these questions are settled.

I love your reply! Clicking on "like" wasn't enough, I had to quote it and give it a "like" again :)

Correlating subjective experience to measurements is something I believe we should be discussing more. I hinted briefly in my post that other people can not correlate the effect of a measurement on a subjective experience for me. I have to do it myself. If someone else says "RT60 of 500ms is too wet, RT60 of 250ms is too dry", I can appreciate that on an intellectual level, but I have no idea what it sounds like unless I listen with my own ears. At best, reading about someone else's experience tells me where to look.

The reason why this is important is because I can often hear things in my system and I don't know what measurement correlates with what I am hearing. Fortunately, I am a bit scientifically rigorous and rather than dismissing the measurement as worthless (what subjectivists do) or dismissing the subjective experience (what some objectivists do), I think to myself that it is because I don't understand the measurement, or I have heard something that I have not captured with my measurement technique. Both humility and confidence are required - confidence to know that I heard something real, and humility to admit that I don't understand. And conversely, humility to admit that I may have been imagining it.

For example, I can easily hear that a speaker is "dynamic" in that it transients have attack and sound clean, but what measurement does this correlate to? I am VERY confident that this is a real subjective phenomenon. I thought it might have something to do with the sharpness of the step response, but I was wrong. I have since found that you need to measure loudspeaker compression and plot the input against the output - e.g. for every 5dB increase in volume my DAC outputs, I should measure a corresponding linear increase in loudspeaker volume. At some point the loudspeaker will compress and distort. At what volume this happens and by how much is part of the explanation. Other measurements which might explain the subjective impression are early reflections in the ETC, ringing which might smear the transient, and so on.

Although Amir does test speakers at different volumes as part of his standard speaker review, he only reports the measured volume and not the input or whether the speaker responds linearly to the input.

I am far from an authority on measuring speakers so I don't know if this measurement I came up with is (1) actually something I came up with and not a phenomenon which is already well known, or (2) whether it is scientifically valid or not.

That is the problem with correlating subjectivity with objectivity IMO. All of us need to take that journey individually, nobody can do it for you. Some people have an unfortunate mindset of dismissing one thing or another without proper investigation. To my mind that limits learning and understanding so I don't do it.
 
To be somewhat fair: At the beginning when Darko said it was important for the viewers to know he's not a loudspeaker designer and "I can't answer technical questions about loudspeakers." He makes it clear that instead he tries to describe his experience listening to loudspeakers.

(Though I agree with some of the criticisms of his replies).
Yet he implies all the time that he does just that. Why anyone listens to this guy.....
 
yeah i like this hobby but i dont like it THAT much

i'd rather go over work compliance docos
 
It is well known experience that measurment of a loudspeaker does not tell really how it will sound. There are details in the constsruction which will not show up in a measurment but influence the sound impression. So all what is said by Darko is somewhat right and wrong because much is just speculation. And some listeners would like some reverb in the room, some want it desert dry. That is just different taste.
 
havent we come to the maxim that...

a speaker that measures good will not sound terrible

a speaker that measures bad will not sound good

true?
 
Darko: Imaging. You could see a low distortion figure and think that will image well but that's not a guarantee. For example, I reviewed two speakers, one went down to 54Hz, the other to 64Hz, but the 64Hz speaker sounded like it had more bass. It's a loose guide for consumers, it's a shortcut. Life is more nuanced than measurements.

This is the sort of thing a lot of audiophiles come out with but at least they have the excuse that audio's not their day job.

He really ought to start learning something about what he's doing, I don't see that he has the excuse not to.

How can he not know why one speaker seems to have 'more bass'? I mean that's just basic stuff.
 
Darko: Imaging. You could see a low distortion figure and think that will image well but that's not a guarantee. For example, I reviewed two speakers, one went down to 54Hz, the other to 64Hz, but the 64Hz speaker sounded like it had more bass. It's a loose guide for consumers, it's a shortcut. Life is more nuanced than measurements.

This is the sort of thing a lot of audiophiles come out with but at least they have the excuse that audio's not their day job.

He really ought to start learning something about what he's doing, I don't see that he has the excuse not to.

How can he not know why one speaker seems to have 'more bass'? I mean that's just basic stuff.

Best to jump to the timestamp and listen to what he actually said, rather than my impression of what he said. To be fair to him, he did speculate on a few reasons why but it was clear from listening to him that he didn't measure. Because if he did, he would know exactly why.
 
