Audio specifications

[Boerd]

I found this very good explanations about audio specifications and how should you read it:
http://www.rane.com/note145.html
and thought I'd share it.

Some time ago I was comparing Parasound P3 with Cambridge 840E preamplifiers and one of the specs that jumped at me was crosstalk for P3 is 55db and for 840E is -80.
Wow! -80 seems much better!!!
If you read the fine print -55db is for 20Khz and -80 is for 1Khz.
Now let's compare apples to apples:
If we have -80 at 1Khz then at 16Khz we have 16=2 to the power of 4;
6db (per octave) * 4 = 24 yields to -80 + 24 = -56db at 16Khz
Now all of a sudden the preamps actually have the same crosstalk (P3 is in the end slightly better) as log2(20) aprox 4.35
4.35 * 6 = 26.10 so the 840E has -54db crosstalk at 20 Khz...

In the end the one who initially looked better has lower specs - you just have to do the math.

 

[highfigh]

I found this very good explanations about audio specifications and how should you read it:
http://www.rane.com/note145.html
and thought I'd share it.

Some time ago I was comparing Parasound P3 with Cambridge 840E preamplifiers and one of the specs that jumped at me was crosstalk for P3 is 55db and for 840E is -80.
Wow! -80 seems much better!!!
If you read the fine print -55db is for 20Khz and -80 is for 1Khz.
Now let's compare apples to apples:
If we have -80 at 1Khz then at 16Khz we have 16=2 to the power of 4;
6db (per octave) * 4 = 24 yields to -80 + 24 = -56db at 16Khz
Now all of a sudden the preamps actually have the same crosstalk (P3 is in the end slightly better) as log2(20) aprox 4.35
4.35 * 6 = 26.10 so the 840E has -54db crosstalk at 20 Khz...

In the end the one who initially looked better has lower specs - you just have to do the math.

Specifying power at 1KHz is even more ludicrous (That's one speed level past plaid). The manufacturers are just shooting themselves in the foot by not standardizing specs, IMO.

 

[Boerd]

Specifying power at 1KHz is even more ludicrous (That's one speed level past plaid). The manufacturers are just shooting themselves in the foot by not standardizing specs, IMO.

There are clear standards but some of the companies are trying to make their products look better - in this case Cambridge Audio (BTW 840E is VERY good, they didn't need to do that )

 

[KEW]

Specifying power at 1KHz is even more ludicrous (That's one speed level past plaid). The manufacturers are just shooting themselves in the foot by not standardizing specs, IMO.

I hate to say it, but I believe this is wrong (as frustrating as it is).
It seems the business mantra of this age is:
"A confused customer is a profitable customer"

 

[highfigh]

I hate to say it, but I believe this is wrong (as frustrating as it is).
It seems the business mantra of this age is:
"A confused customer is a profitable customer"

How long will it take for consumers to become so confused that they just don't bother to buy anything good? We're already seeing a big shift toward convenience and portability over sound quality. Music isn't mastered to sound good, it's mastered to sound loud. CE Pro magazine showed poll results of white collar people, not rich but doing pretty well, and only 26% actually cared a lot if the programming they watched was in HD. I'll look for the article for more details but people really don't care as much as equipment sales make people think. The satellite, cable and U-Verse mantra is "We have more HD channels, with more coming all the time" and more people don't know what that is than those who really care. Look at Bose- they sell the hell out of their stuff and how does it really sound? OK for some things and not so great for most. Look at HDMI with all of the problems we read about and experience- at some point, people will just get fed up with it. I know many installers who hate it and wish it had never been introduced. Then, wireless is being touted as a way to connect all of someone's equipment without running wires but at some point, there will be so many interference that many devices won't work well. Then, who will be called to fix it? Network-enabled equipment is great but at what point will people just shrug and buy the most simple things they can find?

For those who care about what they're buying, accurate specs are a must. For those who don't care, specs are meaningless, anyway.

 

[highfigh]

There are clear standards but some of the companies are trying to make their products look better - in this case Cambridge Audio (BTW 840E is VERY good, they didn't need to do that )

How much crosstalk is bad, and at what frequencies? Most people can't even hear 20KHz at all, so 55dB is fine. At 1KHz, it's more noticeable.

 

[Boerd]

...Music isn't mastered to sound good, it's mastered to sound loud....
For those who care about what they're buying, accurate specs are a must. For those who don't care, specs are meaningless, anyway.

First part (about loudness) - sad but true. Loud all the time during a song means compression (no dynamic range). That sucks!
The finest point you make here is that incorrect specs lead to distrust which leads to disinterest for the specs. That in turn let's a lot of wannabes stay in business; we really need measurements (specs).

