I can't say he is wrong because he's just relaying his subjective impression. On the other hand, it had some classic hallmark posting moments:
-I have xx years experience (another post, I know but same subject)
-I have heard almost every speaker
-I am surrounded by all this different equipment
-I work on SuperWhammy$$ speakers
-Asking JBL if the tweeter will break in...magically

I'm glad nobody had to pay $1.99 for that.

We also don't know if his comments hold merit with objective measurements. Many pro companies provide measurements and some do measure flatter on and off axis than the JBLs. We just don't know if he is just used to the ATCs or if they actually measure better in those frequency ranges. With no objective measurements we simply don't know.

John

Respectfully, thats crap.

The ATC SCM20 is pretty much dead flat On axis and very smooth off axis. It’s a neutral speaker.

FR does not reveal all. There is no info on how the speaker responds to dynamic signals.

FWIW I have 705p and SCM7 here and I prefer the SCM7 on my desk system for music. I prefer ATCs for music, and the JBL SCL series for movies.

Spins are a good start but they are not the be all end all.


Sent from my iPhone using Tapatalk

Do you have a spin for the ATC? The ATC does not have a waveguide. Doesn't that lead to directivity issues?

Exactly. We don't know.. I'd like to see those better measuring ones though :-)



I'd be very surprised if that were the case. None of the anechoic ATC measurements I've seen (Stereophile, German magazines) were exceptionally flat. They lack a waveguide to match dispersion, they're traditional square boxes (diffraction issues). Still solid designs, but they have inherent issues that have been resolved in more modern designs.

You mean compression? I don't believe any dome tweeter will keep up with the compression driver/waveguide combo in the 7 series.. just saying.

Stuff can get very weird on desks. I have a pair of Genelec 8020s on mine, which are good, quite neutral monitors. Yet I had to EQ the hell out of them on my desk, to get them to sound more-or-less good. There is no arguing with preference, but one can't argue the objective facts - and while spinoramas don't provide all the info, they provide enough info to make an educated guess about to expect. They are also taken at 2m - not exactly in the very nearfield. Which is something to consider.

There’s more to the sound of a speaker than just the frequency response, on axis, listening window and power response.

Otherwise why did the Salon beat out the M2 in John’s listening tests?

Spins do not characterize performance with varying music signals, only steady state. And there has been insufficient research done to tie measurements to sound quality in this specific area (Toole has agreed with this).

No, I do not know exactly what these measurement are either, but my current opinion is that the differences we hear between “neutral” measuring speakers are not related to the differences in their frequency responses but to how the speaker is tracking music with a multiplicity of overlayed components of all the instruments at low (reverb / recoding venue ambience), medium (harmonics) and high levels (fundamentals) all combined together.

Maybe something like the “large signal” Klippel type tests are required. Whether it is important or not, I know ATC make a big deal out of driver linearity, that is the ability to reproduce the incoming signal not only at low drive levels but also high.

John also hears differences I believe, in the “silkier” sound of the Salon vs 208, and the “wider” soundstage of the SCL-2 and 708i vs M2 and other sonic characteristics for which there appears to be no explanation in the frequency response.

I very much agree with the need to measure and characterize speakers, and agree that the frequency response on and off axis is a critical part. I do feel that in typical AVS fashion its become a reductionist argument though...”better the spins, better the speaker”. Spins are a great, great start and manufacturers following the principles will end up with neutral sounding speakers.


Sent from my iPhone using Tapatalk

No. It depends on the crossover frequency between mid and tweeter and how the mid driver is designed (is it pistonic or does the outer part “decouple” at higher frequencies).

Some reading: http://www.acousticfrontiers.com/spe...iver-diameter/


Sent from my iPhone using Tapatalk

Too late now of course. But I'd like to have seen the test done at several listening distances.
My assumption is that for two models as close as the M2 and the Salon2, there will correlation between the listening distance, room characteristics and the directivity behaviour of the loudspeaker with the Salon2 having the edge at regular listening distances and the M2 closing in and possibly surpassing it as distance increases.

