How should music sound?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
"2. CRITERIA OF LOUDSPEAKER PERFORMANCE

2.1. Terms of Reference

It is assumed that the ideal to be aimed at in the design of a sound reproducing system is realism, i.e. that the listener should be able to imagine himself to be in the presence of the original source of sound. There is, of course, scope for legitimate experiment in the processing of the reproduced signals in an endeavor to improve on nature, however, realism, or as near an approach to it as may be possible, ought surely to be regarded as the normal condition and avoidable departures from this state, while justified upon occasion, should not be allowed to become a permanent feature of the system.."

Title: THE DEVELOPMENT OF HIGH-QUALITY MONITORING LOUDSPEAKERS: A REVIEW OF PROGRESS

Source: BBC (1958)


I agree !
 
Note that the BBC dictum was scribed in 1958. I agree it's a worthy goal for mono recordings. ;)

Ah, but *stereo* was a means at the time (and today), to achieve greater realism. (..at least when compared to most mono recordings.)

The goals expressed in 1958 shouldn't be any different than today. ;) Note the section heading: "Terms of Reference"
 
Last edited:
Are you suggesting the Beatles shouldn't have done Sgt. Pepper's? :D

You can love the Beatles and acknowledge that their recordings aren't the ultimate in fidelity. Too much doubletracking and studio manipulation.

My desires for a sound system: when the recordings were made with care then I want acoustical instruments to sound as much like the live instrument and venue as possible. For studio recordings and electronic instruments I want great involvement, texture, variety of tone color.

The biggest sin of a system, IMHO, is to impart a strong personality of its own to what it reproduces. I want to hear each recording without an added layer or an everpresent filter.

David
 
Administrator
Joined 2004
Paid Member
Since you are holding the answer you might tell how is this curve measured in a room ? :)
With a microphone! :D

I use a wide band cardioid mic because its directivity is similar to my ears, but an omni will do. I measure at the listening position.

There should be plenty of the B&K curve around the web. I'll try to find one.
EDIT: Attached is the classic curve. I start my roll-off higher, circa 400 Hz.
 

Attachments

  • Bruel-Kjaer-optimum-curve.png
    Bruel-Kjaer-optimum-curve.png
    76.9 KB · Views: 161
Let's consider the following example.

The source material is the Sheffield Drum and Track CD, a good recording of various drum solos IMHO.

The most realistic sound to me for this CD comes from headphones, high directivity horn loaded 3 way speakers, or small 2 way (1" tweeter, 5.25" woofer) speakers auditioned in the nearfield. Environment is a typical untreated living room , about 14 x 20 x 8'.

If I hear it in the farfield through the 2 way speakers above, the transients are "diluted" by the reflections. They do not sound as realistic to me. Have you even listened to music with too much reverb added? The detail or texture decreases as one adds more reverb. This dilution I referred to above is similar but to a lesser degree.

Now if I listen to classical music in the farfield on that 2 way speaker, it sounds more realistic that through headphones/horns/nearfield to me.

What speaker configuration offers the best balance?

David, even if the speakers are neutral, doesn't the room add character with the reflected sound? Not unlike concert halls?

Pano, what is the reference measurement distance for that target curve? What size of room? thanks for graph!
 
Administrator
Joined 2004
Paid Member
I dunno, wish I did. It's a famous curve, so we should be able to find more info. It may need a little tweaking for your room, but it works well in most domestic environments.
The curve is definitely not the 1M response, it's the listening position response.
 
With a microphone! :D

I use a wide band cardioid mic because its directivity is similar to my ears, but an omni will do. I measure at the listening position.

There should be plenty of the B&K curve around the web. I'll try to find one.
EDIT: Attached is the classic curve. I start my roll-off higher, circa 400 Hz.

Is it impulse response measurement ? Windowing, smoothing etc ?

Or averaged bandpass 1/3 noise amplitude response ?
 
... If I hear it in the farfield through the 2 way speakers above, the transients are "diluted" by the reflections. They do not sound as realistic to me. Have you even listened to music with too much reverb added? The detail or texture decreases as one adds more reverb. This dilution I referred to above is similar but to a lesser degree....

I very much agree! Big problem in sound recording and reproduction. With no reverb it sounds dead on some systems/rooms, with a little bit of reverb it can sound horrible on other systems/rooms. Tricky stuff.

Edit:
4! With a punch that will flatten noses en masse! With the purity of newborn childrens minds! With a velvet background that makes everything disappear!
 
Last edited:
Administrator
Joined 2004
Paid Member
Or averaged bandpass 1/3 noise amplitude response ?
Oh, I see what you mean! Usually 1/3 or 1/6th averaged response as measured by HOLMImpulse for me. Don't know what B&K used. But as long as my overall response meets that falling target, it sounds balanced to my ear. Most visitors comment on how well balanced my system is, so it may be working.
 

ra7

Member
Joined 2009
Paid Member
Pano, there are a number of combinations of direct response and power response and room that will give you that curve. Are you saying that in your experience, the curve works in most rooms, with most speakers, regardless of differences in directivity? Or are you saying that in your combination of speakers (which are by now well known on diyaudio :)) and room, the curve works?

I do agree that the general trend in that curve sounds right.
 
Administrator
Joined 2004
Paid Member
I haven't tried it on all rooms with all speakers, so can't say. But in my experience, and the experience of others I know, this tends to work well in a domestic listening environment. If you have a computer music source, it's easy to try.

Direct vs reflected is one reason I use a cardioid mic. The omni picks up more from behind me than my ears do. Not so important outside or in a very large room, rather important in a small room. Perceived tonal balance is dependent on direct and reflected sound together. In what ratio, I do not know. I do think the curve need a little tweaking depending on the room and speaker. But just a little.
 
If I hear it in the farfield through the 2 way speakers above, the transients are "diluted" by the reflections. They do not sound as realistic to me. Have you even listened to music with too much reverb added? The detail or texture decreases as one adds more reverb. This dilution I referred to above is similar but to a lesser degree.
Personally I find it the other way around, the texture "improves", becomes more interesting with greater echo. This greater detail in the music is a 2 edged sword, if the system adds beyond a certain level of audible distortion to the sound then such recordings become difficult or impossible to listen to, the ear/brain overloads trying to sort it all out. But if the system is sufficiently clean then these recordings become the most magical of all, have the greatest impact.

This is why I use "difficult" recordings to test systems, not super simple and clean ones. If you're buying a new car do you test its capability by finding the smoothest, straightest road to drive on ...?

Frank
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.