Okay I've got nothing better to do (at least nothing palatable) so will have a go at the questions:

6:09: "What is one thing mixing engineers should know about audiophiles, and what is one thing audiophiles should know about mixing engineers?"

All audiophiles should be made, at gun point if necessary, to go to a studio and watch a recording being made. This would stop many of their bizarre assumptions.

Mixing engineers don't need to know anything about audiophiles in order to do their job properly.

"What characteristics of engineering that goes into a speaker that most affects the sound stage, and what do you look for in a speaker to get excellent imaging?"

A good speaker - that is one with even on and off axis dispersion and low distortion will image superbly given correct placement in the right environment.

16:40: "As you appreciate the importance of both objective measurements and subjective impressions, how accurately do you think we can predict the sound of a loudspeaker based on measurements?"

You can't 'predict sound' but you can predict from the measurements as to whether a speaker will make a dog's dinner of a recording, or render it reasonably listenable. There will always be some differences in 'presentation' even between good loudspeakers, but that is less important for enjoyment of music.

19:00: "Do you feel there are aspects of a subjective experience that are not portrayed in measurements?"

Yes, since we are talking 'subjective' my experience will differ day to day even though the system and the recording do not change. I am the variable.

21:47: "Some music sounds good on most speakers, but other music will sound good on some speakers but terrible on others. Which audio recording / processing qualities ensure good sound on most speakers?"

I agree with Darko here. Overuse of limiting and compression are the only things that will render professional recordings unlistenable on a competent system.

31:32: "At what level of treatment are we able to hear real differences in speakers? For example, do we need a certain RT60 value? What do we aim for?"

Not entirely sure what this question means. I'd say make sure the room is damped sufficiently to not be an echo chamber, and if bass is boomy or uneven, sort that out with appropriate speaker choice and EQ. Once you've done that other factors are more important. There's no set in stone RT time it will always vary with room size and personal taste.
 
Best to jump to the timestamp and listen to what he actually said, rather than my impression of what he said. To be fair to him, he did speculate on a few reasons why but it was clear from listening to him that he didn't measure. Because if he did, he would know exactly why.
Okay did that. He doesn't mention the most likely explanation. But again, this does show the shortcomings of the purely subjective review. His cop out is 'Go and listen for yourself.' Fine, if that's true then what's the point of someone else's purely subjective review? He might as well not bother doing them.
 
19:00: "Do you feel there are aspects of a subjective experience that are not portrayed in measurements?"
Darko: Imaging. You could see a low distortion figure and think that will image well but that's not a guarantee. For example, I reviewed two speakers, one went down to 54Hz, the other to 64Hz, but the 64Hz speaker sounded like it had more bass. It's a loose guide for consumers, it's a shortcut. Life is more nuanced than measurements.
Comment: It's more likely that he doesn't understand why the 64Hz speaker sounded like it had more bass. All sorts of reasons abound, e.g. the missing fundamental effect, Darko may not actually know what low bass sounds like (lots of people who think they know what bass sounds like actually don't, most often they are mistaking midbass for low bass).

21:47: "Some music sounds good on most speakers, but other music will sound good on some speakers but terrible on others. Which audio recording / processing qualities ensure good sound on most speakers?"
Darko: Everything comes down to the recording and mastering quality. There is a whole genre of "audiophile music", artists nobody has ever heard of. These sell well because gear obsessed audiophiles are into the sound of those recordings. I'm not. I don't know what goes into it. But dynamic range compression definitely affects quality. It is important to maintain the dynamic range.

It's a strange example to give because it is surely measurable. There are aspects of subjective experience like the roadworks outside my house that are far less measurable, at least in the context of usual frequency graphs and whatnot. And far less debatable subjectively, if you can't hear the roadworks or planes going overhead then you are probably deaf.

He doesn't answer the 2nd question but tbf he couldn't really. I kinda agree with his jab at "audiophile music" where you can hear a pin drop and it's recorded impeccably, but it's a soulless cover of a Stevie Wonder tune. I disagree with the premise in the question that some music will sound good on some speakers and terrible on others. I'd say that's true on a larger scale but even then only at the extremes. Frank Sinatra singing through King Tubby's sound system and stuff like that.
 
Back
Top Bottom