The large 15" driver on the M2 that plays well up into the midrange is also much closer to the floor, which can-maybe-perhaps cause some anomalies in the response depending on the listening distance. The small 4" midrange on the Revel won't suffer from these effects.

I don't have the capabilities, but I believe with some research it could be determined why the Salon2 beat out the M2 in no uncertain terms - and I think the above will play a part in it.

Some that would compete very well for on/off axis. The SpinOrama is a great measurement especially for a quick glance but it doesn't tell everything about a speaker. But for consumer comparison it is a great tool. You are correct a waveguide is about matching dispersion at the crossover frequency and this usually shows off axis. On axis is easier to match. An active crossover has a lot more flexibility and precision that a passive design simply can't do. As a generalization I still like to see a waveguide even a shallow one but with the right drivers and carefully matched crossover you can get good 45+ degree off axis without a waveguide.

Here are some examples from the companies own literature. These would do very well on the spinOrama and that 8260 would beat the M2 but not in output. And I included the 1032A as you can see it is a 25 year old design so making accurate speakers is not new concept.

There are a lot of companies that make accurate speakers and do research. Floyd quotes many references to research in his books that did not come from Harman. Sometimes it gets a little thick here with it making it seem like Harman is the only company that makes accurate speakers and does research. That is simply not true. I do applaud Harman for making some of their research available to the public. Other companies I have seen share their research too but only to professional like through AES which you need a paid membership to view and not directly available to the public.

I do give Kudos to Harman for making great products and making research available to everyone!

Attached Thumbnails

         

     

Most measurements are done at 2m and the reason being with larger speakers like the M2 the sound between the drivers would not be integrated at 1 meter. The M2s response would not be good at 1m because you would not see the combined response of the drivers. You can measure at 1m for smaller speakers but most labs usually do it at 2m as a practice and then extrapolate the data for 1m for things like sensitivity and output.

Do you have data to support your off axis claim? To me the SCM20 sounds like a typical 7" 2way, with that annoying upper mid-range bloom that comes from poor directivity matching. Nothing special at all. Monitor Audio offers a similar sounding speaker for less with more finish options...


You don't think that the obvious differences here are treble directivity and possibly treble resonances?

The SCL and 7er both have wider treble directivity than M2. Simple geometry: smaller diaphragm firing through a smaller outlet. So does Salon, I believe.

The marketing propaganda for Be generally is reduction of resonances within the audio band. The BMS or BMS-style plastic rings JBL uses do have some treble resonances.

I do agree with you that more work is needed on large signal performance and correlation thereof to preference

The SCM20 has a 4.9” driver of you exclude the surround.

The difference between the on axis and 45 degree off axis at 1k is 1.5dB, 0.8dB at 2k, 1.1dB at 4k. Crossover frequency is 2.5k.

Despite the waveguide you can see a small change in the directivity of the 708i around the 1.9k crossover: it’s not perfect either.

I have measurements but the are not taken anechoically. Outside off the ground but still affected by reflections. If I had a 708p I could take comparative measurements of frequency response. I do not have a setup to do Harman style high resolution spins.

I was calling John out because he made assumptions about how ATC measures with no data.


Sent from my iPhone using Tapatalk

Just like every other "7-inch" 2-way with the tweeter on a 180deg waveguide...

Again I've heard the speaker and it doesn't offer anything above what, say, and old NHT 1.5 offers.

A 1dB mid-range hump at 45 deg off axis in the ear's most sensitive region is audible in most rooms. IMO the only setups for which a speaker with that kind of sound power problem are unproblematic are a very wide room or nearfield, such that the speaker-sidewall distance much greater than the speaker-listener distance. Even then I'd prefer a higher performance speaker, if only for the placement flexibility in another room.

