Measurements Vs. Sound

Phelonious Ponk said:

Are you sure? If the differences were that audible - and I don't doubt for a second that they were - a very common on-axis FR measurement should reveal them.

Tim

Click to expand...

As I am not an expert I quote a Revel manual about their In-Room Response Relative to Target Response curve:

The in-room response curve is a prediction of how the loudspeaker would measure in a typical room. Research and observation reveals that ubiquitous on-axis response curves cannot distinguish between two loudspeakers with radically different sound qualities

I'd agree with that, especially for the heavily-smoothed responses typically included with most loudspeaker spec sheets, if the two responses are similar (no radical variations in on-axis FR among models).

Keep in mind, what loudspeakers do is even more complicated than normally considered.
Unlike how we measure with one microphone, we hear with two ears with hidden adaptation that lets us hear up and down, front to back as well as left right.

In addition to the familiar response flatness, distortion vs frequency vs loudness, dispersion in time, the speaker also radiates in 3d, as frequency dependent graduated loudness sphere as shown below or as a constant loudness lobe as in link two.

Consider how even the size of the enclosure can alter the radiation pattern and make what it does change when close to the speaker vs farther.

http://www.etcinc.us/tech/nl043_far_field_criteria.pdf

http://www.clfgroup.org/clf_SAC_Newsetter_fall_2004.pdf

Especially for multi-way or larger speaker systems, if one took all the measurements possible including spatial distributing vs frequency like in the links and you found two speakers that were identical in every way, they would nearly have to be another of speaker of the same model.
Best,
Tom Danley

Tom

Somewhat unrelated but would that explain my subjective impression of a small satellites /sub systems tending to sound "small" when compared to larger speakers? Even at equal SPL

There was an excellent presentation by John Atkinson at AXPONA. He covered the important measurements and what they meant. His view is that these measurements help inform our understanding but they don't replace trained listening. You really need both. I agree with this.

FrantzM said:

Tom

Somewhat unrelated but would that explain my subjective impression of a small satellites /sub systems tending to sound "small" when compared to larger speakers? Even at equal SPL

Click to expand...

Not Tom, so FWIWFM: I have found that for me the biggest problem with smaller speakers is not their dispersion patterns, but the fact that their small size means the subwoofer reaches above the localization threshold. This leaves you with two small higher-frequency point sources and a easily-located LF driver, causing a sort of discontinuity in the sound field over frequency. A pair of subs near the smaller speakers helps a lot, but then of course you lose the advantage of small speakers… - Don

DonH50 said:

Not Tom, so FWIWFM: I have found that for me the biggest problem with smaller speakers is not their dispersion patterns, but the fact that their small size means the subwoofer reaches above the localization threshold. This leaves you with two small higher-frequency point sources and a easily-located LF driver, causing a sort of discontinuity in the sound field over frequency. A pair of subs near the smaller speakers helps a lot, but then of course you lose the advantage of small speakers… - Don

Click to expand...

I agree. And would add that many smaller sub/sat systems have a problem moving enough air in the midbass. Most subs can't do it well, and most small sats are incapable.

Jeff Fritz said:

I agree. And would add that many smaller sub/sat systems have a problem moving enough air in the midbass. Most subs can't do it well, and most small sats are incapable.

Click to expand...

Jeff,

I also have the some idea about most sub/sat systems and understand what you mean, but such answer in a thread entitled "Measurements-Vs.-Sound" asks for a question - what type of measurement do you associate with this property you call "problem moving enough air in the midbass" ?

microstrip said:

Jeff,

I also have the some idea about most sub/sat systems and understand what you mean, but such answer in a thread entitled "Measurements-Vs.-Sound" asks for a question - what type of measurement do you associate with this property you call "problem moving enough air in the midbass" ?

Click to expand...

I'm not sure I can provide any definitive evidence. My observation is based on what I've heard from sub/sat systems that measure fairly flat but that seem to lack midbass weight. But I'll give it a shot nonetheless!

In my system I rarely feel that I get good subwoofer performance above about 60Hz. Generally, the lower the crossover the better in terms of integration. A speaker like the Energy RC-10 (a fine speaker for a small space), a small two-way with a smallish mid-woofer, starts to roll off anechoically at about 70Hz.

http://www.soundstagemagazine.com/measurements/energy_rc_10/

I'd rather have a more robust bookshelf speaker to pair-up, perhaps with a more gradual LF roll-off and that plays maybe 10Hz deeper. The lack of midbass weight I hear in some systems may be due to the bookshelf speaker not being able to fully meet the subwoofer within its comfortable operating range.

