Research Project: Infinity IL10 Speaker Review & Measurements

[amirm]

Call me crazy and a newbie, but I wonder: why would one do a listening test on a speaker without applying some equalization first? Particularly should one consider room EQ a transformative thing.

If the room never changes, and nor does the position of the speaker or the listener, then EQ for the room becomes immaterial. The effect of the room becomes constant. Indeed this is the way the research was conducted.

However, I did find a flaw in performing the test without EQ when one is testing speakers with wildly varying bass output. The ones with more bass can hit the room modes and create ringing/boominess that lowers the subjective performance of that speaker, but not another with less bass. For this reason, all of my listening tests use a parametric EQ around 100 Hz to take down the most dominant room mode. Once there, it equalized the playing field, pun intended. You can see this often when I show the correction EQ for a speaker:

That first one in teal color is for room mode.

Of course if you are trying to optimize a single speaker for your room, you would do more.

 

[amirm]

Sorry amir.

They always want more!

Imma just gunna say thanks for everything your doing. I hope you know how much we appreciate it! Dead serious, If your ever in Bath, UK, I would take the day off to show you everything. Beers on me.

Thanks. But how do we keep @Thomas savage out of it??? He is likely to find out and want to tag along. Which would be fine by itself but no self-respecting pub owner would want to let him in their place!!!

 

[booster]

Yes, I have. Here's how it looks with my room in the 500-3000Hz range.

I also encourage you to do more such checkings yourself before making various conspiracy theories about what is audible and what is not. Just take mic in your hand, start REW and see things for yourself. Luckilly for all of us it's very easy and cheap to do such things these days.

View attachment 71645

bs22 eq filter, can I get it?

 

[tuga]

All well and good quoting the great Bruno... But...do you understand it?

True of false: if the THD of the speaker is max -70db, IMD is also low ( inaudible)

TotalHD is rather useless because HD varies in level along the operating range.

 

[thewas]

And also odd order harmonics are psychoacoustically much more severe and have a quite higher threshold of detectability than even order ones.

 

[Absolute]

A small look into IMD in speakers;


"The graphic from Figure 10 shows the spectrum of the excitation signal (green curve), the measured spectrum of the signal emitted by the loudspeaker (red) and the distortion components extracted from it (blue). Both are summed up in 1/6 octave bandwidth. "
https://www.fidelity-online.de/bowers-wilkins-800-d3-messungen/

The interesting things to look for here is the blue noise in the bottom of the graphs (IMD not in the signal, aka the inclusion of new sounds/distortion from the speaker) relative to the output level provided. This is quite complicated since it's not the same between speakers, but the spl level is stated in the text above each graph.

The JBL M2 is measured at 106 dB peak at 8m while the Kef Ref 1 is measured at 101 db peak at 4 m, which is a massive difference.

Here three small active monitors are compared directly, but shown in a easy-to-read manner;

https://www.fidelity-online.de/kompakte-studiomonitore-von-nubert-adam-und-sls/

Unfortunately there's not many online measurements from Fidelity to get a broader view of what is good and what is not. If anyone has a good understanding of these things, please share with my stupid self

 

[GelbeMusik]

That looks nice, now You're talking! Alas, the signal is not representative for a musical signal for instance. But it is the most similar used today.

Let's see if anybody else has an insight to what is going on here.

 

[Lbstyling]

Thanks. But how do we keep @Thomas savage out of it??? He is likely to find out and want to tag along. Which would be fine by itself but no self-respecting pub owner would want to let him in their place!!!

The advantage of being a local with a traditional pub.....the 'lock in' mate.

https://www.google.com/amp/s/www.thrillist.com/v/s/www.thrillist.com/amphtml/drink/nation/english-pubs-vs-us-bars-15-reasons-why-british-pubs-are-better-than-american-bars?amp_js_v=0.1&usqp=mq331AQFKAGwASA%3D#ampf=



I think you really may need a proper British education in such things!

Thomas is welcome, but.... Taunton.... hmm. Been a while since I have done any charity work.

he should be able to handle some cheddar valley though, so the locals might let that one slide.

 

[Lbstyling]

TotalHD is rather useless because HD varies in level along the operating range.

U might want to take a look at my comment again.

 

[Lbstyling]

Well, try to find a loudspeaker with max -70 dB THD (=0,03% !!!) in the lower frequency range, I would buy a truckload of them!