How much crosstalk is bad, and at what frequencies? Most people can't even hear 20KHz at all, so 55dB is fine. At 1KHz, it's more noticeable.

Crosstalk is higher at higher frequencies so it is useful to know the crosstalk at 1Khz (I'd rather post the 2Khz point).
BUT - the spec is clear - it should say 20Khz +4 dbu so this is a MUST to publish for all decent manufacturers.

 

[Boerd]

How much crosstalk is bad, and at what frequencies? Most people can't even hear 20KHz at all, so 55dB is fine. At 1KHz, it's more noticeable.

How much is bad ???
Both these preamps rock in my book so 55db at 20 Khz is a solid number.
Of course you can pay 4k$ and get a preamp with 90Db separation at 20 Khz but unless your hearing is better than your dog's you can't tell the difference.
The way I test for channel separation is I have a CD with a testing tracks. One of them is saying Left then Right only in the respective channel.
What I do is disconnect one speaker, let's say the right one, turn the preamp up to 95 DB and I should not hear "Right" at all for 80 Db at 1Khz (voice is 100Hz-4Khz).
There should be no bleeding between channels but if you put your ear close enough to the speaker there will be a little bit of faint hissing in the mid driver instead of the "Right" word.

 

[highfigh]

How much is bad ???
Both these preamps rock in my book so 55db at 20 Khz is a solid number.
Of course you can pay 4k$ and get a preamp with 90Db separation at 20 Khz but unless your hearing is better than your dog's you can't tell the difference.
The way I test for channel separation is I have a CD with a testing tracks. One of them is saying Left then Right only in the respective channel.
What I do is disconnect one speaker, let's say the right one, turn the preamp up to 95 DB and I should not hear "Right" at all for 80 Db at 1Khz (voice is 100Hz-4Khz).
There should be no bleeding between channels but if you put your ear close enough to the speaker there will be a little bit of faint hissing in the mid driver instead of the "Right" word.

The only way to get true separation is dual mono preamps, sources, processors and power amplifiers. I don't see that happening anytime soon. Even then, everything is connected electrically.

I asked "how much is bad" because even modestly priced equipment sounds like the channels are totally separated in most situations because nobody listens with their ear right up to the speaker and by the time the sound gets to the main listening position, any crosstalk is usually inaudible unless the volume is extremely high. It's very easy to lose that 16dB in the distance from the speakers to the ears.

 

[JerryLove]

How long will it take for consumers to become so confused that they just don't bother to buy anything good?

But isn't that the point? Good stuff costs more to make than bad stuff. High production costs = less profit.

 

[highfigh]

But isn't that the point? Good stuff costs more to make than bad stuff. High production costs = less profit.

And mass production allows a manufacturer to spread the development/licensing costs so it's not as much of a hit for each unit but speeding up the process makes it possible to create more defective units. Good things can be mass produced and bad things can be hand built by a gnome in a tree. It's not only good stuff that has high production costs- cheap things with a high failure rate sometimes cost too much to keep making them.

People rarely care whether the company is making money- they just don't want them to go out of business while the warranty is in effect. Sure, it's contradictory but at times, people aren't very logical.

My point is that when people are confused, they either give up looking or they just take what they can get by dealing with the least crap. Neither gets them what they want and they usually won't be satisfied by the experience. That's not good for the industry.

 

[Boerd]

But isn't that the point? Good stuff costs more to make than bad stuff. High production costs = less profit.

If everybody would post correct (not misleading) specs figuring out what is good would be easy (easier). In such ideal world I agree that higher quality (specs) should be rewarded.
Instead, there are producers that publish misleading specs or no specs at all!
Here is an example - go to:
http://www.headphone.com/headphones/akg-k-702-black.php
then read the "what we think" and here is the SNAKE OIL - no measurements whatsoever for a cable they sell:

"For serious AKG K702 listeners, HeadRoom highly recommends investing in an upgraded K702 replacement headphone cord from Cardas Audio. Cardas headphone cords are renown for significantly improving dynamic range 'quickness' and extension, soundstage imaging depth & spatiality, and overall musical detail resolution clarity. It's certainly a quick worthwhile upgrade for audiophiles looking to make their beloved AKG K702 headphones perform at the very top of their game. Cardas Audio AKG K702 headphone replacement cords are available in 10ft, 15ft, and 20ft lengths: Cardas AKG K702 Headphone Replacement Cord "


Now the headphone cord above is 200$ yet there is no measurement of how and what does it improve ... We know better.
This is the ultimate insult - no specs, no measurements whatsoever - ONLY claims and phony phrases.