Jeff Fritz said:

I'm not sure I can provide any definitive evidence. My observation is based on what I've heard from sub/sat systems that measure fairly flat but that seem to lack midbass weight. But I'll give it a shot nonetheless!

In my system I rarely feel that I get good subwoofer performance above about 60Hz. Generally, the lower the crossover the better in terms of integration. A speaker like the Energy RC-10 (a fine speaker for a small space), a small two-way with a smallish mid-woofer, starts to roll off anechoically at about 70Hz.

http://www.soundstagemagazine.com/measurements/energy_rc_10/

I'd rather have a more robust bookshelf speaker to pair-up, perhaps with a more gradual LF roll-off and that plays maybe 10Hz deeper. The lack of midbass weight I hear in some systems may be due to the bookshelf speaker not being able to fully meet the subwoofer within its comfortable operating range.

Click to expand...

Have you measured the response at the listening position of subwoofers in the typical placement and the speakers in such a setup?

I don't disagree that many compact speakers lack the lower frequency capability required to produce a seamless blend, but often that is more of an issue when any sort of dynamics are required. Discontinuities are much more often functions of the response observed at the listening position and how the response of the speaker interacts with the room's transfer function. For this sort of case, it would not be surprising if there is a depression in the subwoofer's delivered response (@LP) above 50-70Hz. A valley in the loudspeaker's response can also drive very similar perceptions.

Full range speakers vs. separate LF devices is always a conundrum... With separate placement there is more opportunity (no guarantee) to get a smooth delivery of the frequency range to the listening position. The hurdle is that we are then required to get both the levels and crossover region to integrate smoothly. With a monolithic speaker we have the benefit of the crossover being dialed in such that a good loudspeaker will deliver even power into the room from the midrange down to the lowest frequencies. We are then dependent on and limited to placement choices and acoustic treatment efforts in getting that smooth power response to the listening position while also finding a good location for the rest of the frequency range in terms of balance and imaging. In simple terms we're given more rope to hang ourselves with in a modular system, but much more flexibility.

Lower frequency crossovers tend to push the integration problems out of critical/sensitive ranges and in narrowing the bandwidth of the subwoofer, allow simple level adjustment to have greater impact on the end result. Of course never in the past have we had so readily available measurement equipment such as the Dayton OmniMic, XTZ room analyzer or the free REW software (along with a handful of other options).

Mark Seaton said:

Have you measured the response at the listening position of subwoofers in the typical placement and the speakers in such a setup?

I don't disagree that many compact speakers lack the lower frequency capability required to produce a seamless blend, but often that is more of an issue when any sort of dynamics are required.

Click to expand...

Yes, I've done in-room measurements many, many times, and have achieved excellent results with subwoofer integration: http://www.ultraaudio.com/twbas/twbas_20080401.htm

I do think you are right in terms of dynamics. The subwoofer being more capable at the top of its passband versus the sat being less capable at the bottom its passband.

Listening to a good four-way speaker with a dedicated midbass driver is instructive on just how much air needs to be moved in the 60Hz to 120Hz range.

I can't contribute to the technical part of this conversation, but I've heard what you guys are talking about: Floor-standing speakers with multiple small woofers giving a greater sense of girth, or presence, or scale, than bookshelf speakers paired with subs. It makes sense, of course. In the simplest terms, three six inch drivers are more than one, and it sounds so.

I'm just not sure it's always a good thing.

When listening to orchestral music, or large-scale rock music, all that driver surface seems to be a big positive. But sometimes, when playing acoustic music, or a jazz quintet, they actually seem to get in the way, creating an oversized lower midrange that is not as open and natural as what you can get from a pair of excellent bookshelf speakers and a well-placed sub or two. At least that's what I hear. I hear strengths and compromises in both approaches.

Tim

Phelonious Ponk said:

I can't contribute to the technical part of this conversation, but I've heard what you guys are talking about: Floor-standing speakers with multiple small woofers giving a greater sense of girth, or presence, or scale, than bookshelf speakers paired with subs. It makes sense, of course. In the simplest terms, three six inch drivers are more than one, and it sounds so.

I'm just not sure it's always a good thing.

When listening to orchestral music, or large-scale rock music, all that driver surface seems to be a big positive. But sometimes, when playing acoustic music, or a jazz quintet, they actually seem to get in the way, creating an oversized lower midrange that is not as open and natural as what you can get from a pair of excellent bookshelf speakers and a well-placed sub or two. At least that's what I hear. I hear strengths and compromises in both approaches.