Agreed that the 708 could use a lower crossover. 1.9kHz is a little high for an 8" woofer and a 100+ deg WG. In biamp mode is the xover a little lower? Asking because I don't know.

That's funny - I actually like a lot of the ones you've mentioned, and I've seen the data. I'd scratch the Dynaudio LYD off there (measurements are also smoothed on there), as well as the KH310 from personal (although sighted) listening. The 120 and 420 on the other hand were very nice. Genelec 8020 (bought em) and 8050 I liked as well - haven't heard the coaxials or the 1032.

The 708i's impressed me a lot though - enough to get over the fact that I actually needed white speakers for the living room. The 8050 would have been the alternative.

But, one thing we can agree on, they're all a lot more accurate and neutral than MANY older (and some recent) examples still seen in studies today.

It's 1,7khz - the in-wall series are crossed at 1.3khz.

The 708 also has a hump in the 1.5 to 2.5 kHz range, which looks like 2dB difference between on axis and first reflections. Does that classify as a "sound power problem" to you? It certainly doesn't to me, and neither does a less than 0.7dB difference 0 vs 45 off axis in the crossover region.

The original post referred to someone comparing the ATC SCM20 to a 708i in an environment which likely was nearfield. Not that it matters, since the ATC doesn't have "sound power problems".

I believe recent JBL/Revel/.. products aim for a flat listening window and not necessarily the flattest on axis response. Ideally of course, they are as close as possible to one another.

Again, we're splitting hairs here - a lot examples have been presented of well measuring loudspeakers. Maybe we'd ought to ask ATC if they'd be willing to send out some data?

A new review. While not technical in nature, it's a fun read (and being super flattering it's also a lot more accurate than the earlier SOS review ):

https://www.cinemasound.com/review-j...ence-monitors/

Fun, yes - but this just goes to show that many people don't understand the science - positive or not.

Someone ought to tell the reviewer that's not how any of that works.

Is the crossover at like 1kHz? If not, geometry dictates it has the same "mid-range mushroom cloud" polar one sees from other obsolete 2-way speakers such as Ascend, Monitor Audio, etc.

In listening, it HAS the upper mid-range colorations typical of HiFi speakers with the obsolete 7" woofer/180deg tweeter waveguide configuration. I've heard them. The only thing special about them is the brand name.

Nyal - while I don't think I made any claims as to how the ATC measures, I can see why how I worded things could be construed that way. I bolded a few things and changed a couple of words:

A/Bing them to the ATCs it was apparent that ATC rules the mid range.

So, we have a speaker we know is accurate - the LSR708i - and a speaker that the poster is claiming has a richer midrange. The answer could be right there - from what he / she is reporting, the ATC has a warmer midrange and a rolled off treble. No surprise that someone mixing on ATCs would produce recordings with a pronounced top end, that would sound shrill and sibilant on a neutral monitor. Of course, part of this is conjecture, as we have no measurements of the ATC to look at...

_____________

Better?

Still, if you read my original post with my original wording, I said very clearly: Without any measurements on the ATCs we have no idea if they are accurate or not.

That's hardly me "making assumptions about how ATC measures with no data." Per the longer quote above, I was going only by the claims in the "review," and drawing tentative conclusions from there. Agreed that my wording here was not as clear, so made the changes you see above.

Still - I think the original review is full of fallacies and "Circle of Confusion" issues. The ATC speakers could be made by God himself, I don't know. My issue was with how the conclusions were reached, and how they were stated:

The Bad:

Filtered sounding,

What does this mean...?

The mids are not open and the highs are very harsh.

How does the poster know this is not a problem inherent in the recording? (For what it's worth, I have also found speakers like the 708 and M2 a touch more "harsh" on instruments like violins, which you pointed out in a different post, which I will get to later.)

They are sibilant. I wanted to extensively test this on VO recordings I have done for TV. All different VOs. Ones that I've listened to on many different systems. And the JBLs are not accurate in this frequency area.The horn compression tweeter is really intense and will not let up.

