Hacker News Reader

Comment by fatnoah | Hacker News Reader

This is basically the current "holy war" among some audiophiles. There's measurement-intensive approaches likes audiosciencereview and then there's the more subjective things like "soundstage" and "depth" and "clarity".

That's why it's so fun, though. Audio reproduction can be measured objectively but the experience of listening to something can only be measured subjectively. Ultimately, if the $5k D-Link router sounds better to someone, who am I to argue?

I have a mid-fi setup by audiophile standards, but it's still more of an investment than most people I know have made. When asked for advice, I tell people to focus on what sounds good to them. If it sounds good, then it _is_ good, assuming one isn't buying something to impress others.

Also, if they ask nicely, I'll give them a discount on my optimized-for-their-CPU-architecture, hand-coded, machine language version of malloc, free, and memcpy.

View on HN

this is really well summarized.

I'm in the same camp as you, I've got way more in my setup than the average person, but its mid-fi at best, and mostly on the headphone side. I like music, and I have headphones on for 5-8 hours a day, I'd like to optimize that experience. The good news is the average pair of headphones today is SO so much better than it used to be, and in a weird way some of the overpriced stuff from Apple/Sony/etc... is actually getting better sound to people that hadn't experienced it before.

I really wish there was more emphasis on the transparent + objective measurement side of things industry wide. There is in some places, but at the end of the day entire companies are built around not having objectively perfect sound reproduction, because there is a following of people that like the house-sound they engineer in to their speaker/headphones. And there is value there, its just hard to quantify.

About 12 years ago I built a pair of large (relatively) full-range bookshelf speakers for near-field studio monitor type use at my desk with a wild 3 chamber porting setup. No traditional physical filter-network whatsoever, I borrowed a decent mic, measured the response, generated an inverse impulse filter that I could use to preprocess the signal going to the amp to arrive at a "perfectly" flat frequency response. Guess what? it sounded ok at best.

Audio is hard. Your ears are weird.

> Audio is hard. Your ears are weird.

Basically this.

In the interest of disclosure, for my senior project in college (a full-year capstone project), I was on a team of students that built an analysis rig for a company that made club/venue/stadium sized speakers. Our rig reduced analysis time from 3 days to about an hour or so. However, that was just one piece of the puzzle, as the "room" also had a massive impact on sound.

That company's "demo room" was amazing. It was the first floor of an old mill building with speakers on one end of the room, and furniture at the other. At least 100-200ft if distance. It was acoustically treated, but not overmuch since one of the goals was to demonstrate the room adjustment capability. Several 5000W amps powered the whole thing, and the audio quality was amazing. It felt and sounded like listening to headphones at a moderately high volume, but when you tried to speak to the person next to you, you could barely hear your own voice.

Very cool. room correction actually has seen lots of improvement in the past decade luckily. A lot more AV receivers come with a mic and calibration process now. I'm not sure what the story for all the sonos/whatever stuff is like, but I imagine there's something. Thats not a cure-all, but its something.

15 years ago that was only for the nerdiest of the audiophile nerds, and its available to any moderately enthusiastic consumer now.

The fun thing is that generally, people do not like perfectly flat audio. From a subjective perspective, I think that perfectly flat feels like something is missing.

I think perhaps this comes from our own hearing not being objectively flat in it’s sensitivity.

Personally I don’t want to analyse my music (or my equipment) too hard, I want to enjoy it and the emotional response it invokes in me.

>Audio reproduction can be measured objectively but the experience of listening to something can only be measured subjectively

This is very true. I bought an earphone recently because it had the best technicals for its price-point. I was nice, but somewhat overwhelming to listen to for a long time. Some songs were harsh and sounded too airy. As much as I wanted to love it, I couldn't.

Then I bought a cheaper one that lots of reviewers said sounded good on anything they tried (Tin T2). It had "worse" technicals but guess what, I'm using it more than the older one and finding it way more pleasant to listen to.

> Some songs were harsh and sounded too airy.

Chances are you hear production mistakes/errors in these songs that you simply couldn't notice on the "worse" headphones. And if you produce music yourself, such "better" headphones might allow you to avoid making such mistakes/errors.

I have a friend who bought a pair of magneplanar speakers which he then sold because he could hear the pianists fingernails hitting the keys in one of the pieces. I think he decided there was such a thing as too much fidelity.

That has to be his imagination, or else maybe he's hearing some other artifact that he's mistaking for a collision between fingernail and plastic. There's no way in hell he is actually hearing what he thinks he's hearing.

What makes you so confident that is the case? I listen to lots of classical music and certainly hear musicians breathing, even on older recordings. I would imagine fingernails on a keyboard are of a similar decibel level.

Not necessarily. With good quality recording and speakers it's easy to hear things like a finger hitting a guitar string just before the string moves, lips touching a microphone, etc. I started noticing these things years ago when I first bought Blue Note recordings.

One common problem with "headphone producers" is that they often spend too much time "cleaning" audio tracks and worrying about audio leaks.

In older recordings those things were masked by other instruments, sometimes even by the room's reverb or analog noise floor, or sometimes engineers just wouldn't care about soloing tracks as much because they were on the clock and studios were expensive as heck.

Now we have everyone with super low-noise digital recording, great headphones and unlimited (home) studio time. So people spend hours obsessing about things that weren't an issue before.

Same thing with the "vinyl sound". It's a proven fact they are worse a reproducing recordings faithfully than a regular well-mastered audio CD, let alone some high resolution FLAC. Vinyl distorts the sound. But it distorts it in a way people like.

Measurement is complicated. It's not one-and-done and can easily be done incorrectly or misinterpreted. GoldenSound recently did a video on this topic, showing how minor changes in settings for analysis tools and analysis hardware can cause radically different conclusions.

There are also many techniques in hardware that make a piece of audio equipment measure objectively better, but sound worse - like feedback.

Audio is really complicated and can't be boiled down to one or two comparison charts.

ASR is essentially worthless, unfortunately.

> Ultimately, if the $5k D-Link router sounds better to someone, who am I to argue?

You can argue just fine when they fail the blind comparison test.

Reply on news.ycombinator.com