Interview about Measurements with Pearl Sibelius Speaker Designer

[Yuhasz01]

Damn, objective measurements, subjective evaluation( messy- room, music genre and personal aesthetic preferences) all have value to a degree. Neither can be dismissed.
Old dichotomy that keeps raising its head again and again regarding speaker evaluations. VERY boring tired arguments.

 

[GXAlan]

Is it because they are hard/impossible to differentiate?

You can differentiate it, though subtle. Looking at comments, there are those who can find that they prefer the less transparent option.

Anyway we are talking loudspeakers in this thread and whether it is a good idea (or if it even matters) to choose one that is a long way from perfect. Personally I think it does matter and that cannot be waved away by saying it's all personal preference.

It’s not all personal preference, but how you define a long way from perfect is hard. You just have to look at Klipsch and B&W. They don’t measure well, but I certainly believe that owners of those speakers enjoy the special effects delivered by those speakers.

 

[Mart68]

You can differentiate it, though subtle. Looking at comments, there are those who can find that they prefer the less transparent option.


It’s not all personal preference, but how you define a long way from perfect is hard. You just have to look at Klipsch and B&W. They don’t measure well, but I certainly believe that owners of those speakers enjoy the special effects delivered by those speakers.

Some owners maybe, I do come across a fair few people dissatisfied with their modern B&W speakers. Klipsch much less so although I did have a Klipsch owner here last year who commented on how 'refined' my speakers were by comparison. It's clear he hears the problems with his Cornwalls so why he sticks with them is a bit of a mystery.

I often find that owners of bad speakers tend to blame the problem on something else. The recording is a popular scapegoat, but changing DACs and cables as frequently as changing underwear is also common. A year or two down the line they will finally accept that the speaker isn't working for them and change it - usually for something equally poor - and then the process begins again.

What I am saying is that if we accept that a speaker can be objectively bad, then logically it must also be subjectively bad.

 

[thewas]

As scientific audiophiles, we appreciate the value of EQ/room corrrection. But while it’s easy to talk about a target curve for our primary reference system, the idea that it is “easier and more expedient” to use equalizers and digital sound effects to match a specific recording and mood is not actually true.

It’s the analogy of arguing for a pantry of individual spices to season to taste than it is to buy pre-made seasoning mixes.

It is much easier to have two systems and pick the one that you like depending on the mood.

I really don't see why EQs and effect plugins are not easier and more expedient to adapt to different moods and recordings than having different systems, can you please explain it for me?

 

[GXAlan]

I really don't see why EQs and effect plugins are not easier and more expedient to adapt to different moods and recordings than having different systems, can you please explain it for me?

Sure. This relates to “theory of mind” in that it’s hard to imagine it because you yourself don’t approach music this way with a preference for coloration.

Suppose I like the effect of the 300B SET for singer songwriter music but I want maximum transparency for my John Williams classical concert. We haven’t quantified all of the 300B SET effects but one that has been quantified is an inverted loudness. You stay neutral for content that is say -40 to -20 dBFS and during transients you get the smiley face EQ. It makes sense. Smiley face EQ is fun for a while but it gets fatiguing. If it only happens during transients, it could be great in small doses.

1) What software player should I buy?
2) What plug-in software should I buy?
3) What numbers should I use in the variables available for that plug in?
4) How can I assign that album to have enable that effect plug in but have a different plug in for the other album?

It’s a genuine question and ends up being a not-so-straightforward answer.

Even simpler are harmonic distortion plugins. Maybe I don’t need the fancy content level dependent manipulation. What EQ plug-in is available that is compatible with the jukebox software that you can recommend.

A ha! That’s the problem. It seems like it should work easily and in principle, it IS true. But in practice, it doesn’t because the combo of software doesn’t really exist …. Or at least I don’t know of one.

The alternative is:

I go to my home theater room when I want transparency.
vs.
I go to my office when I am playing music that I want special effects.

 

[thewas]

As far as I know there exist quite few good tube emulations which offer also a quite big variability and flexibility compared for example to your 300B amp.

Also the other question is, do you really prefer this specific non-linear characteristics of the 300B so much over anything else or would another tube implementation or combination of variable loudness compressor/expander etc. give you even more pleasure? That question is definetely easier to answer on a digital simulation domain compared to buying, borrowing and carrying dozens of amps.

Now about your Jukebox player, I use and recommend EQ APO where you can enter convolution filters and VSTs system wide (Windows) independently from the playback source.

 

[MattHooper]

Sorry, I know I'm responding to an old comment, but since this thread just popped up....