So now I just throw my own two cents in here, from experience. I have the following monitor speakers on hand: JBL LSR708i, JBL LSR305, and Mackie HR824Mkiis. I have mixed eight independent feature films, the last two (The Creep Behind the Camera and Halcyon) in surround. I have extensive experience with dialogue editing, and know exactly what sibilance sounds like. There are all kinds of tools for reducing it or removing during mixing and mastering. Almost always, it simply means an emphasis on frequencies between 5 and 8 khz. As anyone can see by looking at the Spinorama, the 708 is exceptionally flat in this area. If there is obvious sibilance, my strong bet is that it is in the recording, NOT the speaker.

FWIW, I had the privilege of my mix for The Creep Behind the Camera premiered at the TCL Chinese in Hollywood, which had just received a makeover. Happy to say my mix translated very well One can also Google reviews of the Blu-ray to see what others thought of my mix (and score, for that matter).

I actually called JBL to see if the tweeter would soften over time. Nope.

Silliness...

I messed with the parametric EQ but couldn't get the speaker to sit more evenly like the ATC just does straight up

Again, Circle of Confusion issue big time. How does the "reviewer" know that the ATC is more "straight up" than the JBL? It might be, but without measurements, who knows? The big fallacy here is evaluating a recording mastered on one set of speakers against the same recording on another set from a different manufacturer, then being surprised when it sounds more natural on the speaker it was mastered on. Once again, how does the reviewer know that the problem was not in the recording?

A/Bing them to the ATCs it was apparent that ATC rules the mid range. Nothing can beat them. The 20s with a sub are just a dream.The JBLs don't really need a sub but for post their low end is not tight enough. I cannot recommend them for post. Maybe electronic or dance music. But don't put them next to ATCS!!

"Nothing can beat them." "Don't put them next to ATCs!!" I certainly make room for the possibility that the ATCs could be better than the JBLs - but of course in a sighted comparison - when someone has a stake in the outcome - I am extremely skeptical. As someone has every right to be of me.

FWIW, I still have the acoustically transparent barrier my operations manager Joel created for the M2 / Salon2 speaker shootout - we can always trot it out any time for a double blind comparison. Let's not forget that 86% correlation between the Spins and listener preference, which is almost 100% for bookshelf speakers.

One of the things I'd love to do is bring that to a high end show, and let anyone bring their speaker of choice in to shoot out against a pair of Revel F208s.

Agree with much, but not all. Some thoughts:

Salon2 vs. M2 - remember, it was a pretty close thing. Discussions with those who know more than me resulted in the following, tentative conclusions:

The M2 design has more direct vs. reflected sound, which can result in sound perceived as less "natural." However, many mixers prefer more direct sound which they feel isolates problems more effectively. Conversely, a speaker with more direct sound can sound more natural with certain, close mic'd instruments such as saxophone. As was shared with me:

Yes, a sax, like all horns, is very directional. If your expectation is to hear a sax in your living room, a narrow dispersion loudspeaker will be appropriate for a close miked recording of it - in a single channel; hard panned. However, it may be a less good match for other instruments or recorded sounds, especially if panned - using both channels. If your expectation is to hear something like the mastering engineer heard, wide dispersion loudspeakers are likely to be more appropriate.


It may still be a phenomenon that the ear can localize sounds at the throat of the horn, which can give it a "boxed in" sound to a certain degree. This can also explain why I liked certain recordings better on the M2s than the Salon2s, and why almost all of those recordings involved close mic'd brass or woodwinds. Those instruments bring forth sound in a manner similar to a horn speaker, which makes sense.

Both of those explanations help explain the differences we heard, and at least partially correlate with the Spins...

Lastly, we have no spins on the SCL series yet, so can't draw any conclusions re: what I've heard vs. measurements. And some of the impressions I've posted were done totally sighted, so I'm skeptical of my own findings...

