It is phrases like that that would make me deeply suspicious of any speaker. What about other types of music, such as rock/pop, the type most of us actually listen to? The statement is somewhat self satisfying in that it implies that this aspect (acoustic reproduction) is the only benchmark against which to judge a speaker and it is by definition ‘more accurate’ when other speakers that cannot pull off this trick are not.
3 Likes
That is correct, it was not Toole and Olive themselves calling it a target curve.
Addition: When a speaker with flat anechoic response and constantly increasing directivity index is placed in a listening room with constant or decreasing RT60 towards higher frequencies. Or in other words, the Harman curve is indicative of a tonally imbalanced indirect soundfield with dominating bass/lower midrange and underrepresented brilliance/treble.
To which degree this would mean massive degradation of sound quality in listening tests is a matter of how much dominant the indirect soundfield is, how much its level is decreasing towards higher frequencies and which frequency bands are affected. According to studio specifications, frequencies below 300Hz and above 8000Hz can widely be ignored (as they have to be either judged in subjective listening tests or can be EQed in case no dramatic flaws are detected). In studios the general level of indirect sound is anyways lower and presumably showing a better tonal balance thanks to broad-band absorption. I found a constant or slightly increasing direcitivity index between 500Hz and 8K being ideal.
What I found to be most annoying is decreasing level of indirect sound in this band, particularly with drastic steps of several Decibels down or up within one octave. In many cases this is due to large midrange drivers playing too high or tweeters kicking in with either none or overly narrowed waveguides or a combination thereof.
Unfortunately this is pretty common with speakers which are blessed and explicitly praised as ´perfect directivity´ by the ´Harman curve misunderstood´ followers.
With such a broad baffle, it is to be expected that a certain level of D.I. is already achieved at 500Hz so potential problems with too much midrange and lack of brilliance in the room are not as dramatic. My advice is: take a close look at the behavior between 1K and 5K. The problem is you cannot EQ such problems later, and manipulating the direcitivity index with driver sizes and x-over frequencies is anything but trivial.
1 Like
My initial assessment of the off-axis response of the midrange and tweeter at the 3 kHz crossover frequency at 30°, 60° and 90° shows up a directivity mismatch which is consistent with the Seas data sheets. It wasn’t really obvious in the overall off-axis response in the passive speaker, however, I suspect the steep slopes of the UB jpol11 filters in the active setup may make it more apparent. Of course an easy fix would be to drop the X-over frequency to where the directivities match, however the 27TFFC shows an increasing rise in distortion products as the X-over frequency is lowered.
In the run-up to the active build, I did some measurements with tweeters and waveguides. The Seas DXT didn’t fit the bill unfortunately, a Monacor DT-300 and waveguide was too big, however the existing 27TFFC mated to a Visaton WG148R waveguide looks very promising. There were no surprises in FR or distortion and no new resonances on the CSD plot with the waveguide.
It would be a tight fight modifying the cabinet to take the waveguide without intruding into the midrange enclosure, but it’s possible with a bit of trimming on the rear of the waveguide and the tweeter faceplate.
Adapting the active crossover to a waveguided tweeter is trivial in Acourate. It’s a straight-forward near-field (~1D of the waveguide distance) re-linearisation and a new time-alignment to compensate for the new acoustic centre. Then a simple re-run of the filter creation and tuning.
A couple of hours’ work after modifying the cabinet.
1 Like
That is what I would expect and would personally avoid under any condition. Despite from choosing a smaller midrange, lowering the x-over frequency or choosing filter types with multiple slopes there is not much one can do with a given design.
If you want to know about directivity and do not have the possibility to do a 180deg polar measurement or DI via reverberation chamber, standard frequency responses under 70 and 110 deg are a pretty good indicator for directivity as they show a good average of all angles but are not overly affected by baffle and diffraction effects.
1 Like
I’m not averse to modifying the speaker - I built them myself - DIY Speaker Build - Troels Gravesen's Poor Man's Stradivari
The build took nearly a year, so a few extra hours to fit a wave-guided tweeter isn’t really a barrier.
