B&W CM1...all I can say is WOW! - Page 2

There's nothing that can make up for poor frequency response performance, though.

No argument there, just know whether you're buying high fidelity or not.


Totally false. When biases are taken away, people prefer as close to real as possible. That's why this hobby is called "HiFi", as in High Fidelity.

True, but without a flat frequency response, you know the speaker is not accurate. Accuracy is desired by some while others like a colored sound.

But frequency response is only one dimension of the sound accuracy. As well. few speakers measure anything close to flat in a real world environment, and I have a few TrueRTA graphs in a few typical listening room (mine and others) to prove it. I have never seen any FR graph convey things like the way you can hear different insruments clearly on different speakers, or the way the speaker portrays the positional cues, width and depth of the soundstage, the way the speaker renders the metallic sound of a cymbal or high hats convincingly

I used to be one of those people who was inexperienced, and went by looking at charts, reviews, and testimonials. As I got more experience with live instruments, and high quality recordings, I started to trust my own ears. No the CM1 isn't as technically accurate in the frequency response as others, but it has the right combination of attributes that make it sound pleasing (magic if you please) When I did A/B it with the NHT Classic 3, there were some things each did better than the other which you would never get by looking at their respective graphs. Using good quality reference recordings, both speakers had flaws that were clearly audible but it wasn't as if, "I can tell that Speaker A is more accurate in FR thus Speaker B is crap" Even the most famour speakers engineers that based their work around NRC measurements will say that listening is always the primary arbitrator of sound quality.

Personally, I could care less if people like markwriter think Axiom, Energy, Paradigm, Revel makes the best speakers because they engineer them to be flat, and that FR flatmess is a prerequisite to good sound. I've owned and still own many of them (lent them to family), and certainly some of them are competent (except the one or two Axiom which I find unlistenable) but I kniow from experience that sound quality goes FAR beyond what you see on a graph.

Oh yeah, I forgot that fact

Of course, I was careful to say that without a flat FR the speaker is not accurate, I didn't say anything that a flat FR is the only variable that determines accuracy.

I know from my own experience that poor measurements goes hand in hand with poor sound quality to my ears. That doesn't mean the audition process is rendered unnecessary, measurements helps to narrow down the candidates. I know I do not like speakers that have the mid-bass bump in FR around 100hz, that is valuable information for me to know. If published anechoic chamber FR shows that bump (beyond what would be caused by the quasi-anechoic measurement technique), I know not to waste my time auditioning.

In the end, the final choices are down to different compromises that the speaker engineers took. I let my ears decide if it is worth a slightly less linear FR for less cabinet resonance, or less distortion, or better dispersion, or more dynamics. However, without measurements, I'd be auditioning forever in a non-scientific guess and check approach which is exhaustive and not needed in this day and age.

No disagreement there, you have a method for figuring out what works for you. I have my own based on my technical knowledge (as an engineer), and many years of trying/buying different equipments. I find exaggerated sibiliance to be annoying and I really dislike speakers that have midrange peaks, a boxy or loose bass, and sometimes I can ascertain that by looking at the graphs, but in the end, a quick 5 minute listen with a few seconds of selected tracks can often eliminate those speakers, even if their FR graph don't point to anything obviously wrong.

A speaker's performance is the sum of many design considerations, and "wide bandwidth linear response" is just one of them, but that doesn't mean that there aren't other performance characteristics that don't affect the sound in profound ways (of which there may be not standard way of measuring them unless you have access to the design tools that the speaker designers use). I guess there are some things that CM is doing very well right even with an obviously flawed FR plot

Here is another take......

Without a reasonably flat FR, what ever the artist intended, will not be truly represented.

If a note or instrument is in the dip of an unbalanced FR(or peak/lump for that matter), that note or instrument will not have the emphasis the artist intended.