I didn't like much of what was in there. His logic that he uses colored mics so speakers can be colored made no sense to me. Why would I want to perform double coloration that way? Wouldn't I want a neutral speaker to hear what coloration he added specifically by the use of the specific mic? Why would I want every piece of music colored all over again by the speaker?

(my emphasis)

The answer is, at least in principle: for roughly the same reasons one would use an equalizer.

People use EQ to "fix" or make things sound better, often due to deficiencies or colorations in the recording process. Equing down exaggerated sibilance being a classic example. When you do that you are "performing a double coloration" in the sense that a coloration was added by the recording process, and you are altering the signal - adding another coloration - that ameliorates or "corrects" for the original coloration. So you can actually end up with "better sounding" or "more natural."

In principle, how could this work in speakers? We've bee discussing in another thread the Devore O speakers which add resonance to the sound. I found they added a coloration that beefed up and gave body and resonance to the overall sound. Since I find most recordings, and stereo playback, reductive relative to the real thing - that is typically instruments will sound thinner, less dense, less body and presence - the Devores to my ears seemed to give back some of what is lost in the recordings, and so lots of instruments sounded more like the real thing to me. I found the resonance was a bit too obvious on some stuff, but cannily integrated enough that, on most tracks, it "worked" for me.

So, what if we even grant that is true? Wouldn't that still mean that, while gear with such coloration may sound good on some tracks, it won't sound right on other tracks? Sure. But that's precisely the situation we face anyway with accurate speakers, given the variability of recordings: you will constantly be greeted with some tracks with vocals or instruments sounding rich and natural, others where they sound odd and off. So there is no automatic advantage for an uncolored speaker *in that sense* of the argument. If it turns out that someone finds a certain coloration aids his enjoyment of most of the music he likes, then that could be the way to go.

(Keep in mind I'm addressing the internal logic of Amir's question about "double coloration." An argument about what blind research shows regarding general speaker preferences is a slightly different conversation).

 

[GXAlan]

As far as I know there exist quite few good tube emulations which offer …

Best guitar VSTs 2025: Get the best simulated guitar tones for your next production

My pick of the best guitar VST instruments will get great guitar tones for your productions, even if you’ve never played guitar before

www.musicradar.com

Again, I don’t easily find the solution.

It is very possible that tweaking could be even better - but where do you start? This is the gap that still exists.

In digital photography, you do have film simulations. But not so obvious for playback.

I use and recommend EQ APO where you can enter convolution filters and VSTs system wide (Windows) independently from the playback source.

Thanks for the recommendation!

 

[312elements]

I think the assumption that a speaker should be good at playing all kinds of music is a flawed one. I also find the assumption that everyone cares or should care about off axis measurements flawed.

I don’t own any Pearl speakers, but I did build a pair of monitors using a similar mark audio driver.

For reference, I have a couple of pairs of Kali monitors in my home theater, a pair of LS50 metas in my photo studio, and a pair of L82 classics in my living room.

The DIY speakers are in my listening room where I spend most of my time enjoying music. In that room, I don’t care about any seat besides the mlp. The room is heavily treated and the speakers are crossed over at about 150hz. There are things those speakers do better than any of the other ones and I attribute that mostly to the shape of the enclosure and the lack of a crossover. I love them for what they do differently. They excel at jazz, singer songwriter acoustic, and rock up to a point. That covers about 90% of what I listen to. At the MLP the frequency response is relatively flat and with this system I prefer a little harmonic distortion with my music.

The Kali’s sound good as long as you’re not sitting too closely, the LS50 meta are enjoyable with lots of clean power, DSP, and a pair of subwoofers. The L82’s sound shockingly great in the open floor plan room they’re in with clean power and a little DSP.

These Pearl speakers are a niche product for sure. I’d likely only enjoy them in my listening room, but I’m confident that I would enjoy them.

I’d probably rank my enjoyment of my own speakers from DIY, L82, LS50, Kali. Most would probably tell me that the metas are the best speaker in my possession. Different strokes for different folks.

 

[maty]

I attribute that mostly to the shape of the enclosure and the lack of a crossover.

Or simple first crossover with good/very good components + few and not agressive EQ + rePhase -> convolution filter... if the driver(s) is/are very good.

 

[Geert]

Harman, in spite of its wealth and Toole's praiseworthy efforts, tried and failed.

No they didn't. The result of their work is a system that allows them to very reliable predict speaker preference in an objective test. Their work revolutionised speaker design.

You're mixing this up with people including all kinds of subjective elements in their judgement of speaker quality, in most cases not even doing the slightest attempt to objectify their listening test.

