JBL Synthesis: 1400 Array - measurements interpretation :)

[youngho]

Basically, Stereophile has the measurement accumen of an inexperienced hobbyist. The fact that JBL's measurements show a flat response, and Stereophile shows something like a 7db hump in the bass is a classic example of this.

I do not believe Stereophile has done anything to improve their measurement set-up. I believe Audioholics has a decent anechoic chamber they use for measurements, and two online reviewers who don't even run ads (ASR and Erin's Audio Corner) now have Klippel NFS systems which are state of the art acoustic measurement platforms that enable better accuracy than the world's best anechoic chambers.

That Stereophile hasn't seen fit to buy an NFS shows you how seriously they take their data presentations.

Stereophile has been doing measurements for decades, so maintaining a relatively consistent method allows for more of an apples-to-apples comparison when looking at their website, in my opinion.

Young-Ho

 

[617]

Stereophile has been doing measurements for decades, so maintaining a relatively consistent method allows for more of an apples-to-apples comparison when looking at their website, in my opinion.

Young-Ho

It's not consistent. When you take a nearfield measurement of a woofer you are typically only 1cm away from the cone. At these distances, small changes in mic position change the level of the measurement dramatically - going to 2cm makes it 6db quieter, going to 5mm makes it 6db louder.

You then need to splice the nearfield measurement to the gated measurement. How do you do this? Ideally you'd do a calculation on baffle step frequency, adjust the nearfield response accordingly, and then make an educated guess as to the bass level compared to the rest of the speaker. Ideally instead of splicing at a fixed frequency you blend the responses together.

Stereophile clearly doesn't do this. The splice it in any which way and they get it totally wrong.

It's laughably inaccurate. None of their measurements portray the overall tonality of a speaker correctly.

 

[pos]

The full engineering data sheet, including the spinorma, can be found here:

1400 Array

... Submitted by Giskard

www.audioheritage.org

 

[changer]

I support youngho’s case that Atkins must have made a measurement error if the other shown measurement ought to be correct, and disagree with both Duke and 617. If done correctly, and this is mostly a question whether the gated measurements right window is long enough to ensure correct alignment, spliced near field and far field measurements can be aligned correctly. Have a look at Napilopez’ work, his measurements are often very close to the Klippel results and never radically different. Detail ist not lost in the bass frequencies, they are measured near field and hence with high resolution. It is the lower frequencies of the far field measurement, between roughly 300 to 1000 Hz,which do not have a high resolution. Near field measurements do introduce artifacts, such as notches at a multiple of the radiators diameter, which is why they are only used to a calculated maximum frequency per speaker type, but a hump at 100 Hz as a result of near field measurement is highly doubtful. Either Atkins made a mistake or the other measurements are wrong, both cannot coexist.

E: types this in parallel with 617 and I support his last point that Stereophile probably did not match the measurement correctly.

 

[youngho]

Near field measurements do introduce artifacts, such as notches at a multiple of the radiators diameter, which is why they are only used to a calculated maximum frequency per speaker type, but a hump at 100 Hz as a result of near field measurement is highly doubtful.

All of Stereophile's measurements have this hump, and all the texts accompanying the measurements note something like "Around half of the apparent boost in the bass region is an artifact of the nearfield measurement technique" for the Array 1400 or "At lower frequencies, the broad rise in output in the upper bass will be mainly due to the nearfield measurement technique" for the Salon 2 (https://www.stereophile.com/content/revel-ultima-salon2-loudspeaker-measurements). Knowing that this is "relatively consistent" across all of Stereophile's measurements, you can still see that the Salon 2 would be expected to have deeper bass extension than the Array 1400 and that the Array would have more upper bass, even without seeing Harman's spinoramas. That's what I meant about apples-to-apples.

Also, knowing that they made a mistake (possibly understandably so, since it could be argued that this is part of the "relatively consistent" method that I mentioned before) in the choice of axis, one can basically hinge down the response above 5 kHz or so and come up with ballpark-similar measurements above 2 kHz or so compared with Harman's, at least until 10 kHz or so, but the range several hundred above and below the crossover frequency between the woofer and midrange horn is likely adversely affected by the choice of axis.

Young-Ho

 

[changer]

I never understood why they would end up with a boost in bass region, if working correctly. They do not seem to care if they put it in the disclaimer ..