New panels - yet REW thinks they're rubbish!

[JohnA]

I've been running the SL3s with the new panels for a few weeks now, and they do sound nice.

However according to REW frequency response is complete garbage, especially from 10K onwards. we're talking over 10db down, like something is wrong with them. That's with 1/12 octave smoothing.

I had a Loki+ lying around, and I've ended up with the 8K band setting almost at full blast (+10db at least)
It is dropping less steeply according to REW with this setting.
Listening for a while everything feels normal, then bypassing the Loki it is immediately like someone wiped all the high frequencies away.
So yes, it has a lot to do with habituation, the brain adapts quickly.

Have others got similar experience with measuring frequency response of MLs?
Maybe it's normal, or maybe mine will get better with the passage of time..

 

[Gordon Gray]

Do you trust the measurements or your ears?

 

[RCHeliGuy]

Do you trust the measurements or your ears?

Measurements ALWAYS! To your brain EVERYTHING is relative for all of your senses.

The link below is to a very simple experiment that anyone can do at home by simply dipping your hands in hot cold and luke warm water.
Your hands will feel the exact same temperature water completely differently.

https://www.thenakedscientists.com/get-naked/experiments/strange-temperatures
If you live next to a paper mill, you eventually block that smell out.

This is exactly why so many companies want you to have a breaking in period for your brain to become accustomed to what you are hearing. And it works. You actually feel like it sounds better after you grow accustomed to it.

There are similar tests to that link to show how easy it is to fool your eyes.

Audio is no different. The brain adapts.

 

[JohnA]

..exactly

which is why I'd be interested in other people's REW measurements with the mic in their actual head position.
Maybe my results are normal, hard to tell otherwise.
From 10K upwards the drop is steep, at 20K 10db down roughly

Yes, a microphone cannot approximate two ears and a brain, but for a rough comparison it should do. The new panels weren't cheap, so I want to know that I got an objective improvement.
Maybe they need to break in a lot more?

 

[RCHeliGuy]

FWIW my Ethos were pretty flat which is likely strongly influenced by my room. The higher frequencies on average were close to my average centered at 0dB. I have some ringing which is expected, but the overall average had very little drop off. Maybe some over 15-16KHz but only a few dB and I doubt I could hear that. There wasn't much of a bass null to be found and I got better bass response than the speakers are rated for. I had to dial the bass back a bit on both sides.

 

[SomeGeoffGuy]

I looked up the original SL3 reviews on Stereophile's website, and the frequency plots show exactly what you are hearing. I kind of expected this, because my Aerius panels behave the exact same way, and they are from the same era.

I just bought my Aerius used and put them through the shower PDF procedure. I was amazed at how good they sound, but I could definitely tell they dropped off starting at 10k. I have them in my man cave, and had them wired to the "B" speaker output of the receiver, with "A" being a cheap pair of Dayton bookshelves. You could definitely hear the treble come in as I pushed the "A" button on and off, but the ML's sound so "Big", it doesn't really bother me.

-Geoff

 

[JohnA]

Of course, the Stereophile graph, I had forgotten about it.
Thanks Jeff!

Although I can't hear much over 13-14KHz myself, it's good to know that the response measurements I got are not all that weird

 

[Robert D]

I looked up the original SL3 reviews on Stereophile's website, and the frequency plots show exactly what you are hearing. I kind of expected this, because my Aerius panels behave the exact same way, and they are from the same era.

I just bought my Aerius used and put them through the shower PDF procedure. I was amazed at how good they sound, but I could definitely tell they dropped off starting at 10k. I have them in my man cave, and had them wired to the "B" speaker output of the receiver, with "A" being a cheap pair of Dayton bookshelves. You could definitely hear the treble come in as I pushed the "A" button on and off, but the ML's sound so "Big", it doesn't really bother me.

-Geoff

The only thing I discovered with my old Prodigy panels was that the amp had to work a lot harder to drive them to get the db sound level out of them. The pre amp was set to +12 db in order to be proper sound level. Now with the new panels, it is at +2 db.

Im hoping i didnt wear down my amp more than i should have by pushing it so much harder.

 

[SomeGeoffGuy]

Although I can't hear much over 13-14KHz myself, it's good to know that the response measurements I got are not all that weird

I hear you on the hearing - mine drops off at 14.5k. It's almost ironic - as soon as I can finally afford to get some nice stereo gear, old age starts re-possessing my hearing!

-Geoff

 

[JohnA]

OK, some graphs overlaid

The bottom is the old panels. Measured at lower SPL, but it is the shape we are interested in.
Same microphone, same location, same smoothing (1/24 oct), same power amp, same speaker location, same room acoustics
Displaying 250Hz - 20K, as Xover is 250Hz

Mic is aiming upwards in both cases. Would probably be different if aiming straight at a panel..

Top one is the new panels.

 

[Russr]

OK, some graphs overlaid

....

The two top ones are the new panels. Green without equalization, red with the Loki+ 8K set at slight boost (10 past if it were a clock)

Doesn't look like the old panels were that bad eh

Great info! And up to about 16KHz, not that much of a diff's. My dog (if I had one) would notice quite a difference with that roll-off >16K!

This thread has brought back some old "head them off at the pass" type battles I used to get into. It's about the measurements and graphs, no it's about the soul, no, it's about the $$$ I spent on cables and feet!

