ATC SM75-150S Dome Mid Measurement Data

Just in case you weren't joking, the Fox he was referring to was the fellow ZGer, not the film/TV/news/media company of the same name...

Dunc

 

Well... they say the midrange is the most important bit to get right...

Dunc

 

I actually am interested in what is being discussed here, and I have a vague grasp of the technicalities - certainly I can see why it is being suggested that ATC should cross over lower down.

On the face of it those Dynaudio measurements look very good. What's your take on them ShinOBIWAN?

Also, I've got a pair of Jamo Concert 8s so I thought I'd look at the specs of some of the SEAS Excel drivers, and they don't look so hot. Any comments on them would be cool as well.

 

I had a brief demo of the Concert 8s and remember them sounding pretty excellent. Interestingly, ATC uses SEAS Excel tweeters on all of their speakers from the 20s on up. Funny that they would choose more drivers that appear to measure badly (where did you see that Antony?) Maybe they know something we don't?

Dunc

 

For the sake of completeness I've redone the THD measurements.

The initial measurements I posted were done with a laptop, onboard AC97 audio and a Behringer MIC100 mic preamp. The ADC/DAC's on this are dogshit and the noisefloor terrible, although its still within the realms of acceptable.

This time I did it with an EMU 1820m which has 120dB SNR DAC's and ADC's. I believe they're actually superior to the ones in the Lynx TWO B. If there's a weak link in the chain, its now not the sound interface.

Anyway the top end THD has flattened out, is much less erratic and lower end is a lttle smoother. Overall THD is slightly lower overall with an average of around 0.15% between 500hz and 10Khz measured using the same condition of 90dB/2m. You can also more clearly see the distortion rise at driver resonance.

The result has now been posted in the first post of this thread, you may need to hit refresh on your browser to clear the cache and show the updated image.

 

Cool, thanks for that. It's nice to be reminded why [anybody's] measurements have to be taken with a pinch of salt...

Dunc

 

Maybe I should have said..."Oh Boy, all this over THAT feckin midrange driver".

 

I'm not surprised that even given measured data the debate continues. Its a good driver and the best of its breed ie. 3" dome mid. What it isn't is the best midrange on the planet. It has a couple of significant problems and implementing a design using this speaker forces you to use steep slopes and more likely digital implementations - those are things that audiophiles HATE.

So its a driver that will always have a love it/hate it following. And the lovers and haters on here are particularly vocal so we see plenty of mud slinging and, I use the term loosely, debate.

 

For the sake of completeness, I wanted to include my recent experiences at a kit-loudspeaker manufacturers recently.

I heard most of their product range, which includes drive units by ATC, Volt, Seas and Scanspeak.

It was obvious to me, that despite using excellent drive units (and what appeared to be well made cabinets), that I had heard much finer examples of speakers using the same drive units by other manufacturers.

It shows how much more there is to loudspeaker design than simply putting figures into a computer based calculator to provide some kind of theoretical design.

I'm sure it is possible to make a DIY speaker which is of notable achievement, but the less gifted amateur is more likely (IMO) to produce something which is less than the sum of it's parts.

Knowing my own limitations, I believe experts are experts for a reason.

 

I can only assume your talking about Wilmslow audio

There are lesser DIY or kit outlets just as there are lesser manufacturers.

The point was that a designer can extract a more linear speaker from using measurements to characterise and identify the optimum operating ranges.

Of course making a given set of drivers within a specific enclosure as linear as possible isn't everyones cup of tea. BUT I'd have to dispute ATC's claim of low distortion and accuracy based on these and other measurements. Some people obviously like ATC's implementation but for me a good design is one that uses the drivers in their most linear regions for the benefit of accuracy.

Pseudo science and voodoo is why these and other audio forums are so contentious. If you like something that's not what the label says it is, where's the shame in that? But lets not stick are willies up the manufacturers arses based on reputation alone and the idea of 50 guys with PHd's running around slapping each other on the back.

 

The figures appear to be somewhat misleading as 90db at 2 m is presumably non anechoic and as such quite low - indeed some 10db lower than the average level I might listen at at 15db off peaks.