Tim

Click to expand...

Driver integration in any multiple-driver speaker system is key to good performance. That said, certain types of music play into the strengths of certain designs. Acoustic music may not present a large amount of the problem frequencies of sub/monitor configurations.

Lee

Listening to a good four-way speaker with a dedicated midbass driver is instructive on just how much air needs to be moved in the 60Hz to 120Hz range.

Click to expand...

Hello Jeff

I have always wondered about that. It seems almost like the lower the driver excursion the better. I have 6" 2 ways and larger systems with 15's covering that range and if you look at the systems from an excursion point of view the 15's are barely moving where the 6's are really working. So what's the difference if both systems are actually moving the same amount of air?? You won't see an SPL difference may be a difference in distortion but how audible is it??

Rob

Robh3606 said:

Hello Jeff

I have always wondered about that. It seems almost like the lower the driver excursion the better. I have 6" 2 ways and larger systems with 15's covering that range and if you look at the systems from an excursion point of view the 15's are barely moving where the 6's are really working. So what's the difference if both systems are actually moving the same amount of air?? You won't see an SPL difference may be a difference in distortion but how audible is it??

Rob

Click to expand...

If there is an audible difference, it's more likely to be a matter of dynamic compression/linearity (ie does the observed sound level jump 20dB when the recorded signal does). That is not to say that you need huge drivers to have good dynamic performance, but rather you need enough linear capability for the use, including the brief peaks which are well above the average level which might only use a fraction to a few watts.

Driver integration in any multiple-driver speaker system is key to good performance. That said, certain types of music play into the strengths of certain designs. Acoustic music may not present a large amount of the problem frequencies of sub/monitor configurations.

Click to expand...

Understanding that this is a purely subjective evaluation, it sounds exactly the opposite, as if the floor standers are creating a low-mid emphasis that masks the clarity and space around the instruments in smaller ensemble work. It could be that I'm hearing that wrong, but it could also be that each configuration has its own strengths and weaknesses. I'd love to have the opportunity to compare a couple of pair of stacked active monitors and a couple of subs to floor standers of a comparable configuration. My guess is that the scale would be achieved without the loss of clarity and separation.

Tim

Jeff Fritz said:

Yes, I've done in-room measurements many, many times, and have achieved excellent results with subwoofer integration: http://www.ultraaudio.com/twbas/twbas_20080401.htm

I do think you are right in terms of dynamics. The subwoofer being more capable at the top of its passband versus the sat being less capable at the bottom its passband.

Listening to a good four-way speaker with a dedicated midbass driver is instructive on just how much air needs to be moved in the 60Hz to 120Hz range.

Click to expand...

Hi Jeff,

Thanks for the link and graphs. Many compact loudspeakers are increasingly limited in dynamics below 150Hz, especially compared to most subwoofers. That discontinuity in linearity can be quickly audible in real listening.

Looking to your curves nicely shows some of the hurdles often encountered. It's clear that the two locations you measured have very different response below 100Hz. The very full range Altair yields a useful tapering down to the subwoofer (especially if there is no electronic high pass on the main speakers) in the first position measured as seen here:


When you moved them closer to the front wall, there was more efficient delivery of the lower range, with trade offs in the spacial qualities:


While loudspeakers with woofers in very different locations (such as bookshelves) will interact somewhat differently, the differences in locations should be more similar than different. A more bass limited speaker which is already rolling off in the 50-100Hz range might not fare so well in the first location.

While difficult to say clearly given the scale and size, looking at the following graph hints at why deeper extension from main speakers is important in your system:


To better see what is going on we would want to see the response of the subwoofer with the low pass filter switched off, as this would better show how much of the shape in the 60-110Hz range was a depression in the response vs. just a peak sticking up near 120Hz. Given what looks like a suckout/recession of these upper subwoofer frequencies due to the room interaction, it is not surprising at all that a speaker with power to at least 60-70Hz would be preferred. In a different situation/room where you might see smooth response from 50-100Hz prior to EQ or low pass filtering, limited low frequency extension of a compact speaker might not be as apparent.

Of course there are always matters of truth with any observed trends... The range above 40-50Hz is often a roller coaster response from room to room and location to location, making the integration of a subwoofer in this range less conducive to plug-n-play across various systems.

Mark Seaton said:

Hi Jeff,


Looking to your curves nicely shows some of the hurdles often encountered. It's clear that the two locations you measured have very different response below 100Hz.

Click to expand...

