Home Theater Forum and Systems banner

1 - 20 of 64 Posts

·
Registered
Joined
·
266 Posts
Discussion Starter · #1 ·
Ok lets open a can of worms.

The audio market is such a mess, manufacturer are playing with number and make them meaningless or sometime they gave you no number at all (like Bose)

I would like to know what do you think would be the minimal or acceptable specification for different audio and why not video gear.

I will start this up with speakers (cabinet not drivers)

Sensivity: in DB/watt/meter (giving sensivity in DB with no reference means nothing)
frequency response: lower limit - upper limit @ -3DB ex: 40hz - 40Khz @-3DB
maybe the radiation patern, but I am not sure this would be use by a lot of people

amplifier
power is frequently given for one channel at 1Khz
I beleive it would be better to have at least the power bandwith ex. 105W from 10hz to 40Khz
and possibly power all channels driven simultaniously

give your input on that

minimal spec for receiver I am not asking about how many HDMI input and output a receiver should have, but just the spec so we can compare.

Like a reference when they say that a receiver can put out 105W per channel, put a standard (like distortion level, but they should all have the same reference for example number of watt @ .1% distortion

you can add other gears with spec

Please keep it cool, not sure manufacturer will check this tread, so be cool and do not start a fire
 
G

·
Almost all speakers have specs listed. Bose is the only I know that doesn't.

As far as amps and AVRs go, I do wish their could be some SAE standards or something. Some are under rated, some are over rated. It's a mess.

Without any standards though, the information on quality equipment is out their. Not really fair to the normal consumer who isn't really in to all of it. On the other hand, it is so easy to do research now days.
 

·
Registered
Joined
·
1,060 Posts
The FTC has regulations which specify the minimum criteria for test conditions and reporting of amplifier power ratings. It provides for multichannel amplifier sine wave continuous average power output ratings based on two channels driven into the specified impedance at 1kHz (or the frequency response for the rated power output) and a manufacturer selected and stated THD.

http://ecfr.gpoaccess.gov/cgi/t/tex...;view=text;node=16:1.0.1.4.52;idno=16;cc=ecfr
 

·
Registered
Joined
·
266 Posts
Discussion Starter · #4 ·
Almost all speakers have specs listed. Bose is the only I know that doesn't.

As far as amps and AVRs go, I do wish their could be some SAE standards or something. Some are under rated, some are over rated. It's a mess.

Without any standards though, the information on quality equipment is out their. Not really fair to the normal consumer who isn't really in to all of it. On the other hand, it is so easy to do research now days.
I agree that they have some spec for speakers, but spec like:
frequency response 40hz to 22Khz is meaningless, you probably saw those little speaker for computer that have spec like 20hz to 20Khz, we know it is impossible, those spec are probably at -60db
 

·
Registered
Joined
·
1,747 Posts
Sensivity: in DB/watt/meter (giving sensivity in DB with no reference means nothing)
frequency response: lower limit - upper limit @ -3DB ex: 40hz - 40Khz @-3DB
maybe the radiation patern, but I am not sure this would be use by a lot of people
Sensitivity should be given in db/V/m because watts change with respect to frequency. 2.83v/m is the general standard as it translates to 1w @ 8 ohm and 2w @ 4 ohm.

Frequency response lower/upper limits are mostly meaningless because they're affected strongly by things like room influence, our own ear's hearing transfer function, and +/-3db is a very wide, 6db window with lots of room for coloration. I'd prefer something closer to showing that a speaker is +/- 2db or better between 200hz - 10khz. An actual, high resolution graph is prefered to just a one-line specification as this will also indicate if/where the speaker starts to roll off. Remember though, a graph given at 75db may not represent the speaker's response at 90db. Thus there is also a strong need for a power compression information. Even if a speaker is perfectly flat to 20hz that doesn't mean it meaningfully reproduce 20hz.