Properly carried out professional measurements will be 90 db @ 2m anechoic at a minimum and often 95db for more accurate devices.

 

Measurements taken outdoors on the field that backs onto our garden. Bottom of the satallite enclosure was ~80cm off the ground and the baffle tilts 16 degrees from vertical for time alignment.

Distance from the mid dome center to the mic capsule was 2m. SPL was measured at 90dB@mic using whitenoise generated by NCH Tone Generator.

 

I love my Concert 8s.

The 'measure badly' comment came from my un-expert quick look at the break up modes of the SEAS drivers compared to the ATC and Dynaudio. They looked a bit nastier to my untrained eye.

I don't know anything about loudspeaker design, but can follow technical discussions when they are explained like in the OPs discussion of the breakup modes of the ATC driver.

I was really just interested to know how someone in the know would rate ATC vs Dynaudio vs SEAS.

 

You can? I can't:

I agree totally with bottleneck. ATC have designed these drivers themselves from scratch. For Shinpads to claim that he knows far better than ATC how their drive units should be used, suggests to me that he is an optimist, to put it politely.

Exactly how reliable are these home-made measurements of Mr Shinpad? All scientific experiments have some sources of error. What are the margins of error here, and what are the sources of error?

 

When we talk about crossing the ATC lower down and/or with steep filters were talking about the lowpass of the XO, this is the part that concerns the mid-treble crossover and not the highpass which is the one low down. You really wouldn't want to cross the bottom end of the ATC mid any lower than ATC already do as distortion would become quite daft.

Just think of the highpass section as cutting low frequencies and passing the higher stuff and the lowpass similarly passes only lower frequencies and cuts the high ones.

Yeah we heard that from you when someone posted the stereophile ATC measurements late last year. More stick your fingers in your ears action again.

I've shared data with a fellow ATC fan in Australia and we both show near identical data sets. We both have differing measurement setup, with him using a B&K mic, connected to an M-Audio Audiophile USB and LspCAD software whilst I use a Behringer ECM8000 mic, EMU 1820m interface and ARTA/WinMLS software.

Anyway I'm quite sure if I'd shown measurements that proved ATC's implementation was a near perfect optimisation for the drivers characteristics then you wouldn't have even question accuracy, quite the opposite I suspect.

The measurements are easy to understand, I can't see what the problem is. Don't go flying off the rails thinking I'm saying ATC have made a bad speaker, not at all. What I'm saying is ATC aren't using the driver in its most linear region, infact its very non-linear outside of 4Khz and 500hz. So to me their claims of low distortion and accuracy aren't the whole truth. Even if they used the drivers in their optimum regions its still not what I'd call a linear speaker(very, very few are) and neither is the design that I've got, however its more linear than ATC's implementation so I can claim more accuracy. And accuracy and linearity are things that matters in studio monitoring which is ATC's primary market. People also talk of slight harshness with ATC speakers featuring the dome mid and I hear the same thing when I alter my implementation and start to use the driver in its more non-linear regions. It become all the more apparent when you AB the results in a direct comparison by switching between the two XO presets.

That's all I'll say on the matter because you have a real habbit of parroting the same lines over and over again and I'm starting to do the same in attempt to explain my reasoning.

 

Scientists and engineers (or doctors) are not perfect Bub. They have limited time, resources, budgets and imagination.

 

Scientists and engineers (or doctors) are not perfect

I know, and I'm glad you agree with me. That is precisely why I am asking a perfectly legitimate question about the margins (and sources) of error in Shinpad's experiments.

 

And shin also equally right to say that atcs work has no hallowed status.

 

No-one has suggested that it has "hallowed status", whatever that may mean.

OTOH, they conceived, and designed, and manufacture this driver in-house, so an educated guess might say that they know what they are doing with it, and Shinpads doesn't. Even more so, if he can't even validate his published data.

 

I have a pair of Wilmslow Audio Prestige 2 speakers with these Atc mids in and I love them to bits . I have heard JM Lab Altos and prefer these. I front them with Michel Orbe SME V Ortofon Jubilee. Trichord Delphini phono stage. Cary SLP 98 and Cary Aniversary Mono blocks. Dont think Ill ever change this system now