I’ve got some tweaks to do with the current filter set. Once I’m happy, I’ll move on to the off-axis behaviour between the midrange and tweeter.
I have a spare of each driver and lots of plywood panel, so I can non-destructively model various options before undertaking any irreversible surgery on my carbon fibre cabinets!
Interesting discussion but may be should be moved to your DIY thread.
And surely you have some proof of that. Or is this just speculation?
Assumes facts not in evidence.
Ditto. Do you have any proof of this?
So, we’re back to trusting vendors and “trust me, bro.”
Facts not in evidence, again. They might just as well be deathly sick of the house sound.
Of course. Does anyone claim otherwise? As a generic preference seems to work well enough though.
Are high-end magazines anything but mindless entertainment?
If those fuses and green markers provides some measurable differences, it would be interesting to measure them, too. Of course the difference is that cables and fuses do not produce any results, while different electronic gizmos do, and often inversely proportional to their price, which makes measuring them worthwhile.
1 Like
@Boris you are asking for facts or evidence seemingly ignoring that I was writing about subjective listening preference and how it can be connected in some cases to measurements, in some not. By definition there cannot be any ´facts´ on this topic and some mass tests which were made within certain manufacturers to subsummarize an average taste are by no means facts.
Sorry to disagree, but cables produce a lot of objective differences when being measured. Many high end magazines were measuring cables publishing the results.
It is just widely accepted among people with basic understanding of psychoacoustics that these measurable objective differences do not constitute a base for explaining or predicting audible differences as they are too far off. Same situation with SNR, SINAD and similar measurements in DACs which are far from being audible with most of electronic components.
Soo far I see no difference in plausibility between the cables guys and the SINAD guys except for the former accepting that they are subjectivists and the latter forming some kind of religion based on their contradictions and misconceptions.
Exactly. And general subjective preferences can be measured, across population. You keep saying that existing available data is wrong, but somehow any and all of those supposedly better ones are unpublished, or proprietary, ort, most likely, just don’t exist.
Yeah, they have wildly different price tags. Weight and color are quite different, too. Are those important measures? Do they make any difference in signal? Despite what all those magazines might publish (and they will publish anything) no audible differences have been demonstrated, despite even Randi’s $million offer, either objectively through measurements, or subjectively once the price tag of the cable has been hidden.
Except that SINAD etc can be different, and some kit, be it a $5 dongle or a $50,000 audiophile DAC can actually measure badly enough to be audible. It makes sense to measure those, especiaslly when vendors make some rather outlandish claims that can be proven or disproven. For cables its known apriori that they make no difference, so why bother.
1 Like
Moderator note.
Endless Contrarianism
It’s ok to disagree with others, but if you find yourself engaged with the same battles over and over again, it’s time to move on. If you are developing a reputation around an issue or position, then you’re not building constructive discussion. Move on.
3 Likes
Guys, you both stop that nonsense, please!
Everyone should know these days, that some analog signal cable designs, being arrived at via painstaking listening sessions while being hand built by Nubian virgin princesses on full Moon only, do indeed show measurable differences, since they’re simply of an inadequate material mix, cross section and/or geometry.
Period!
2 Likes
Out of pasture-raised, free-range mandrake root fiber, no less!
If we amend Boris’ statement slightly, both @Arindal and @Boris_Molodyi are correct:
Sure, there are measurable differences between cables, but unless something is something is seriously out of whack (input/output impedance of devices or deliberate cable filters), the differences aren’t audible.
Of course, there was the peculiar quirk with some vintage Naim amps that certain speaker cables with any appreciable capacitance could drive them into oscillation.
3 Likes
As I recall, MIT Cables also started because certain equipment required cables with very certain parameters. And they probably could affect sound somehow, with those big boxes of fairy dust connected to them. To amend though, with properly designed equipment, no cable will be more accurate than a zipcord 
2 Likes
It cannot be measured, it can be avaraged and that will still be valid only for particular circumstances. And it is usually not very useful. I mean you can do a survey how many grams of salt people on an average base like in their burger, but knowing this amount does not give you to ability to make a good burger.
