Dr. Klaus Heinz of HEDD Audio (ex ADAM Audio) - measuring speakers, in particular speaker dynamics

[andreasmaaan]

Yes, phase-unwrapping means taking the allpass response of the XO (also known as the excess phase), inverting it in time and use it as convolution kernel to undo the original allpass behavior. The system now behaves like a single way with the same magnitude response. No ringing at all, at least on the design axis, as noted. By design, literally.

Assume a simple crossover at 1kHz with ideal drivers (infinite flat response from DC to light).
Let LP(s) be the original lowpass and HP(s) the highpass. LP(s)+HP(s) form the system sum which is an allpass AP(s).
The time inverse allpass be denoted as IAP(s). It is immediately clear that IAP(s)*[LP(s)+HPs(s)] is the same as IAP(s)*LP(s)+IAP(s)*HP(s).

Well it's hard to argue with the maths: I stand corrected

 

[Head_Unit]

...Spendor BC1 I sought out a dealer and went for a listen. On speech and string quartet they were really impressive but on orchestral and rock music very, very disappointing...

This is true of many speakers, and a reason I find audio shows frustrating for always playing like solo female vocal, classical, jazz. Many speakers can sound great with that stuff but turn in to a nasty hash of distortion if you try to play heavy metal and hard rock. Back in the day I wanted to start some kind of standardized multi-sine-tone testing as implemented on one of the speaker measuring systems and also incorporated by Audio Precision (FASTTEST). I thought it would be interesting to see how much "hash" filled in between the sine tones at some high reference SPLs. Still not a dynamic test but at least a step towards separating the truly high SPL capable speakers versus lesser creatures. I never got too far as I pivoted into product planning and marketing and many other things.

 

[pozz]

Back in the day I wanted to start some kind of standardized multi-sine-tone testing as implemented on one of the speaker measuring systems and also incorporated by Audio Precision (FASTTEST). I thought it would be interesting to see how much "hash" filled in between the sine tones at some high reference SPLs. Still not a dynamic test but at least a step towards separating the truly high SPL capable speakers versus lesser creatures. I never got too far as I pivoted into product planning and marketing and many other things.

Do you think it would be possible to implement this test using home measurement gear?

 

[Purité Audio]

Chris and Steve ( CUK Audio) brought a pair of ‘Tower mains’ around , six modules, total 280 kilos, , they are awesome in every regard.


Keith

 

[pozz]

What are your impressions Keith?

 

[Purité Audio]

They are super Pozz, they definitely have punch!
Keith

 

[Soniclife]

Chris and Steve ( CUK Audio) brought a pair of ‘Tower mains’ around , six modules, total 280 kilos, , they are awesome in every regard.


Keith

They just blend into a domestic room don't they. I've no idea if I want a pair, but I definitely want a room big enough they wouldn't dominate.

 

[pozz]

You would have to sit fairly low to get correct on-axis response. The Towers would only work in this configuration in a living room/listening room. For studios I would think that soffit mounting would be more ideal.

 

[Purité Audio]

Tweeters are at the ’usual’ sort of height , unsurprisingly they sound similar to the other fine monitors I have here, they just have more presence.
I will have to get some, they are really super, ordered some of the type 20’s too, just to hear how they compare.
Keith

 

[j_j]

So the ammount of pre and post ringing is related to the number of taps used in a filter?

Yes, but not quite in the way you might think. If you make a filter shorter, with different ripple structures in passband and stop band, you can have WORSE pre-echo with the shorter FIR filter.

Yeah. Hence my "it's complicated" comment.

 

[Krunok]

Yes, but not quite in the way you might think. If you make a filter shorter, with different ripple structures in passband and stop band, you can have WORSE pre-echo with the shorter FIR filter.

Yeah. Hence my "it's complicated" comment.

Thank you.

My 44.1kHz FIR filters currently have 65536 taps. Do you think it would make sense to try with smaller number of taps, say 16384?

P.S. delay is of no concern to me, my question is related to reducing pre and post ringing

 

[j_j]

Thank you.

My 44.1kHz FIR filters currently have 65536 taps. Do you think it would make sense to try with smaller number of taps, say 16384?

P.S. delay is of no concern to me, my question is related to reducing pre and post ringing

The answer depends on the impulse response you are correcting. Even with enormous filters like you describe, most of those coefficients are likely to be teeny-tiny, and therefore harmless.

BUT, the answer is non-trivial and depends on the individual corrections you are making, including any phase corrections.

 

[Krunok]

Here is how it looks, blue is with 65536 and red with 16384 taps. Obviously some precision, more in amplitude than in phase, is lost when number of taps is reduced.

 

[j_j]

Blue and red are the measured results or which? There is an obvious time delay in what I presume is the phase plot of course.

Both look kind of overkill to me, but since you are correcting phase, you don't have a symmetric FIR, so the chance of pre-echo is pretty small. Probably.

 

[bennybbbx]

Thanks @dc655321 cool!