There are SOME things that are inviolate:

1) to gain (or lose) 3db of audio level output, requires a doubling (or halving) of the power

2) Due to tech advances, mass-production techniques and capabilities, etc., the most humble knock-off digital components available today rival the most esoteric, "Gold Standard" performance in digital domain components (ADCs, DACs, etc.) of years gone by.

3) Similar to "separation of church & state" (ha!), letting your Ears drive your decisions as opposed to the statistics/measurements, etc., is always something audioholics have struggled with.

My personal take on some areas of this hobby are probably seen as somewhat heretical:
I love the sound of my MLs, driven by my Bryston amps. Whether I'm watching/listening The 5th Element, or taking on some Mellencamp or Bach, I pipe it all through my Anthem AVMxx ("xx" depends which body part I have sold off recently to upgrade to a newer version of there pre/pros!). I refuse to believe that my wax-filled "handles" on the side of my head can tell the difference between a decent processor in Stereo mode v.s. a purpose-built, multi-$K stereo pre. Sorry, that's just me! My world is so imperfect, I cannot justify it, even tho' I can still hear "teenagers' (silent) ringtones!
When an idea strikes me as good (e.g. nice heavy plinths for speakers), I do it (65lb pavers from Home Depot).

Anyways, this is a hobby where personal approaches and tastes reign supreme... I enjoy the heck out of it, and I think everyone on this board does, too, whether of a similar mindset or diametrically opposed.

Y'all have a great Thanksgiving!

 

[JohnA]

Assuming at the higher SPL level the old panels would perform similarly, the main difference I see is the 12db dip from 7K to 12K has been replaced by two smaller 6db dips 8-9K and 10-12K.
3K-5K the new panels are probably a few db louder

Did I desperately need new panels?
Probably not, paranoia had the best of me. Equalisation would probably have done the job, let's be fair..

 

[spkrdctr]

The measurements are correct. Martin Logan speakers generally measure quite poorly. People buy them for other reasons, such as:
1. Very big sound especially if you are in the sweet spot for two channel. If you have surround sound then it is just fantastic big sound.
2. They can easily be made near perfect with EQ.
3. The speakers have "the look" of a Martin Logan which is unique.
4. Newer generations are not hard to drive. They sound good with most anything as an amp/receiver.
5. Lastly, and it can't be over emphasized, "The Look". I needed to say it twice!

If I ever have to buy new speakers (I doubt it) I will be buying a different brand of regular speakers as speakers have made huge advances in the last 5 years. New regular speakers be very flat and then it is just room issues. The MLs bring very uneven or low highs to the room before you even start. It is what it is. I like mine and to be honest there is very little sound over about 12Khz anyways. So, enjoy them!

 

[Gordon Gray]

Measurements ALWAYS! To your brain EVERYTHING is relative for all of your senses.

I didn't ask you.

 

[Gordon Gray]

The measurements are correct. Martin Logan speakers generally measure quite poorly.

That simply cannot be true. How dare you to not believe measurements. Just listen to RC, he knows everything.

 

[JonFo]

Top one is the new panels.

That actually looks pretty good.

Here is a metric I took of the brand new SL3 panel I used on my center channel project. It also starts dropping off at ~7KHz with the usual resistance pre-step-up transformer.
If we run it without any, and I mean zero, nothing between the amp and the step-up xformer, we see the cyan line, which actually has a resonance peak at 12KHz before diving back down.

Note: This is a panel driven by solely the diagram energizer, the step-up audio transformer and either 1 Ohm or zero resistance.
The low-end roll-off is the natural panel response, as there is no crossover either in this test.

that LF roll-off is why I XO at 340Hz with a steep slope. The highs kind of model the 'reference' curve from the room corrector, so not much, if any, boost is required.
I run this panel with no resistance, some as the Monolith panels, but that requires some seriously gutsy amps.

So the reason most ML ESLs have a sagging high-end FR is the resistance used to both tune the xformer response and present a bit friendlier impedance load to the amps.

 

[JohnA]

So Jon do you think the panels will perform better crossed over at 340Hz?

You're giving me ideas now

As for my obsession to keep the signal path to the panels 'clean and unmolested' that's gone gone gone...
I will soon be experimenting with surround-extraction from stereo (Surround Master v3 from Involve Audio), so the front signal will be manipulated by more than just analog eq, I'm talking an extra AD/DA step at 24/48 (hides in shame)

Electronic crossover doesn't sound half as bad now that I've seen the response of new panels

 

[JohnA]

...I run this panel with no resistance, some as the Monolith panels, but that requires some seriously gutsy amps.

The power amp drivng the panels can provide up to 1200 per channel at 2 Ohms, so miss daisy it ain't.
Every time I fire it up it's a gamble if the circuit breaker will kick in ha ha..

Still, would it be safe to get rid of the resistors?

 

[JonFo]

So Jon do you think the panels will perform better crossed over at 340Hz?

Definitely, nothing less than 315 for sure, and using a steep slope (24 dB / octave at least).
Also, remember the panel is a dipole radiator, subject to dipole cancellation the lower you go, so not much of a win to try and push it.

Still, would it be safe to get rid of the resistors?

I'd check with MF before trying that, as they'd know what kind of impedance and back-EMF the nuristor can handle.

 

[RCHeliGuy]

That simply cannot be true. How dare you to not believe measurements. Just listen to RC, he knows everything.

I'm so sorry. Apparently you know everything.

Pot meet kettle.