Spinorama! (also known as CTA/CEA 2034 but that sounds dull, apparently)

[QMuse]

@amirm Per the discussion in this thread and how you said the Klippel calculated the ER curves, I decided to check for myself and use the klippels own data to calculate the ER curves manually using the process described by Todd Welti above. I had noticed that the klippel-generated ER and ERDI curves looked usually a fair bit smoother than the Harman ones around the crossover, and this shows why.

Here is the M16 ER and ERDI curves as generated by Klippel (white) vs the process used by HATS(blue) mentioned by Todd Welti above:

View attachment 56561

Not too much of a difference, but you can see the notch/bumps in the blue curves are worse and match Harman's a little more closely than the klippels.
View attachment 56562

Now on the Revel F35:


View attachment 56566

A more prominent crossover issue, and once again, closer to the HATS result:View attachment 56567
The bump on the ERDI curves are now nearly identical.

While probably not something worth going back to fix, it's something to perhaps keep in mind with measurements going forward, and that I hope klippel is informed of. Once again, it could be the difference between a good and great DI curve, especially on a speaker with particularly egregious vertical problems. And it would be good for the CTA-2034A to process to actually be used as, you know, a standard

As an interesting contrast, while almost every speaker I've tested gets worse with the HATS process, the Harbeth actually seems to get better. It has less of a tilt in the ER and smoother directivity. The vertical and horizontal dips actually do a better job of balancing out with this interpretation.
View attachment 56563

But of course Harbeth was the one that got better. It's only purpose is to serve as a universal exception for all rules!

 

[napilopez]

But of course Harbeth was the one that got better. It's only purpose is to serve as a universal exception for all rules!

Haha I was thinking the same. It does seem like the focus was on balancing various nulls for a smoother in room response.

 

[thewas]

Some more Harman spinoramas (also from other brands) added https://speakerdata2034.blogspot.com/2020/04/various-brands-spinorama-data-from.html?view=classic

 

[tuga]

Some more Harman spinoramas (also from other brands) added https://speakerdata2034.blogspot.com/2020/04/various-brands-spinorama-data-from.html?view=classic

803D3 fares very well in terms of ER- and SP-DI.

I wish that these plots include Estimated In-Room response...

 

[thewas]

803D3 fares very well in terms of ER- and SP-DI.

Yes, the D3 generation is a positive step back to the good directivity loudspeakers they had till they 2000s (800 Matrix and N Series).
Even their 90s entry 600 series measured seriously well https://www.stereophile.com/content/bw-dm603-s3-loudspeaker-measurements

 

[napilopez]

803D3 fares very well in terms of ER- and SP-DI.

I wish that these plots include Estimated In-Room response...

A shortcut is to just look at the Early Reflections curve with a little crick in your neck. The estimated in-room response is almost always nearly identical to the early reflections curve in shape - it just tilts down a dB or two more. In his book Dr Toole also often used the ER as an analog for the PIR.

 

[hardisj]

Have you guys noticed this in the ANSI/CTA-2034-A R-2020 spec?

4.1 General Test Setup and Conditions
.
.
.

If a small loudspeaker (satellite, bookshelf, etc.) without a manufacturer-specified filter network is being tested then a two pole Butterworth 80 Hz high pass filter network shall be used.

I don't plan on implementing this, personally. But, I found it interesting.

 

[edechamps]

Have you guys noticed this in the ANSI/CTA-2034-A R-2020 spec?

That's… really weird, and I'm having trouble making sense of such a recommendation. I would certainly like to see an explanation on this. Looks like no-one follows that guidance and intuitively I'd think that's a good thing! Maybe add that to the pile of stuff that should be revisited/clarified in CTA-2034…

 

[JohnPM]

I would certainly like to see an explanation on this.

Looks like an emulation of the usual bass limited speaker management in a receiver, 80 Hz being the default bass management crossover if the manufacturer hasn't specified something different.

 

[Juhazi]

These CTA ER curves are anyway just calculated sums of some x/y-axis measurements. Quite irrelevant which is "right" , but for comparisons sake we must stick to a standard way of measuring/calculating - differencies between "good" speakers are small.

 

[napilopez]

These CTA ER curves are anyway just calculated sums of some x/y-axis measurements. Quite irrelevant which is "right" , but for comparisons sake we must stick to a standard way of measuring/calculating - differencies between "good" speakers are small.

I definitely agree about the differences between good speakers being small, so being consistent matters when comparing across sources.

But also, aside from keeping the standard a standard, the 'wrong' method significantly masks the influence of vertical reflections on timbre and in my experience seems to be noticeably worse at predicting the in-room response because of it. Not huge, but enough to probably make an audible difference when EQing

Especially given how we tend to evaluate speakers based on small differences around these parts. In the 'proper' ER method the vertical components make up 40 percent of the final curve. Put another way, the 6 'ceiling' and 'floor' angles count for as much as the full 17 curves in the front horizontal hemisphere.

Meanwhile in the 'wrong' method, the vertical portion only accounts for 23 percent of the ER, with each direction counting for the same. The rear hemisphere also only accounts for 11 percent of the wrong method vs 20 percent in the correct one, so it would rear its head more with omnidirectional speakers and such.