 

[Boerd]

But wait - there is more

Go to:
http://www.akg.com/site/products/powerslave,id,1082,pid,1082,nodeid,2,_language,EN,view,specs.html
and please note the specs.
These headphones have low impedance and high sensitivity: 62ohms (when, in what conditions???) and 105db sensitivity (when, in what conditions???).
Both specs much better than my Grado R125.
Now on this website:
http://www.headphone.com/headphones/akg-k-702-black.php
go to the "what we think" and you read:
"For listeners wishing to use the AKG k702 with portable audio players like iPods, MP3 players, & PCs/laptops, a dedicated headphone amp of some sort is simply a must for maximized audio performance due to the headphone's VERY low sensitivity -- despite its fairly efficient 62-ohm impedance rating. Our entry-level HeadRoom 'Mobile Line' portable headphone amps ($99-$159usd) will do the trick nicely but the more powerful, accurate, and dynamically-quick HeadRoom 'Micro' series of products come very highly recommended for this superb headphone."

Now really, my Grado R125 can blast my ears (I don't abuse that) and these cans have much better sensitivity AND lower impedance yet I DO NOT USE a head amplifier.

Now either:
105 db / 62 ohms is misleading and incorrect (unlikely in my opinion)
or
headphone.com got some BS ...

 

[son-yah-tive]

I can remember a time when you didn't see anything but All-in-one junk at department stores. Now there is equipment out there that's touted GREAT, at these same stores. Now, the choices for those of us who want really GREAT SOUND, have to almost go all internet. The OLD stores that carried GOOD to High End equipment that use to be down the street, are gone. You almost can't compare anything before buying. That's where educating yourself first comes in. And, it's a SHAME, and hate to say it, but, 'I MISS THE OLD DAYS'. That's the reason I ended up with a BOSE system 8 years ago. There was no store with good enough equipment to really do any comparison. Well, the internet brought me back in the High End circle. Thanks for that at least!!

 

[highfigh]

I can remember a time when you didn't see anything but All-in-one junk at department stores. Now there is equipment out there that's touted GREAT, at these same stores. Now, the choices for those of us who want really GREAT SOUND, have to almost go all internet. The OLD stores that carried GOOD to High End equipment that use to be down the street, are gone. You almost can't compare anything before buying. That's where educating yourself first comes in. And, it's a SHAME, and hate to say it, but, 'I MISS THE OLD DAYS'. That's the reason I ended up with a BOSE system 8 years ago. There was no store with good enough equipment to really do any comparison. Well, the internet brought me back in the High End circle. Thanks for that at least!!

Department stores always carried less than great equipment but once they saw that money could be made on better, they jumped into the pool. They didn't train their people to deal with technical questions but they did hire people who know when to ask for the sale. When it comes to sales, that's more important than knowing all about the equipment.

The high end stores- you do know why they're gone, right? Because people used them to get information and bought someplace else, for a little less. Sometimes, they found a whore who would sell it for a lot less but buying habits have changed completely since the hey-day of high end stores.

 

[AcuDefTechGuy]

If we have -80 at 1Khz then at 16Khz we have 16=2 to the power of 4;
6db (per octave) * 4 = 24 yields to -80 + 24 = -56db at 16Khz
Now all of a sudden the preamps actually have the same crosstalk (P3 is in the end slightly better) as log2(20) aprox 4.35
4.35 * 6 = 26.10 so the 840E has -54db crosstalk at 20 Khz...

Extrapolations are fine, but they are just estimations and don't always hold up true.

For example, I've seen crosstalks for Bryston amps that are -70dB @ 1kHz, but remains @ -70dB even @ 20kHz.

Anyway, I think the standards most people believe to be good include:

SNR: 90dB (20Hz - 20 kHz)
THD: 0.1% (@ full rated power 20Hz - 20kHz)
FR: 20 Hz - 20 kHz +/- 3dB
Crosstalk: -60dB (20Hz - 20 kHz)

 

[Boerd]

Extrapolations are fine, but they are just estimations and don't always hold up true.

For example, I've seen crosstalks for Bryston amps that are -70dB @ 1kHz, but remains @ -70dB even @ 20kHz.

Physically this isn't possible UNLESS you overestimate the crosstalk at 1Khz.
Please see http://www.rane.com/note145.html
That Bryston amp really has -94Db separation at 1Khz if it has -70 at 20 Khz.
It is great however to see good manufacturers under reporting how good their stuff is just to make the specs easier to read.

Anyway, I think the standards most people believe to be good include:

SNR: 90dB (20Hz - 20 kHz)
THD: 0.1% (@ full rated power 20Hz - 20kHz)
FR: 20 Hz - 20 kHz +/- 3dB
Crosstalk: -60dB (20Hz - 20 kHz)

Yes - these specs are achievable and anything beyond this is precision for dog's ears...