There is no existing data which would be applicable outside a very limited set of parameters.
Evidence, please, that an electronic component like a DAC does show results of standard measurement like THD or SNR bad enough that the result is surely audible (and we are not talking about tube amps or turntables).
Yes, perhaps some of this has drifted off-topic.
1 Like
I may be splitting hairs there, but to perform an average, you need to have a set of values, so you are measuring something. Some measurements are by definition averages.
(In fact, many things we measure are, under the covers, averages. For example, the pressure of a gas in a contaier is an average of all the impact forces exerted by atoms or molecules hitting the container walls.)
1 Like
I’ll attempt some spring cleaning this afternoon (Oz time).
Must say that this thread is interesting to view. So much knowledge there is here. Without having dived deep inte every single reference, there is one area that our human hearing is especially critical to. If it compares to meaurements I don’t know. In the natural world where our hearing once evolved, there were a few conditions that challanged our sensitivity/perception, while our modern world constantly do. The zero transitions of sound is according to findings/knowledge at BBC, one of the most critical parts of how sound is perceived. We are apperently very sensitive to this particular part of the wave. The conditions of how we enjoy music/perceive soundtoday is highly different and is probably matched only by living in caves or being close to natural reflecting walls, ie mountain cliffs.
I don’t know how to call this myself, but I guess I soon will hear from you guys. But the thing I think matters is the impossibility to actually perceive music as it once was presented to us, no acoustic intermodulation was present, unless in a cave. The problem is best described by testing how to talk to eachother oven 20 meters distance on an open field and in a concrete staircase of a modern rental complex.
Now the concrete staircase can represent all the various interference we perceive, wanted or unwanted, while the pure perception of sound would be in a field, perhaps also in a forrest. The process of making sense of the massive noise in a staircase conversation is highly dependent of the brain, while isolating the wanted sonic message in the field is far more simple.
Doing this exercise in thoughts around audibility for these two very different cases, I beleive explain a lot of the problems involved in hearing and the beauty of the system when it is working great and all conditions are beneficial.
It also demonstrates the problems for the brain in the latter case, the fatigue we suffer from poor invironments and finally how many properties that needs to be present for us to perceive a sound from an instrument or voice to be realistic.
I beleive the concrete staircase showcase all of the problems with hearing irrespective if their origin is acoustical, mechanical or electrical.
Sorry for not refering to any reports, my work with audibility is highly hands on and empiric and my focus nowadays is trying to eliminate or dampen as much potential sources of sound not belonging to source.
I do not beleive the manufacturers datasheet. I read them as a potential, but when I audition products connected as a system, I find it generating “stuff” making it hard to listen to. By experiencing this, I come to the conclusion that every device, every system and every room is a concrete staircase and it is impossible to clean everything. I have attended a demonstration in a recording studio, where two musicians where playing live and then we listened to the recording from analogue and digital sources. The gear was expensive as hell, we are talking about well over 1 million Euro and also the fancy looking highend cables. The recorded music was not really close. The discrepancy as I perceived it was the attack, the transients of particularly the marimba instrument. Such an impossible instrument to record… A nightmare in a concrete staircase. Instruments and methods for measuring all this, ie making the best possible estimate accuracy, need to be as good as the brain. We are still at the step AI, not equal machine or man I, meaning that measuring a system reproduced audio in a man-made room, still resembles a concrete staircase to much to make the reproduced audio realistic. As long as we are measuring the resemblance between source and playback from the media as source without success and not from original source, horses mouth, we will not able to to get closer to how man perceive sound. The source itself is not reality and showcase the technical imperfection as is.
As @Marian already mentioned, you can’t average without measuring.
Good… maybe not. A McD one, that lots of people find inoffensive, or good enough, for sure.
I think this is something tht needs to be demonstrated first.
Quite a few DACs in the lowest bucket on ASR are audibly different. Even something like a PowerNode can be distinguished by ear, although it’s harder.