JBL LSR 308 in the house

[Thomas savage]

Unfortunately, very few of my music recordings contain pink noise as tracks, or other pure noise as part of their creative design. Therefore, listening to such types of test tracks tells me very little about how well music is played back - strangely, listening to music replay tells me a great deal about how well the system is at, well, music replay. Very unscientific I realise, but hey, I'm that sort of guy! But if you know about some fabulous non-spatial recordings that will keep me entertained for hours, I'll be pleased to hear about them ...

I use pink noise, it's useful as Ray points out..indeed I use a few speaker setup cd's they contain content that it known and standard..

This eliminates chasing ones tail and deluding ones self... Well possibly not the last one

 

[fas42]

Yes, I do have a a couple of test CDs - with pink noise, too! But the main use for them has been frequency sweeps, conditioning speakers by driving them with sine waves, and listening for digital artifacts in music attenuated by specific amounts. Everyone will have their own methods - I have found that certain styles and qualities of recordings of music make it trivially easy to hear problems; they're my diagnostic tools, .

Deluding oneself comes about when you try to decide whether something sounds "better" than something else - my audio friend got himself into strife over and over again, using this poor method! Far more effective is to stress test the playback, until there is an obvious flaw - which you then proceed to resolve ... it's the same process as for cars, you abuse some part, like a suspension assembly, until it breaks - you then know the true limits, rather than pretending you do ...

 

[RayDunzl]

I, even with my deaf leaden ears, I keep sensing something fundamentally different between the panels (ML) and the cones and domes (JBL).

Let's use the word fizzy as a JBL critical listening impression.

So, I tuned them both up with some DRC and level matching.

The 1/6 octave smoothed two speaker frequency response is here (red ML , blue JBL)

Now for some exciting detail you (and even I) may not usually look at even if you look at this stuff.

Here is the unsmoothed response, zoomed in to the range of 3kHz to 8kHz.

How can that be a good thing?

 

[fas42]

So, we're saying that some frequencies almost completely disappear acoustically with the JBLs - diving into a valley 40db below the average level?

 

[RayDunzl]

So, we're saying that some frequencies almost completely disappear acoustically with the JBLs - diving into a valley 40db below the average level?

That's what the microphone sees in-room.

It's not specifically the JBL. Here is the left Infinity P-363 (green) vs the left ML (blue) from 18 months ago:

Similar result, maybe to a slightly lesser degree. Green has more instances of cancellation, anyway.

I surmise it is the wide dispersion speaker interfering with its own reflections from my un-deadened walls and ceiling vs the less dispersive dipole.

Taking one of the JBL outside and attempting a somewhat anechoic test would tell something more.

 

[amirm]

Here is the unsmoothed response, zoomed in to the range of 3kHz to 8kHz.

View attachment 4096

How can that be a good thing?

I don't know about a good thing but those troughs are smaller than the resolution of your ear at those frequencies. See: http://audiosciencereview.com/forum/index.php?threads/perceptual-effects-of-room-reflections.13/

 

[DonH56]

Looks like normal interference effects from multiple drivers, baffles, room/boundary reflections, etc.

 

[fas42]

So, what's the point of measuring in-room response of a speaker, then - what's that telling one that's useful, in terms of "how good it is" ?

 

[RayDunzl]

I don't know about a good thing but those troughs are smaller than the resolution of your ear at those frequencies.

I think I understand that analysis. You can't "hear" narrowband defects.

Not one, not two. How about 10? 100? 1000?

Here's an unsmoothed "mic a few inches in front of the JBL" (blue) vs ML (red) at the listening position.

And JBL vs itself at the listening position (red is close mic):

I hear "something". Kalman's description of the narrow/wide difference is remarkably similar to my own here.

What are we hearing if we can't hear that?

 

[fas42]

BTW, you used the word "fizzy" above - I'm more prosaic, I find the term "distortion" more relevant. If the system makes me aware of a quality which is not part of the recording then I'm hearing distortion - fiddling with FR, etc, will not make this go away. There will be a cause for that "issue", and the interesting thing then becomes working out what that cause might be, and solving it.

 

[RayDunzl]

So, what's the point of measuring in-room response of a speaker, then - what's that telling one that's useful, in terms of "how good it is" ?

It's fun.

There will be a cause for that "issue", and the interesting thing then becomes working out what that cause might be, and solving it.

This is how I work.

You other posts describe how you work.

 

[fas42]

I hear "something". Kalman's description of the narrow/wide difference is remarkably similar to my own here.

What are we hearing if we can't hear that?