Oh lord, I pity the people that depend on FR graphs to determine the performance of a speaker. It's like trying to measure the enjoyment one gets from viewing art using a pie chart with form, color, and angle as slices. David Wilson of Wilson Audio put it best when talking about making adjustments to his speakers post-measurement: "Aw, you know this isn't what the ideal is, but darn, I prefer this other approach, which doesnt measure as well."

On a side note, markwriter, do you happen to work for Harman or one of their subsidiaries? It seems like every other post you makes just happens to gush about Harman's engineers and labs and R&D and so on. I hate to break it to you but with all of the money, development, R&D, and engineers in the world, if you set out to make a $300/pair vinyl-wrapped speaker, you will end up with a $300/pair vinyl-wrapped speaker. Now Revel's Ultima series is a different story...

You can look at it so many ways...there is no reason to take pity. Do you think your way is the absolute correct way?

Of course not but to see someone define a speaker as better solely because it measures better is reason to take pity on them.

If your opinion is that there is no way of judging a speaker on technical merits, then OK.

I suppose assuming people can imply ideas without having to have them literally explained is a bit much to ask for on the internet.

Do you judge wines based solely on year and vineyard? Do you judge cars by only looking at power and acceleration? Do these play a role in decision making and do they give good information? Of course. But to base your decisions solely on graphs, charts, and numbers is a bit absurd hence the ideas of wine tasting, test drives, in-home auditions, etc.

Yes, all things being equal, a flat frequency response is better compared to a speaker that has an more uneven frequency response. But the key word is "all things being equal". How much does a -5dB dip in frequency response related to the overall sound quality compared to a slight metallic resonance in the tweeter at 14Khz, or a dip in frequency response due to floor bounce (which a speaker would not exhibit in anechoic response). If a crossover design causes poor time/phase relationship, or a cabinet exhibits resonances, does that still mean the -5dB dip is still the primary detractor of sound?

These are rhetorical questions, and I'm not looking for specific answers, but all the variables interplay to form the resulting sound. The goal of flat frequence response is a design parameter but not every speaker designer agrees how flat it should be at the expense of other variables if any.

A person who evaluates speakers in a controlled fashion needs not your pity. However, your pity is not misplaced for the person who blindly accepts objective measurements without checking with his/her ears. To be able to quantify your subjective requirements is a very powerful tool in selecting what speakers to drive x hours to in order to audition. As I mentioned earlier, ignoring objective measurements and simply going to auditions is a waste of time and archaic trial and error at best.

I listened to the CM1 and while it does measure "poorly" I can see why some people would like it. Some people claim to like warmer laid-back balanced speakers. The CM1 tends to this crowd and this can be seen quite simply in its FR. The boost in the upper bass grants the warm balance and the absence in the 2 kHz range where our hearing is most sensitive allows for the laid-back sound. While the FR graphs did not exist at the time I auditioned the speakers, they do make quite a bit of sense compared to my auditioning notes of the speaker.

gchanjam -- As for wine and cars, these fields are completely for subjective purposes where the human pallet comes into play and is the only concern for some people. There are people who are more serious and do objective analysis of these things. There is no Utopian wine or car. However, whenever you want either of these to "perform" to a certain degree, this brings in the science. There are factors that allow a car to reach a certain benchmark. While there may not be enough measurements to fully define all of the factors, the existing measurement techniques can lay out minimum requirements that need to exist in order for a car to have a hope of reaching that benchmark. For wine, tannic structure, residual sugar (if applicable), amount of sulfites, are things that can be objectively measured and can explain tenancies in people's palettes. Objective chemical analysis can also explain why once vintage might be favored over another.

Same goes for speakers.

While the science may not be able to explain all, it most certainly helps for those inclined to learn and understand.

Warpdrive -- you aren't the only engineer.
The objective measurements are plenty enough to narrow the playing field when it comes to auditions. A speaker like the Zu Druid, for example, is not worth my time to even consider for audition. However, a speaker like the Thiel CS 3.7 is, and I have. Narrowing the playing field is, of course, only possible if you have auditioned enough to be able to cross-compare subjective and objective notes. For me, most of my subjective notes have paired up well with objective data. Of course, even this process is not very easy as the room certainly has its own contributions which need to be considered.