Mine does. TD15m

 

[thewas]

Mine does. TD15m

Measurements or it didn't happen

 

[amirm]

I think the measurement differences are already pretty evident from switching between all the various charts for the two speakers created by you and everyone else on here to be honest. The IL10 actually scores 5.17 according to @MZKM ’s charts by the way (looks like his score post at the beginning of this thread needs correcting). I would say this is a high, rather than very high score (it’s in the upper half of the third tier of six in his rankings). The M16 got a score of 5.54, is ranked 10th among the speakers tested so far, whereas the IL10 comes in at 18th, not an insignificant difference.

You are mixing our work with the research. The papers do NOT give a computed score for IL10. They give a subjective mean of 6.11:

Its timbral scores sans amount of bass were almost perfect:

This matched my subjective listening tests with respect to tonality:

So if this is all that matters, then it doesn't get better than this in the views of the listeners in controlled testing conducted by Sean. This speaker should knock the socks out of anyone listening to it who hangs around here. You and others should hunt it down and use it instead of whatever you have. Why do I not see a surge of people going to buy it? It only costs $40 I think, sans shipping.

Now, I did present that maybe this is down to 1 dB difference in frequency response mattering. If so, we are once again past the bounds of the research as the modelling is not remotely that accurate.

Then again if we say we can get more perfect than perfect, then you are in my house with respect to the research not reading on comparison of well designed speakers which is most of what we are dealing with. As such constantly quoting the research is of little value.

 

[Lbstyling]

Measurements or it didn't happen

Here's Brandon's proper multitone measurements of the TD 12m (at 100db no less.!)

https://5e8772ee-a-62cb3a1a-s-sites.googlegroups.com/site/drivervault/driver-measurements/tang-band-75-1558se/ae-speakers-td12m/non-linear-distortion/multitone_500hz_100dB.PNG?attachauth=ANoY7cqIe3iOtA-StKIKKa28C9ane4klUoGOzunFTZCHQpLBsJ1fckHq9j2-taxIpgp6YLYvAl1rUGtnqAd_3oGr_2mclyq72QYFDyYg_YIcaFOdsPV7lIgBoqGanuh3E15RibtbDZswMdmwRKe9T5-cQ0fg-wq1Kf9RCsM6lTewBMQVMPXENgZIkEtcf4Jsg3lbTkXGkEz6GWH63-XcMq-wsW0qWM_U3p6ofbPP_vsJTdV7N9pN6WmYuiLkayCsjrtlxLugkvj-VugJZvcrGOFbeFS171fB8lVjuIphNh23T2VDQrTwQBiU4BLvv7DyJHWUkc7N7sncIv5Baw00o-WwwshUMru8Bg==&attredirects=0

The TD 15m measures significantly better than this.

The 15m has an identical motor, but larger (and profiled for better off axis) cone.
It's considered by many to be the best mid/bass driver currently available.
I don't have my old PC with the measurements on to show you the 15m unfortunately.

On the top end I have a JBL 2452SL on a 350hz horn. Also around -70db distortion across the pass band.

 

[amirm]

The IL10 actually scores 5.17 according to @MZKM ’s charts by the way (looks like his score post at the beginning of this thread needs correcting). I would say this is a high, rather than very high score (it’s in the upper half of the third tier of six in his rankings). The M16 got a score of 5.54, is ranked 10th among the speakers tested so far, whereas the IL10 comes in at 18th, not an insignificant difference.

Another thing: I am not at all in favor of using two decimal places to predict listener preference subjectively to a speaker. The mere computation of mean in the original paper indicates we better allow for variations. It was a mistake as such to provide two decimal places in the research, and in our models. I know I had trouble assigning even whole numbers when I took the test let alone two decimal places. We are not instruments with that kind of accuracy.