Isn't the ear/brain amazing? It unscrambles that mess of a transition between extremely direct sound, to messed up sound. Good thing it can, otherwise we would have all died very quickly in the jungle ... but then people turn around and obsess about the minutest subtle variations in how the sound is directed by the drivers into the room. I don't know about you, but I find there is something pretty nutty about the thinking here ...

 

[fas42]

It's fun.

Can't beat that one !!!

 

[amirm]

I hear "something". Kalman's description of the narrow/wide difference is remarkably similar to my own here.

What are we hearing if we can't hear that?

With two ears you hear a much different thing than the single microphone measures. Acoustic measurements in higher frequencies simply is not revealing. You can filter them to 1/6 octave to get a sense of overall tonal curve but still does not tell you how the speaker sounds in the room. You would need measurements of multiple angles with right weighted average to predict the likability of the speaker.

 

[fas42]

This is how I work.

You other posts describe how you work.

One of the aspects of how I work is the listening "a few inches in front of the JBL" thing. Then, I'm getting the 100% undiluted output of the driver. With many systems this is not very pretty, the level of sewage being added to the mix by the playback flaws can be pretty nauseous - however, you now do have the 'truth', and so, what are you gonna do about it ... ?

People might be surprised to know that hearing this way can be a 100% clean experience - in fact, this is what gives one the 100% "disappearing speakers" ...

 

[RayDunzl]

With two ears you hear a much different thing than the single microphone measures.

Acoustic measurements in higher frequencies simply is not revealing.

You can filter them to 1/6 octave to get a sense of overall tonal curve but still does not tell you how the speaker sounds in the room.

You would need measurements of multiple angles with right weighted average to predict the likability of the speaker.

Ears are not microphones.

Something is revealing. Maybe it is related to what can be measured at a single point.

1/6 octave smoothing for the 3kHz to 8kHz region:



You're right, that does not tell me how the speaker sounds in the room.

I'm not predicting the likability of the speaker.

 

[Blumlein 88]

I, even with my deaf leaden ears, I keep sensing something fundamentally different between the panels (ML) and the cones and domes (JBL).

Let's use the word fizzy as a JBL critical listening impression.

So, I tuned them both up with some DRC and level matching.

The 1/6 octave smoothed two speaker frequency response is here (red ML , blue JBL)

View attachment 4097

Now for some exciting detail you (and even I) may not usually look at even if you look at this stuff.

Here is the unsmoothed response, zoomed in to the range of 3kHz to 8kHz.

View attachment 4096

How can that be a good thing?

I would say that isn't a good thing. The spacing is almost like a comb filter in 100 hz steps. I wonder where that would be coming from? Do you get a result like that Ray if you do such a test say 1 meter in front of the tweeter of the JBL?(yes I am sort of asking you run such test)

EDIT to ADD: Okay I see you already have done what I mentioned above. And the speaker isn't putting out a signal with all those dips in it. So it likely is what you surmise especially at 3khz and above your panels are interacting much less with the room, and you are in something closer to a nearfield situation with the panels and very much far field with the more dispersive and closer to point source 308s.

Maybe that is why the 305s I have sound even better when I monitor recordings with them from 1 meter distance compared to sound in a larger room from 11 feet away. That sort of quiet clarity from my Soundlabs is what the 305s (assisted by a sub in the large room) are missing. You get some of the clarity back in nearfield monitoring. While you don't get a benefit if you listen to Soundlabs up really close.

 

[Blumlein 88]

You could test this to see if that is causing the fizzy sound. Take a music file let us say reduce a copy by 6 db. Then move it to represent a 100 hz comb filter (delay it by .01 seconds) and mix the two together. Listen over your ML speakers and see if they sound fizzy like the 308s. What do you think?

 

[fas42]

I was curious what people would call, "fizzy" - and found one simple description, "pronounced upper mids and treble". That would be a subjective impression, and is very much an interference distortion anomaly - I would suspect the class D amps inside the box need a bit more taming, refining - a little too much very high frequency noise is getting to where it shouldn't ...

 

[Blumlein 88]

I was curious what people would call, "fizzy" - and found one simple description, "pronounced upper mids and treble". That would be a subjective impression, and is very much an interference distortion anomaly - I would suspect the class D amps inside the box need a bit more taming, refining - a little too much very high frequency noise is getting to where it shouldn't ...

I don't think the class D amps are doing that. The effect went away when Ray measured up close to the speaker. Didn't Ray also show a distortion plot of both speakers, and while the panels were lower, the 308s didn't have high distortion levels other than in the low frequencies as I recall?