It matches the sim in the DSP software (which includes the filter delay in ms). @Krunok it does settle, again, matching pretty well with @dc655321 Python code:

View attachment 27584

The commercial audio DSP software I use is very, very good

what speaker is this step response ?. there stand eva nr i did not find in internet. step response look very good. In my tests i notice that large bass/mid with slow step response reduce stereo width much. the hedd 05 step response look worse and on a german test of hedd05 on bonedo he write that the reverb sound not realistic. thats what i hear too on speakers with slow step reponse. here can see and hear recordet examples and step responses https://www.audiosciencereview.com/...of-the-synth-sound-in-the-attached-mp3.18401/ maybe can post more step responses of speakers.

 

[mitchco]

what speaker is this step response ?

Speaker and room... https://www.avsforum.com/threads/of...woofer-thread.1214550/page-1834#post-57390652

Another example: https://audiophilestyle.com/ca/bits-and-bytes/what-is-accurate-sound-r923/

 

[Head_Unit]

And I guess that 'something' that he feels is missing from smaller boxes is dynamics - even if they can measure the same. He really emphasises that measurements tell you nothing about how the speaker will sound. He does have a lot of experience. I ask respectfully but do you have professional speaker design experience?

I tend to agree with his statements, and yes I'm actually a loudspeaker design engineer. Having said that, I must caveat that my feelings about small-driver speakers are based on listening to not really so many. And a couple-some long ago very expensive 4" KEF, and the 4" ELAC Debut-actually sound pretty lively.

That's a huge problem in audio (or conversely, a huge opportunity for selling snake oil): very very few people really listen to a lot of different stuff, and almost never under any kind of controlled conditions. Therefore conclusions get made from very limited and unqualified data sets. As a personal for instance, I don't believe power cables or speaker cables or RCAs* make much difference unless perhaps a system is very expensive...but while I have indeed changed each of those at various time the experimental set is very small.


*aside from shielding/RFI issues...where yet again every system differs and someone could have an amazing experience, which means NOTHING to anyone else's system.

 

[Head_Unit]

Once you add a sub to the small loudspeaker you will typically reduce it's bass extension thus it's cone displacement would be pretty much equal that of the large speaker.

Yes...if if IF there is a highpass filter, at a significantly high frequency. Which from reading a lot of posts on various forums, often doesn't happen, especially since a lot of 2-channel equipment is not capable. Even, stupidly, most units with "subwoofer" outputs

 

[Head_Unit]

Can you recommend some primary literature as well? I'll get to scaling that mountain at some point.

Heh, that is how I got sucked into going to engineering school to design loudspeakers, it is so seductive! The more you learn the less you know, it's quite a rabbit hole. After reading David Weems' How To Design, Build And Test Loudspeakers (bless you David, wherever you may be!!!) I made an older friend who had a drivers' license drive me to Northern Illinois University's Physics Library where I photocopied a huge stack of Audio Engineering Society Journal articles referenced in the book.

https://customer258769455.portal.me...ontextID=dfb59c4a-0066-c63d-c975-0b3b93bf27d5
links to the Audio Engineering Society Anthologies, which were basically collections of AES Journal articles. Frustratingly, it does not show detailed contents, but IIRC Loudspeakers Volume I contained Small and others' seminal articles including my old boss Jim Novak. Now if you want to climb Olympus Mons, Venmo me and I'll make you a copy of my copy of Small's entire thesis, whew that is great but a big slog. If you really LOVE math you could try to find J. E. Benson's articles in the journal of Amalgamated Wireless Australia.

 

[bennybbbx]

Speaker and room... https://www.avsforum.com/threads/of...woofer-thread.1214550/page-1834#post-57390652

Another example: https://audiophilestyle.com/ca/bits-and-bytes/what-is-accurate-sound-r923/

intereseting your page. do you think step response is usefull to know too ?. can you please post step response of KEF LS50 ? . from the hedd 05 i find a step response too. it look worse . http://nuke.nonsoloaudiofili.com/HEDDType05SeriesONENearfieldStudioMonitor/tabid/504/Default.aspx and here is explain in german what i hear with LP6 too.

https://www.bonedo.de/artikel/einzelansicht/hedd-type-05-und-type-07-test/2.html

Auffällig ist, dass die HEDDs sehr direkt und plastisch klingen mit einer guten Abbildung im Stereopanorama, aber gleichzeitig etwas trocken wirken. Räume, auch künstlich erzeugte Hallräume, werden nicht so fein aufgelöst und abgebildet wie ich das von anderen 2-Weg-Systemen kenne. Es wirkt zum Beispiel, als ob der Sänger direkt vor dem Hörer steht in einem relativ trockenen Raum steht. Man kann die Feinheiten und eventuelle Ungenauigkeiten des Sänger und der einzelnen Instrumente auch im fertigen Mix noch heraushören, fast wie mit einer akustischen Lupe. Dies macht aber das Finden der richtigen Balance zwischen Hauptsignal und Hall nicht einfacher.

translate with google to english

It is noticeable that the HEDDs sound very direct and plastic with a good image in the stereo panorama, but at the same time appear a bit dry. Rooms, including artificially created reverberation rooms, are not as finely resolved and mapped as I know from other 2-way systems. For example, it seems as if the singer is standing directly in front of the listener in a relatively dry room. You can still hear the subtleties and possible inaccuracies of the singer and the individual instruments in the finished mix, almost like with an acoustic magnifying glass. However, this does not make it easier to find the right balance between the main signal and the reverb.