Speakers should also be specified not only with their axial frequency response (which can be just a matter of passive equalization) but also their radiated frequency response into various directions (20 degrees vertically, 15/30/45/60/75 degrees horizontally) as this more closely represents the speakers consistency as far as what we actually hear in the far field (a combination of direct and reverberant response). Dr. Floyd Toole also suggests replacing the on axis frequency response itself with "listening window" response - an averaging of the response within 15 degree axes.

Also necessary is complex impedance (Z-chart) with phase angles. Most speakers claiming 8 ohm loads are much more taxing on an amplifier than that suggests.

Finally, I feel any good speaker will have a clean waterfall graph. Waterfall graphs don't separate the best from the rest, but they'll quickly throw out the worst from consideration. Waterfall graphs are just one form of showing decay and there's other valid ones as well.

A few companies that do IMO give great examples of how commercial speaker specifications should be include:

http://www.jblpro.com/catalog/support/getfile.aspx?doctype=3&docid=569
http://www.ascendacoustics.com/pages/products/speakers/SRT/srtmeas.html
http://philharmonicaudio.com/philharmonic1.html
http://www.bambergaudio.com/s5tmw_specs_r2.pdf

amplifier
power is frequently given for one channel at 1Khz
I beleive it would be better to have at least the power bandwith ex. 105W from 10hz to 40Khz
and possibly power all channels driven simultaniously
All Channels driven is useful for a 2 channel amp, but less so for a 7 channel amp because that's not a real world scenario being represented. Thus I am most interested in output at low impedances into 2 channel loads - the Power Cube.

ACD does help us see power supply robustness, but can also be irrelevant in the grand scheme of things. When do you drive all seven channels to full power?

The other important measurement for an amplfiier is its FFT Distortion spectrum @ < 1w where it will spend most of its power.
 

·
Registered
Joined
·
237 Posts

·
Registered
Joined
·
139 Posts
Its all about marketing its not actually what the rating is its how many Sheep can you make drink from the sand....It should be simple this many watts this frequency range at all channels driven.....period....But here is the reality of it Company A: Our receiver named the One has this much power Company B: Well we will call our receiver the One and Only and it has the same amount of power. Customer: Hmmmmmm do I want the one or do I want the One and Only wow how could i not go with the One and only it just has got to be wonderful...Sheep lead to the Oasis to drink...Result: Massive sales of the Bose cubes
 

·
Registered
Joined
·
246 Posts
The problem with amplifier ratings has been going on since I can remember in the mid 60's (yes that makes me old). I even remember at some point in the 70's where there was a law passed to require the ratings be stated in RMS power, at which bandwidth, and at what distortion. For about 5 years there was fairly good compliance with this law. But that was 35 years ago and the manufacturers have gradually gotten back to their old ways. The better brands will be quite accurate in their ratings (not always) but the lesser competition has gotten where they will cheat to try and make a sale.

The thing you are wanting with accurate amplifier ratings is the law already, but how are you going to enforce it ?

The biggest problem is with consumer grade gear. Most professional stuff can be trusted.
 

·
Registered
Joined
·
237 Posts
The problem with amplifier ratings has been going on since I can remember in the mid 60's (yes that makes me old). I even remember at some point in the 70's where there was a law passed to require the ratings be stated in RMS power, at which bandwidth, and at what distortion. For about 5 years there was fairly good compliance with this law. But that was 35 years ago and the manufacturers have gradually gotten back to their old ways. The better brands will be quite accurate in their ratings (not always) but the lesser competition has gotten where they will cheat to try and make a sale.

The thing you are wanting with accurate amplifier ratings is the law already, but how are you going to enforce it ?

The biggest problem is with consumer grade gear. Most professional stuff can be trusted.
I believe that that was not a law, but an FTC regulation. As it is no longer enforced, I believe that the regulation has been rescinded. All the 500+ watt HTIBs make a joke of power ratings.