 

[HarmonicTHD]

Best guitar VSTs 2025: Get the best simulated guitar tones for your next production

My pick of the best guitar VST instruments will get great guitar tones for your productions, even if you’ve never played guitar before

www.musicradar.com

Again, I don’t easily find the solution.

It is very possible that tweaking could be even better - but where do you start? This is the gap that still exists.

In digital photography, you do have film simulations. But not so obvious for playback.


Thanks for the recommendation!

But finding the „solution“ by keep switching gear until one finds the gear (tube amp) which effects meet (by accident) one’s personal preference is also time (and money) consuming. So I don’t see it as an argument.

If I was into tube sound, I get the VSTs and start tweaking until I like (prefer) it. And if my preference (mood) changes, I go back and tweak some more …. much less time consuming and expensive and more flexible. (Especially after you having measured the 300B, which even gives a good level for comparison between the real thing and the simulation eg derive a convolution file from it and put it into eg APOEQ or a DAW. Or just use the DAW and with a bit of effort you can mimick the 300B behavior including the dynamic FR change. The 22dB SINAD is easy to dial in).

I use simulated guitar tube amps (and speaker cabinets) in my DAW all the time and I bet people would be hard pressed to distinguish those from the real thing in a blind test. Just saying… and no I don’t want to criticize your approach (happy you enjoy it) just give another perspective. (Btw. The amp simulations often start with measurements and derive a convolution file from them).

 

[Geert]

I also find the assumption that everyone cares or should care about off axis measurements flawed.

Because your "room is heavily treated". So that's an exception that proves the rule.

 

[tuga]

No they didn't. The result of their work is a system that allows them to very reliable predict speaker preference in an objective test. Their work revolutionised speaker design.

Very realiable? Even Olive says the contrary...

"You shouldn't focus on preference scores. It was only intended to help naive people interpret spinorama measurements and reduce it to a single number."

You're mixing this up with people including all kinds of subjective elements in their judgement of speaker quality, in most cases not even doing the slightest attempt to objectify their listening test.

Preference IS subjective and has nothing to do with speaker quality (measured performace).

 

[Geert]

Very realiable? Even Olive says the contrary...

Reference?

Here we have a few quotes of F. Toole stating the opposite:

The simple answer is "no", listener sound quality ratings are remarkably consistent in double-blind tests.

In general the most neutral, least colored, loudspeakers are awarded the highest scores.

And very relevant to this topic, where people claim a speaker can sound 'fantastic' despite a flaw in the frequency response:

The first surprise, beginning with the initial crude blind listening tests in 1966, was that most people, most of the time, agreed on the loudspeakers they preferred. The sound quality scores were highly repeatable in successive randomized presentations. Where was the much touted “personal preference”? People who volunteered their loudspeakers for these early tests sometimes ended up disliking the high-priced “audiophile” products they had been living with. They had clearly adapted to the colorations sufficiently that music was enjoyed. They ended up changing their loudspeakers, having heard ones that were more appealing.

 

[tuga]

Reference?

I've edited my post.

And very relevant to this topic, where people claim a speaker can sound 'fantastic' despite a flaw in the frequency response:

In my view the methodology is flawed and consequently the results are void.

 

[Geert]

Very realiable? Even Olive says the contrary...

"You shouldn't focus on preference scores. It was only intended to help naive people interpret spinorama measurements and reduce it to a single number."

No one said you should go blind by preference scores, it's what behind these scores that matters. Toole knows that very well, obviously: Post in thread 'I cannot trust the Harman speaker preference score' https://audiosciencereview.com/foru...n-speaker-preference-score.31454/post-1127611

 

[thewas]

Preference IS subjective and has nothing to do with speaker quality (measured performace).

Another quote from Dr. Toole:

"Almost 50 years of double-blind listening tests have shown persuasively that listeners like loudspeakers with flat, smooth, anechoic on-axis and listening-window frequency responses."

 

[Geert]

In my view the methodology is flawed and consequently the results are void.

I still need to hear the first speaker that's not build to the rules of flat anechoic on-axis and smooth listening-window frequency response where I can't point out the flaws.

 

[tuga]

Another quote from Dr. Toole:

"Almost 50 years of double-blind listening tests have shown persuasively that listeners like loudspeakers with flat, smooth, anechoic on-axis and listening-window frequency responses."

I've pointed out countless times how I find the testing methodology flawed (single-speaker listening, inadequately positioning and toe, stereo listening without toe-in, etc.).
A lot of care was, rightly, put on eliminating bias but that made correct positioning impossible or impracticable.
Toole has posted here that they did the best they could.