 

[Xyrium]

Time stamped:

Thank you for this!

So, what make and model is that $1800 loudspeaker that they measured in the video at 1:03?

 

[napilopez]

Thank you for this!

So, what make and model is that $1800 loudspeaker that they measured in the video at 1:03?

Someone in another forum mentioned they thought is was an infinity il50. Can't find specific spins for it though

 

[Xyrium]

Someone in another forum mentioned they thought is was an infinity il50. Can't find specific spins for it though

Wow. I believe it. That's a nice design.

 

[thewas]

Several JBL synthesis spinorama have been recently uploaded to Speakerdata:
https://speakerdata2034.blogspot.com/2021/01/jbl-synthesis-spinorama-data.html?view=classic

 

[TiFi Speakers]

On wall and flush mount speakers by standard can be measured 180 degrees, horizontal and vertical as standard defines. However, I am in trouble finding the correct averaging formula for these measurements as "Rear" is always missing. Should it just to be skipped? And for sound power just double the weights measured in front hemisphere (as then total weight will be 1, as it should be for formula)? More confusing is that some Klippel measurements (i.e. https://www.erinsaudiocorner.com/loudspeakers/bw_cwm73s2/) present Rear for flush mounted speakers. Any hints, what part of standard I have missed?

 

[NTK]

On wall and flush mount speakers by standard can be measured 180 degrees, horizontal and vertical as standard defines. However, I am in trouble finding the correct averaging formula for these measurements as "Rear" is always missing. Should it just to be skipped? And for sound power just double the weights measured in front hemisphere (as then total weight will be 1, as it should be for formula)? More confusing is that some Klippel measurements (i.e. https://www.erinsaudiocorner.com/loudspeakers/bw_cwm73s2/) present Rear for flush mounted speakers. Any hints, what part of standard I have missed?

The CTA-2034 standard is, IMHO, more than a little unclear on the matter. Section 5 (Directivity Response — Standalone Loudspeakers) began by saying that in-wall/on-wall/in-ceiling/on-ceiling speakers aren't covered.

Section 6 (Directivity Response — In-Wall, In-Ceiling, On-Wall, On-Ceiling Loudspeakers) describe how in-wall/on-wall/in-ceiling/on-ceiling speakers are measured. However, it referenced section 5.3 for data presentation.

But for the 6 curves as shown in Figure 4, only the on-axis curve is sort of applicable (section 6 does not have the data post-processing subsections as in section 5 that produce the other 5 curves). And hardy anyone listens to in-wall speakers on-axis (or at least more than one in-wall speakers on-axis, or unless you have a purpose built room with angled walls). So IMO I would only present the raw measurements, and not any post-processing. There is, AFAIK, no published study that relate directivity of flush mounted speakers to listener preference. Thus the data, IMHO, is mostly of engineering interest — we have a good idea on what the ideal should be, but don't know how much we will tolerate which imperfections.

 

[TiFi Speakers]

Thank you for detailed reply. Just went into detail in spinorama.org and there are many flush mounted, some measured by producers (JBL mostly). There can be seen, that they present the directivity part (this seems reasonable). And as standard defines - average of measurements. If not measured by standard guidelines, this can not be included and the averages are calculated without. This still follows the ruleset, just n is smaller. For real life some probable listening angles should be covered, not for standard, but for customers

Someone should tell to Klippel also, that for flush/wall mounted it is not good idea to include Rear graph...

 

[napilopez]

On wall and flush mount speakers by standard can be measured 180 degrees, horizontal and vertical as standard defines. However, I am in trouble finding the correct averaging formula for these measurements as "Rear" is always missing. Should it just to be skipped? And for sound power just double the weights measured in front hemisphere (as then total weight will be 1, as it should be for formula)? More confusing is that some Klippel measurements (i.e. https://www.erinsaudiocorner.com/loudspeakers/bw_cwm73s2/) present Rear for flush mounted speakers. Any hints, what part of standard I have missed?

I'm not sure about the exact answer to this, but one thing to keep in mind is that "rear" technically includes +/- 90 degrees. So it's possible to make an incomplete "rear" measurement from simply the 90 degree response.

Granted, the graphs on Erin's site for 90 degrees and for the "rear" curve don't fully match, but looking at the contour plot, it seems the klippel tried to capture data out a little further than 95 degrees or so. Maybe there's weird averaging going on in that respect.

Or maybe Klippel decided to handle this some other way. One possible approach

All that being said, having made a few dozen spins myself, ultimately, I don't think it's a big deal. It's a fifth of the early reflections and an even smaller portion of the PIR. What I'd do in your case is to simply take the 90 degrees measurements as the rear component, make a note of that somewhere on the measurement presentation, and call it a day.

 

[TiFi Speakers]

Standard states: Predicted In-Room (PIR) amplitude response, also known as a “room curve” is obtained by a weighted average consisting of 12 % listening window, 44 % early reflections and 44 % sound power. (p55)

Is there additional data available on this topic?

Specifically, I am interested in understanding the deviation, as it seems that in real life in rooms with open plans and abundant soft furnishings, these weights might vary in favor of listening window. The question is how much they differ and whether the main variation is in the listening window weight or if, in some situations, early reflections versus sound power also have a significant impact.