For someone starting out with this speaker game, it is probably most helpful to listen to as many speakers as possible and then looking at their objective data (if available) to find trends within the audition notes and the objective data.

I am not disagreeing with either of you.....just the notion of "taking pity" on someone. The guy even has studies to back up his posts.

Using studies to back up your arguments is also different from being able to use them to understand them in the big picture.

The earlier NRC research are the mantra of companies from the API group, PSB (albeit much less so nowadays), Axiom, Paradigm, and definitely the defects from the NRC who now spend time at Harman, and I've certainly enjoyed using their white papers (the subwoofer placement ones are the ones I read and utilized directly).

OK....I understand that and is perfectly fair. (I saw the "shake my head" part. )

Not saying I was.

Since you are an engineer, having some experience with basic principles as logarithmic scales, decibels, harmonics, allows one to put some sense of pespective on the objective data, which is often a downfall of some people who jump into this hobby without technical understanding.

I don't know about you, but I was always a tinkerer. I took a semester of acoustics and that little knowledge is a dangerous thing, my wife screams every time she hears some impulse signals or sine sweeps

edited out the wiseass parts

True, commonsense is only common to those who have the experience. FA Everest Master Handbook of Acoustics set off the craziness for me

Kevin Voecks of Revel depends quite a bit on FR graphs (actually it's a family of curves, all dealing with FR in different ways).

From "Sound Science" in Test and Measurement World:

"the Harman team learned that listener preferences correspond not with a single technical measurement such as total sound power but rather to a family of curves such as those indicated in Figure 1. Those curves represent on-axis response, the response averaged over a typical listening window, the early reflected sound response, the sound power, and directivity indices related to the sound power and early reflections."

True.


This is the type of point I was trying to make.

Is there something wrong with someone is making references to scientific studies on the AV Science forum?

When properly double-blind tested with speakers level-matched and positionally matched, people prefer accuracy the vast majority of the time.

You're comparing speaker evaluation to wine evaluation?

Some people care about what was "intended" and some people just want to listen to sound they way they like it.....both options are equally viable. Its all a simple and personal choice. Hence we have thousands of speakers to choose from and its also why we all have EQing abilities.

Of course its easier to start with a "flater" FR then EQ it in room as needed but we can EQ almost ANY speaker in room to actually create the sound we like, that option is something that few here seem to understand.

Sure! Speaker are truely nothing more then something like wine, those who think they are more and want to sleep with their speakers nightly are not normal people that should offer advice

Heck, you evalute speakers simply on a few names out there Can't think for yourself (one trick pony on here in every thread?). I couldn't care who built the speakers since Im not one of those "name brand" recognition people

If it sounds great then its a good speaker regardless of who built it or how much it cost.

If the guy on the first page finds some good quality DIY speakers for a cheap cost then he should see if he can listen to them in his house...they might sound incredible.

I have yet to hear any comercial speaker that can not be beat by DIY at a lower cost.

For those want to spend some time and have speakers that are better then speakers 3 to 4 times the cost, check out zaphaudio.com diyaudio.com or even our DIY forum on this site.

Its great to have links but that is all you do is rave about Revel or anything Harmon does....there are thousands of great options out there. I would rave about someone like Mark Seaton before the names you drop every time because Mark Seaton actually will post on here and answer questions.

If you were to use a given speakers Freq. response as a measure of accuracy, it would have to have the EXACT same in room response as the mixdown sessions equipment. Given that albums are mixed down in multiple facilities across the globe using multiples of speaker systems during mixdown playback, there's really no way of determining what is 'accurate' to the artists or engineers intentions. Todays engineers mix down for multitudes of playback environments with a new practice of 'loud' or 'high headroom' playback being used on a lot of pop and rock recordings.

Speakers with low distortion drivers and an intelligent crossover design that can cover the spectrum from 20hz-20khz will get you closer to a mixdown studios acoustics than a flat frequency response.