In looking through my AV Receiver manual, I find that the FTC does have some rating specs, but they must be optional. The manual says:

Rated Output Power All Channels:
110 watts minimum continuous power per channel, 8 ohm loads, 2 channels driven from 20 Hz to 20 kHz, with a maximum total harmonic distortions or 0.08% (FTC)
130 watts minimum continuous power per channel, 6 ohms loads, 2 channels driven at 1 kHz, with a maximum total harmonic distortion of 0.1% (FTC)
120 watts minimum continuous power per channel, 8 ohm loads, 2 channels driven at 1 kHz, with a maximum total harmonic distortion of 0.7% (FTC)
7 ch x 170 W at 6 ohms, 1 kHz, 1 ch driven of 1% (IEC)

If I was the marketing manager, I'd advertise the 1190 watts IEC number, or better still, some bogus 'peak power' that would be more like 10,000 watts!
 

·
Premium Member
Joined
·
15,054 Posts
Receiver specifications are all over the board. Most manufacturers only rate the power output 2 channels driven not all 7 or 9. This is where receivers generally fail poorly usually only outputting half of what they are rated to do. I think this is very misleading to the un aware.
 

·
Registered
Joined
·
266 Posts
Discussion Starter · #11 ·
Sensitivity should be given in db/V/m because watts change with respect to frequency. 2.83v/m is the general standard as it translates to 1w @ 8 ohm and 2w @ 4 ohm.
OK, you are right, speaker are not a pure resistance

Also necessary is complex impedance (Z-chart) with phase angles.
Wow, a Z-chart, you have just lost 99% of the population (except on this site of course where peoples are very knowledgeable.

Finally, I feel any good speaker will have a clean waterfall graph.
I know what a Z-chart is because I studied electronic(it would probably take me a while to figure it out as I have been away from the field for quite some time), but I am not sure about the usefulness of waterfall in an anechoic chamber. I thought it was use more to check the acoustic in a particular room (don't flame me if I am totally wrong, I am not an expert in speaker)


ACD does help us see power supply robustness, but can also be irrelevant in the grand scheme of things. When do you drive all seven channels to full power?
Never, but this could be a good indication of the power supply and amp robustness.
 

·
Registered
Joined
·
1,747 Posts
Hi.

Wow, a Z-chart, you have just lost 99% of the population (except on this site of course where peoples are very knowledgeable.
Hi. You're right in that a Z-chart might not be immediately comprehensible (mostly because of phase angles) but it helps consumers make educated purchases. For example, you don't need to know a world about electronics or chart-reading to see this so-called 8 ohm nominal speaker:



Is not as easy a load as this so-called 8 ohm nominal speaker:



Of course, if you can analyze the effect of the phase angles, and recognize which frequencies most content you listen to/watch has its power demands centered on, you can get some idea that speaker A might need a more robust amp than speaker B assuming their sensitivity is the same. And if you can't, it doesn't hurt for the information to be there so that someone who can, can point it out and help that person out. :T



but I am not sure about the usefulness of waterfall in an anechoic chamber. I thought it was use more to check the acoustic in a particular room
waterfalls show what's going on as time elapse - the decay of the signal. Now a really long gated waterfall shows the decay of the room. But a short gated impulse response/water fall can still show a lot of information about problems with speakers. Floyd Toole does argue that they're "cosmetic" and I agree - good/mediocre speakers will have sufficiently good waterfalls - but poor speakers certainly won't.

Observe a speaker with little if any aberation:



and a $5000 audiophile speaker that isn't behaving pistonically



Now frequency response plots alone will normally tell your a lot of what you need to know, but I like CSD/waterfalls.

(don't flame me if I am totally wrong, I am not an expert in speaker)
No flaming here :T

Never, but this could be a good indication of the power supply and amp robustness.
could be, or it could be an indication of an amp with discrete power supplies to each of its amplifier channels - and thus actually lacking headroom in real world conditions dollar-for-dollar. I'm not against measuring ACD, but I'm against emphasizing it over 2CD into various loads. If a receiver can do about 60% into 7ch, 80% inch 5ch @ 1% distortion of its rated power, I think that's more than enough.
 