Spins may not reveal everything we need to know, but they sure seem to reveal most of it - hence the 86% - 95% + correlation during controlled listening tests...

Others may make the best speakers on the planet and put Revel and JBL models to shame. But what impresses me about Harman - besides doing the research AND publishing it in scientific journals - is that they also go through the extra step of validating the results in the double blind tests at the MLL. As I think you know, we've tried (on a very limited budget) to try and replicate a "human-powered" MLL here We have a standing offer for anyone to bring in any speaker to be shot-out double blind using what we have here.

Of course, someone could come in tomorrow with a speaker that blows away our demo models. My thought is - let the chips fall where they may

Believe me Harman is not the only company doing double blind tests and other listening tests. They might be the only one with the cool quick speaker changer that gets the not used speaker totally out of the way I know of one on a rotating platform that puts both speakers back to back with absorption between on the platform and it just rotates 180 degrees. Of course it is not as ideal as Harman's but does do quick changes. And I'm not sure if the platform is big enough to do 3 or 4 speakers and rotate 90 or 120 degrees. I think they only compare two at a time. But, that company doesn't make their testing public but does submit papers to AES. I give Kudos to Harman for making their material available to everyone. I think a lot of companies would be afraid to do that.

Interesting. PM'ing you for more. Can guess at a couple

Mr. DS-21...

You are of course right that dispersion is related to driver width. I wrote a whole blog post about it: http://www.acousticfrontiers.com/spe...iver-diameter/

First, an issue is that a 6.5" driver is not actually 6.5" in terms of the width of the radiating element. The ATC SCM20 driver is 4.9"

Second, the assumption in the math is that driver behaves as a piston. Pistonic drivers (e.g. metal, ceramic, etc) will have worse off axis than a non-pistonic driver at the top of their pass band. It is possible to engineer a driver such that the outer part of the cone "decouples" at higher frequencies, such that the radiating area is less, and hence off axis is better. The SCM20 driver has multiple components that provide such a function - it's easy to see from looking at the driver that the inner and outer pieces are made of different materials. As a result the SCM20 has wider dispersion at the top of it's pass band than a equivalent diameter metal cone from Ascend or Monitor Audio.

Of course, a competently designed speaker with tweeter on waveguide should have better off axis, particularly at extreme off axis angles, relative to a non-waveguided speaker.

But, there is much, much more to how a speaker sounds than just speaker directivity. Once we are in the "neutral" ballpark, like we are with ATC (and many other good speakers such as KEF, PSB, Revel, etc), then other factors reveal their importance. To name but a few: resolution of low level signals, low levels or no power compression when used at the SPLs the speaker is designed for, inaudible distortion / cone breakup, inaudible box colorations, inaudible port operation. Perhaps that is why people are reporting favoring one speaker over another?

Of course it's possible to design a well-behaved loudspeaker without a waveguide. It's just easier and more cost efficient to do so. As for inaudible effects contributing to lower preference ratings. Wouldn't that make them.. audible?
I believe the main reason for people defending their speakers, is more ego than antyhing. No one wants to believe they did not make the best choice possible, not for that amount of money anyway. That goes especially for industry folks, whose job relies on their ability to hear into the mix. Defeating their ability to pick the best set of speakers reflects badly on them. It's not all black and white of course - loudspeakers have become so much better over the past decade than anything that came before. The more companies invest in scientific research, the better for us in the end.

I've seen this term used several times over the years but I don't think I've ever seen a measurement demonstrating "low level resolution" differences between loudspeakers. It kinda goes up to the line where I'd judge it to be magical thinking. I just wonder if it's misnomer used to describe perceptual differences in previously unheard ambiance cues that are caused to happen under the right speaker radiation and room reflection conditions.

I've sent an inquiry to ATC asking for some detailed technical data on their loudspeakers. Hopefully I'll get a positive reply.

Audible yes, but I'm pretty sure they don't show up on spins.