The Unix Philosophy of Audio - Page 2

Here's one engineer's excerpted take on dsp:

Take a stream of bits coming off a CD, HDD, SD card, or any other digital source, we now want to perform some kind of equalization on it in the digital domain, for example, let's say we want to add a 4dB boost at 5kHz with a Q of 0.5. If the process were transparent and 100% perfect, the only difference between the input and output signals would be a 4dB boost at 5kHz with a Q of 0.5; phase, noise-floor, and every other possible parameter shall be 100% identical. Nothing added and nothing taken away except for the EQ which we wish to perform. Sounds easy enough but we're not quite there yet.

I wouldn't underestimate the power of signal processing. Barring distortion or interference, you can get extremely close to the original signal. If those minute differences in the signal destroy your audio experience, it's safe to say that nothing will satisfy you, and that all the bad things you hear in a system are all made up in your head. If you want to be picky, vinyl has its own problems and I'd consider all the microscopic contaminants in the grooves to be distortion as well.

Yeah, just bar distortion, ok. Distortion is the name of the game.


But only if you want to stay true to the signal. I've already thrown out the idea that many people obviously don't.

I didn't understand exactly what you were trying to say there, but I don't stress over details in my audio. Every speaker/headphone is unique, even within their same models. I listen to whatever sounds good and doesn't cost more than I'm comfortable spending.

You just have to realize that after a certain point, you're paying for hype.

HDTV's for example.Manufacturers have got them down to a science. You can buy the best for relatively cheap.

You can't tell me that speakers based on simple electrodynamicsare harder to make than HDTV's. They're not. It's just that nobody seems to be able to agree on what sounds good and what doesn't. This gives manufacturers free reign to charge what they like and amazingly, people pay for it.

This is a completely different topic. HDTVs are filled with discrete components and designs easily licensed and mass produced, even rebadged. Speakers are a different beast entirely. The fact that the demand for high fidelity speakers is exponentially lower than for HDTVs alone accounts for the drastically different marketplace. Look at floor space for TVs compared to speakers and number of customers in each respective area in any big box store. While it is my opinion that mass market speakers are as overpriced as low and mid end sony tvs, hand built and finished cabinets simply have a huge amount of overhead and labor. Those rebadging tv distributor brands hardly have to do a lick of R&D. Many of them are almost nothing more than resellers or middle men - not so for a legitimate speaker company with their own engineered vision for sound reproduction.

It's true that it's much easier to fool the ears than it is to fool the eyes. Even so, disregarding those speaker companies with no integrity, it's easy to see why speakers cost what they do, once you know what goes into them.

If you have a few hours, or better yet a few days to spare, I'd suggest reading through the following thread on diyAudio, yes, all 1000+ posts. The technical issues faced are far from simple.

http://www.diyaudio.com/forums/showt...hreadid=100392

If the link doesn't work, go to the Loudspeakers forum and search for the thread entitled "Beyond the Ariel".

It's not so much the manufacturing capability as it is design. A TV is much more complex than a speaker will ever be. Even high order crossovers with upwards of a hundred components can't hold a candle to the circuits found in TVs.

I might add that hand built cabinets say nothing about the sound of a speaker either.

But in any case, it's your money. You can do what you want. Some people spending $1,000's on rare metal cables... well you get the idea. Electromagnetics and circuit design have been around for ages. There's no hand waving in these sciences. If you're going to blow money, at least let it go to a charitable cause.

edit: I want to add that the acoustics beyond the electronics of a system are an entirely different realm. I'll agree at that point it becomes an art more than a science. But keep an eye on your pocketbook. I'm an electrical engineer by day and a concert violinist/pianist by night. I'll tell you that nobody has gotten a violin down perfectly, not even Stradivarius could make a perfect violin 100% of the time. Pianos are a little easier. But by god, there are a crap load of violins out there that may cost 10 to 200 thousand dollars simply because of their name that sound well... like crap.

I guess my philosophy to audio is not to lose one's head in the hype. I like to balance science and art. Sometimes they go together. It's really amazing.

How complex something is means nothing if you can mass produce it. You could say that a tv is much more complex than a car, too. A digital camera is much more complex than a film camera, but digital cameras are just as cheap now as the next electronic gadget. What about a house? Jeez, there's not much to a house. Have a care about the design and manufacturing issues.

I've totally forgotten what we were originally talking about. Anyway, I think I was originally making a reference to the design work that goes in developing speakers and TVs and not the actual manufacturing.

There's a lot of 'mysticism' in audio. Some of it has some logic behind it. Other stuff is bs. That's all I'm going to say. CD players that measure the same will sound the same. That's where I stand.

Speakers are more of an art, but there's only so much I'm willing to pay for art. Why would you want to pay exorbitant amounts of money for a TV that plays everything with a persistent green tint? It could be a really pretty green, but I'd never pay good money for a TV like that. What's the point? Speakers are the same thing. I'm not shelling out several grand for 'defective' speakers. If I want a hand in dictating the sound I'm listening to, I'll go ahead and perform the music myself as the artist.

I'll semi-quote the manufacturer's of my speakers because I simply love their slogan. You don't listen to your speakers, you listen through them.

Designing flat speakers that are coveted by audiophiles for their pure sound is a noble goal. I'll pay for that. Make speakers that color the sound, put a glossy finish on them, slap a gold logo on them and tout them as better than the rest... you've got to be kidding me.

Recap: Too much hand waving in audio. It's not necessary.

There are plenty of engineers and other objectivists that disagree about how you make a "flat" speaker. I already went through a short list of measurements that a speaker needs to be checked against. But even then, not all of them are supposed to be "flat." There are too many variables. For another example, what do you do about the off axis response of a speaker? It's not going to measure flat unless you had perfect spheres radiating. Well, there're lots of ideas. Recently, the flat off axis (as much as possible) high frequency monopole approach has been popular amongst the diyers and other companies with increased disclosure about their design principles. They think it is better to have the same sound throughout the room, at least in one direction. However, another camp says that not only is the diffuse field response of the ear tipped up, but all real concert environments have little to no upper frequency room response, and therefore the off axis response of the speakers should be tipped down as low as possible or it will sound too bright. Another camp says that it's more important that the polar response of the speaker remain as constant as possible for as large a range as possible so as to prevent our ears from being confused. We haven't even talked about how these different dispersion designs couple to the room. No matter what, none of these designs will measure flat off axis with current technology. There's plenty of room for R&D. Lots of room for distortion analysis and interpretation, too.

Now I don't know about high end cd players, but I can easily see justification for costly speakers - to a point. What I can't stand are conventional box speakers in crappy 1/2" particle board cabinets costing a couple thousand bucks, or speakers that are +/- more than 2-3db, like the Zu Druids or Silverline speakers. However, the real engineered speakers are out there.

I subscribe to...

the simple theory of audio:

I look at measurements (if available) to make sure a component isn't completely screwed up. Then use my ears to make sure it sounds good. And add a dose of common sense to figure out if I'm getting a fair value for the price.

You can split hairs about jitter/feedback/cables/testing/whatever (which, IMHO is pretty good for passing the time and nothing else) but the bottom line for me is, if it doesn't get my toes tappin', then it doesn't go in my system.