·
Registered
Joined
·
20 Posts
The problem with amplifier ratings has been going on since I can remember in the mid 60's (yes that makes me old). I even remember at some point in the 70's where there was a law passed to require the ratings be stated in RMS power, at which bandwidth, and at what distortion. For about 5 years there was fairly good compliance with this law. But that was 35 years ago and the manufacturers have gradually gotten back to their old ways. The better brands will be quite accurate in their ratings (not always) but the lesser competition has gotten where they will cheat to try and make a sale.

The thing you are wanting with accurate amplifier ratings is the law already, but how are you going to enforce it ?

The biggest problem is with consumer grade gear. Most professional stuff can be trusted.
I remember that so well. I used that info to buy my first integrated amp in 1975. I wish they had stuck to those standards. I think they should bring them back.
 

·
Registered
Joined
·
919 Posts

·
Registered
Joined
·
20 Posts
Waterfalls are not of any real utility to the vast majority of people. Rather, I believe they mislead because of people's tendency to believe they can hear with their eyes better than is actually possible.

There are many reasons. Example: the waterfall shows you something about the energy storage of a system, but it says nothing about the audible threshold as a function of level/time/frequency. So, how do you know which features in a waterfall are audible and which are not? Second, this is the response to an instantaneous hard gate of a signal with uniform energy across the entire spectrum. Very few pieces of music will contain anything like this (I've yet to find music that has a single area that looks like this). Therefore, even if it did represent what you can and cannot hear, it's unrealistic. Third, another poster mentions that most of this goes out the window in a real, non-anechoic environment. In most cases, the room will have much more reverberation and room energy will mask this the stuff you see in such a waterfall. Yes, if there is an obvious storage of energy at some narrow frequency and the decay is really slow it might be noticable on a hard gated sound, but we've already mentioned that this doesn't occur in the vast vast majority of music.

Radiation patterns: these are a little more useful if they're averaged over sufficient angle, but still the character of the room and speaker to listener transfer function is going to make a very big contribution to the sound. If you have these directional transfer functions, you can make some sort of prediction of the speaker to ear behavior. But, does that mean you know what it's going to sound like?

Finally, let's not forget that most of the music we listen to is stereo (meaning content coming out of both speakers). For this stuff, the interaction of the two speakers in the listening environment is far more important than almost any spec I've ever seen in any manufacturers literature or audio review (including John Atkinson's admirable tests). The ability for a stereo pair to convey a highly coherent sound to each of your ears is going to have a substantial effect on the quality of the phantom imaging (other than hard left/right, it's phantom guys). So, if you want a meaningful spec, let's get a "stereo" frequency response to each of your ears as a function of left-right channel differences. That means as a function of magnitude pan, as a function of delay pan, and as a function of time (early, mid, late). But, as mentioned by previous posters, that will leave 99.99% of the audio buying public in the dust.

So, are companies like Bose so evil to dispense with piles of specs which ultimately tell you nothing about how it will sound in your own home? Bose at least tells you to take it home and live with it for a while. If you don't like it you can take it back for a full refund. Ditto any really good high-end audio emporium. I've never bought a piece of audio gear (and I have some near nose bleed stuff) without a home audition period.... Or do you prefer those companies that pile on spec after meaningless spec in an attempt to convince you that you can hear those good numbers?

Audioengineer
 

·
Plain ole user
Joined
·
11,121 Posts
Your points are certainly valid, up to a point, but I don't see an alternative suggestion that justifies the design nor documents the performance of Bose or other products in your post. Is engineering really all about "take it home and try it?"

Audio performance is complex. Specifications and measurements have to be taken in context. Most of the discussion of waterfalls that you will see here relate to performance in a particular room that a user is trying to optimize.