 

[thewas]

Here's Brandon's proper multitone measurements of the TD 12m (at 100db no less.!)

https://5e8772ee-a-62cb3a1a-s-sites.googlegroups.com/site/drivervault/driver-measurements/tang-band-75-1558se/ae-speakers-td12m/non-linear-distortion/multitone_500hz_100dB.PNG?attachauth=ANoY7cqIe3iOtA-StKIKKa28C9ane4klUoGOzunFTZCHQpLBsJ1fckHq9j2-taxIpgp6YLYvAl1rUGtnqAd_3oGr_2mclyq72QYFDyYg_YIcaFOdsPV7lIgBoqGanuh3E15RibtbDZswMdmwRKe9T5-cQ0fg-wq1Kf9RCsM6lTewBMQVMPXENgZIkEtcf4Jsg3lbTkXGkEz6GWH63-XcMq-wsW0qWM_U3p6ofbPP_vsJTdV7N9pN6WmYuiLkayCsjrtlxLugkvj-VugJZvcrGOFbeFS171fB8lVjuIphNh23T2VDQrTwQBiU4BLvv7DyJHWUkc7N7sncIv5Baw00o-WwwshUMru8Bg==&attredirects=0

The TD 15m measures significantly better than this.

The 15m has an identical motor, but larger (and profiled for better off axis) cone.
It's considered by many to be the best mid/bass driver currently available.
I don't have my old PC with the measurements on to show you the 15m unfortunately.

Thats a multitone measurement around 500 Hz (!) where membrane amplitudes are several orders smaller, I am talking about bass region and (T)HD.
Even the fantastic 15" driver of the JBL M2 had at a 91 dB Leq at 8 meters (106 db peak Lpk) total distortions (IMD & HD) of 3,4%, anything unter 0,3% especially in the low bass is utopian.
https://www.audiopro.de/medias/item/16109/jbl_m2_soundandrecording_1214.pdf

 

[Lbstyling]

Thats a multitone measurement around 500 Hz (!) where membrane amplitudes are several orders smaller, I am talking about bass region and (T)HD.

It's for a smaller driver, (half the SD) at 100db output.

https://5e8772ee-a-62cb3a1a-s-sites...XerWgSrWs9l00HaiGUanGyTkx16LMx&attredirects=0

This is the 12m again at 100db sweep.

I had the 15M measured at -70 THD @ 9X (cannot remember the exact SPL).

I'm sure you can extrapolate the driver would be close to this from this measurement.

And I agree, proof would be nice for you, but am not sure I can be bothered to setup all my kit for you as I'm busy building an extension on my house for them to live inTBH!

 

[tuga]

This matched my subjective listening tests with respect to tonality:

So if this is all that matters, then it doesn't get better than this in the views of the listeners in controlled testing conducted by Sean.

It would be a bit disturbing (to put it politely) if we were to accept that frequency response / tonal balance is all that matters.
I wonder whether this the result of cutting too many corners when designing the research tests... Could the listeners have been that undemanding? Hard of hearing? Drunk?

 

[MZKM]

The IL10 actually scores 5.17 according to @MZKM ’s charts by the way (looks like his score post at the beginning of this thread needs correcting)

Fixed (put 3 instead of 2 by accident).

 

[thewas]

It's for a smaller driver, (half the SD) at 100db output.
https://5e8772ee-a-62cb3a1a-s-sites...XerWgSrWs9l00HaiGUanGyTkx16LMx&attredirects=0
This is the 12m again at 100db sweep.

And it shows that it has in the lower bass region THD around -30dB / 3% which is a good and normal value and endless far away from your -70dB / 0.03% limit...

I had the 15M measured at -70 THD @ 9X (cannot remember the exact SPL).
I'm sure you can extrapolate the driver would be close to this from this measurement.
And I agree, proof would be nice for you, but am not sure I can be bothered to setup all my kit for you as I'm busy building an extension on my house for them to live inTBH!

Sorry but I don't trust such measurements and you don't need to measure them for me, just post any official measurements of any loudspeaker or driver which has those -70dB / 0.03% or lower THD in the lower bass region at a reasonable SPL.

 

[amirm]

It would be a bit disturbing (to put it politely) if we were to accept that frequency response / tonal balance is all that matters.
I wonder whether this the result of cutting too many corners when designing the research tests... Could the listeners have been that undemanding? Hard of hearing? Drunk?

The motivation for the test was to counter the damaging effects of Consumer Reports review by the way which put Harman speakers in bad light. Once that was accomplished, maybe the motivation was not there to keep going as the research stipulated they would in areas of distortion and directivity.

The tests involved those speakers in CR review so all they had to do was compare those. I am wondering what the error bars would be if they were given the speakers we are dealing with now.