Mark, here are two graphs taken of two full-range systems in their optimal locations in my room for such speakers:

http://www.ultraaudio.com/opinion/20100501.htm

These are 1/6th octave, so a little more telling. Still, you can't tell in these graphs that one speaker is a four-way and the other a smaller three-way, yet they measure very close. The sound, however, was quite different (the Arrakis being superior in every regard).

I recently had an opportunity to A/B the Magico Mini II and the new Q1, and it was far more dramatic than what you might expect to find from a pair of two-ways from the same maker. I could easily see where they might measure similarly in-room (a guess). Yet the Q1 was simply better in every parameter -- stunning differences. A writer of mine asked me recently, "How do we measure real improvements in detail?" I know all the usual answers relating to prominences in certain bands, noise, etc., but still the answer is not always clear.

I believe in getting the FR smooth and neutral in-room, but you still have to listen to really absorb the sound. (That from someone that has spent a LOT of time measuring and getting my room right.)

Jeff Fritz said:

I recently had an opportunity to A/B the Magico Mini II and the new Q1, and it was far more dramatic than what you might expect to find from a pair of two-ways from the same maker. I could easily see where they might measure similarly in-room (a guess). Yet the Q1 was simply better in every parameter -- stunning differences. A writer of mine asked me recently, "How do we measure real improvements in detail?" I know all the usual answers relating to prominences in certain bands, noise, etc., but still the answer is not always clear.

Click to expand...

Do the Magico guys have any ideas about this?

Jeff Fritz said:

Mark, here are two graphs taken of two full-range systems in their optimal locations in my room for such speakers:

http://www.ultraaudio.com/opinion/20100501.htm

These are 1/6th octave, so a little more telling. Still, you can't tell in these graphs that one speaker is a four-way and the other a smaller three-way, yet they measure very close. The sound, however, was quite different (the Arrakis being superior in every regard).

I recently had an opportunity to A/B the Magico Mini II and the new Q1, and it was far more dramatic than what you might expect to find from a pair of two-ways from the same maker. I could easily see where they might measure similarly in-room (a guess). Yet the Q1 was simply better in every parameter -- stunning differences. A writer of mine asked me recently, "How do we measure real improvements in detail?" I know all the usual answers relating to prominences in certain bands, noise, etc., but still the answer is not always clear.

I believe in getting the FR smooth and neutral in-room, but you still have to listen to really absorb the sound. (That from someone that has spent a LOT of time measuring and getting my room right.)

Click to expand...

Hi Jeff,

My comments earlier were only referencing the sub-bookshelf integration comments. The comments I made above were entirely in reference to the low frequency behavior, not the full range sound quality which involves so many more dimensions than an RTA measurement. I regularly post that no single measurement will define nor reveal the total subjective experience. We are on common ground there.

The many spacial and time domain characteristics of a loudspeaker are no where to be found in a smoothed RTA measurement, and won't be easily revealed by frequency response measurements at the listening position. Measuring the frequency response at the listening position is more an effort in confirming what of the intended energy is making it to the listener in a given room. It can be helpful in identifying differences in what makes it to the listener's ears between the L/R speakers which might impact imaging or highlight some room acoustic or setup mis-alignment. As you have explained well in various articles such as those linked, various measurements make for great tools in coaxing the best performance from a given loudspeaker design in your room, and can highlight some deficiencies, but do not absolutely define what makes for an enjoyable listening experience.

In my experience I would expect the two curves of the Dynaudio and Rockport examples you posted to sound quite different, even if they were emitted from the identical loudspeaker. When we add the differences between the two speakers in how they radiate sound into the room, their internal or structural resonances and even diffraction off enclosure edges we find many contributing factors toward very different subjective results in the same room and position. As you noted, readers need to be careful to not get too excited over any single measurement, especially one taken in-room.

I do believe that measurement of loudspeakers in isolation such as you have facilitated with the NRC are much more telling of differences in loudspeakers. Yet again it is tempting for many to focus on a particular measurement or 3, where they really don't provide much insight until an array of correlated measurements are assembled, with interpretation being far from obvious...

In the case of the two Magico 2 ways, your experience isn't all that surprising. Changes in drivers and crossovers alone can easily make for very obvious sonic differences where an on-axis measurement is fairly similar. I am certain that some exploring of the on/off axis response variations, in and out of band driver resonances, and overall spectral balance would reveal plenty of differences. The problem is that those are not an easy set of measurements to take, the measurements of interest will be different for different speakers, and they are not simple to decipher. The work done at Harman is probably the closest to an aggregate view I've seen, and even in that case there must be some assumptions in "proper/normal function" which do not always hold true in the exotic audiophile market.