I am not sure what your agenda is here but if you are attempting to justify Bose products by criticizing the use of specifications and measurements, I don't find your perspective very convincing. Most here who have opininions on Bose don't base them upon measurements but upon listening. I would welcome more objective assessments that would be meaningful. Do you have suggestions?
 

·
Premium Member
Joined
·
19,397 Posts
So, are companies like Bose so evil to dispense with piles of specs which ultimately tell you nothing about how it will sound in your own home?
No, Bose isn't evil, they are actually pretty smart with their marketing. There are no specs listed for the Acoustimass 10 Speaker System, it costs $1000, the Bose website says:

Our best 5.1-channel home theater speaker system for large rooms

Powered Acoustimass module adds drama to music and movies. Downward-firing drivers and proprietary technology deliver the lowest audible notes and effects with clarity and balance.

Ever heard one? I have. My sister bought one a month ago.Two 5.5 inch drivers that can't do 35 hz let alone the lowest audible notes and effects. A great system for the uneducated and uniformed. Let your ears be the judge.

 

·
Registered
Joined
·
20 Posts
(snip) Do you have suggestions?
You've made some excellent points. Engineering is NOT just about taking it home and listening to it. Engineers need to figure out how to measure the things that matter so they can produce better designs in an intelligent manner. FWIW, I'm not saying a purely empirical approach is not intelligent (e.g. trial and error). Unfortunately, most of the good engineering on the issue of why speakers sound the way they do will likely never make it into print because companies consider it IP. Practically everything Ffloyd and Sean (Olive) write about audio is valuable, if you understand the assumptions and caveats. A good place to start if you want to learn more on why traditional frequency response measurements and information are near useless and what you might do instead is Lipshitz and Vanderkooys 1985 paper on "Experiments in Direct/Reverberant Ratio Modification." One take away is that you can't design for a target frequency response and expect to know what it'll really sound like without taking radiation pattern into account. By extension, you also need the room acoustics.

This is just the tip of the iceberg though (nearly 30 years ago...).

GDSTUPAK: I agree with you that Tom's THX work was a real benefit to consumers because it says, "trust our judgement that systems which meet our requirements should be capable of providing some minimum level of audio reproduction quality." That gets the consumer away from meaningless or hard to understand specs. I don't think THX is enough to distinguish between good and really good, but it'll certainly separate the wheat from the chaff.

Mike P: I've tested a fair amount of audio gear in my life. I'm not saying your sister's Bose system could reproduce 35Hz, but it's not necessarily the system's fault. I remember an incident where a friend of mine bought a fairly capable home theater system and was showing it off to me. First thing I noticed is that his amp was clipping on bass heavy material yet it wasn't playing that loud. A simple reciprocal "walk-around" revealed he'd put is subwoofer in a spot where there was a intensely deep room null in the lower midbass range (I think it was something like 70Hz). His receivers autoEQ function had put a huge EQ peak there and it was clipping when trying to drive +15dB of power into that frequency. He needed a 2000watt amp rather than a 100watt amp. We moved the sub, to a better spot and now his system can shake the house across the whole bass range. Someone might have said, brand X subs suck big time, but that is not necessarily so.

And ... I don't think you are being rational to suggest that low frequency response is limited due to the fact that there are only two 5.5" drivers in that box. It really depends on the box, and specifics of the driver design. Bose uses ported boxes and waveguides. If the port or waveguide is tuned at 35Hz, a 2" woofer with the right performance characteristics could drive that port/guide to produce a fair amount of output at 35Hz. This is a case of not judging a book by the cover. We don't hear with our eyes. They just bias our opinions. I've been thru a Bose store demo, and in that setup their system had plenty of low frequency output. Nothing close to my main system at home (The Bose felt like it was around 15dB lower in maximum output), but it did play fairly deep.

In summary: if a sub sucks, make sure it's not the room or sub placement. If you think there's no 35Hz output, it's easily confirmed with a slow sine sweep on a CD/DVD and a cheap SPL meter.
 
1 - 20 of 64 Posts
Top