GRIMM Audio LS1c & SB1 DSP Speaker Review

[Davey]

I know it's largely semantics, but I agree with Davey's next sentence:
"The measured system response (at typical distance) is modified by the room. (That is quite a different thing.)"

I think Amir understands this. (Thank you for reading my post closely.)

Nevertheless, the "Estimated In-Room Response" plot does not reflect the approximate plus 6db you'll get in the bass area when the woofer is put on the floor for normal usage.
And, indeed this could be the case for some other speaker measurement types as well.
This is well pointed out by @bmc0.
I hope it's understood by other readers trying to reconcile the sloped off measurement at the bottom of the Grimm response.

 

[DearSX]

Wow, great to see expensive speakers tested and delivering something great/amazing! Price is up there, but at least its not junk or at least as good as other more affordable speakers. @amirm thanks!

Would love to see a Magico/YG/Wilson tower tested.

 

[capslock]

Nein. Both Amir and Erin use the same Microtech Gefell MK225 pre-polarised mic capsule. The NFS doesn't come with a microphone, you have to supply your own, or buy one they recommend. Now the sales engineers at Klippel will suggest an industry standard 3Hz to 20KHz 50mV/Pa because it's a) low noise and b)measures up the standard audible range 20KHz. The G.R.A.S. or Microtech Gefell they have on stock; and the very good but more affordable MtG is usually recommended.

This is a pre-polarized electret condenser 1/2" mic capsule, and here is manufacturer's datasheet for it: capsule
Klippel specify it's Max. SPL before clipping as "135dB, 1% clipping point" after attachment to the pre-amp (eg. MV210- IEPE; or MV220- 48V phantom power). If 146dB is the 3% (-30dB) clipping point then it follows then it follows that
The -80dB distortion point of the microphone is ~50dB lower. i.e. 146-50 = 96dB.
The -70dB distortion point of the microphone is ~40dB lower i.e. 146-40 = 106dB

In short, if one wants to characterise a transducer who's true H2 is between -70dB to -60dB (error margin +/0 0.5dB), then the observation limit for the Microtech Gefell MK255 is 96dB to 106dB.

Which makes it fine at 1m for distortion measurement in the midrange and treble frequencies. But the swept sine method to measure distortion relies on accurate measurement of the frequency response, and so it's very hard to characterise bass distortion properly when measuring at 1m non-anechoic. This is why the bass distortion appears a lot higher than it truly is, compared to measuring bass harmonics with a nearfield measurement. But of course, that would subject the mic is huge SPLs near it's clipping point.

Erin's NFS doesn't support 48V phantom power, but if Amir's does, then he can upgrade to the MK202 with the MV225 pre-amp.

So as it stands, the SB1's Digital Motion Feedback's effectiveness (or lack thereof) cannot be assessed as it is presently done.

Anyway, thanks for the review @amirm . Interesting, as always, what the price of a car can buy...

Then you have done more research than I have I didn't realize Amir and Erin use the same capsule. The MK225 seems to be the same thing as B&K 4155 which is a prepolarized version of the good old 4165 (I know, the product numbers are different these days, but I remember the old ones). The distortion spec of the capsule is exactly the same.

You raise a valid point that constant current preamps always distort more than those with a separate supply. I am not even sure 48 V is the end of it. The older B&K preamps use 120 V, and given that the amp is essentially a bootstrapped follower, the higher the operating voltage, the lower the distortion. Then there is also the question of input impedance. For the classic B&K preamps with polarization voltage, it was on the order of 50 MOhm (I have a schematic somewhere, would need to look it up). That will also result in some HD2. Not sure how the preamps without polarization fare there.

I am not familiar with how the NFS measures distortion. Does it keep the microphone at 1 m? How can you then know what is the fundamental of the driver and what is room gain? Is that what you mean by needing to know the frequency response?

Anyway, I am more interested in midrange. I did a couple of comparisons of speakers that have the same drivers and similar XO and found that within Amir's measurements, there is more consitency than within Erin's, and when similar speakers were measured by Amir and Erin, Amir tended to measure lower midrange distortion. I tried to do the same for Nuyes' data and got the impression he was in between but had to few pairs to really confirm that.

 

[Matias]

So cool to also have high end price and high end performance speaker reviews. I wish more manufacturers would bother lending review units.

 

[Hayabusa]

I almost don;t dare to say it...
But it looks like the measurements were done without the sub.....

[EDIT: reading back the whole thread I see my concerns are already covered.. something is wrong here, must be settings I guess]

This is the measurement from stereophile:link

They also speak about a 2 way mode of LS1c and the high pass filter of the sub, 20Hz/40Hz.....

 

[olieb]

Nevertheless, the "Estimated In-Room Response" plot does not reflect the approximate plus 6db you'll get in the bass area when the woofer is put on the floor for normal usage.

In respect to which case do you think that a placement on the floor will give "plus 6dB"?
At 50 Hz you have a wavelength of 7m and a placement of 60cm above the floor will probably give you at least 4-5dB "gain" from floor too. And this does not even take into account the other 5 walls that will give (a rather complex) gain too.
Basically in this frequency range the speaker/sub is not much more than the "motor" and the room itself is the speaker. The situation is more like sitting in the speaker than anything else.
The only (somewhat) meaningful measurement that can be done generically is the "motor measurement" (anechoically) without (random) inclusion of walls. The final adjustment of levels (and FR) then has to be done in/with the room.

 

[FTB]

40k $ for no bass ?

 

[AudioSceptic]

I think Amir understands this. (Thank you for reading my post closely.)

Nevertheless, the "Estimated In-Room Response" plot does not reflect the approximate plus 6db you'll get in the bass area when the woofer is put on the floor for normal usage.
And, indeed this could be the case for some other speaker measurement types as well.
This is well pointed out by @bmc0.
I hope it's understood by other readers trying to reconcile the sloped off measurement at the bottom of the Grimm response.

The usual 3-way speaker, where the bass driver just *is* under the mid and tweeter, is one thing, but with separate sub(s), which can be placed anywhere, where do you even start? With 2 subs, does near or under the mid and tweeter end up being best? How much does the cross-over frequency matter?

 

[AudioSceptic]

I almost don;t dare to say it...
But it looks like the measurements were done without the sub.....

This is the measurement from stereophile:link

View attachment 452224

They also speak about a 2 way mode of LS1c and the high pass filter of the sub, 20Hz/40Hz.....

You didn't read post 8, did you? ;-)

 

[Hayabusa]

You didn't read post 8, did you? ;-)

only 5 mins ago,
you did not see my edit did you?

 

[Davey]

The only (somewhat) meaningful measurement that can be done generically is the "motor measurement" (anechoically) without (random) inclusion of walls.

Which is what the NFS does.

My original statement (that got lost in all the hand-waving that followed) was "the room does not modify the bass response of any speaker."
I think some missed the basic premise of that. But, of course, it's a clearly correct statement.

If the room is big and has only one boundary (ie, the floor) you'd get +6db when moving the speaker from full-space to half-space.
I used the word "approximate" for a reason.

 

[AudioSceptic]

only 5 mins ago,
you did not see my edit did you?

OK, no pack drill then...

 

[Hayabusa]

BTW, compare my measurements to that of stereophile:

They really need to correct for their near-field response error (the bump). It makes speakers like LS1c look bad for no reason.

I just learned this sub has DSP high-pass filter setting, What high-pass filter setting did you setup for the sub in this case?

 

[Pearljam5000]

I wonder if the LS1be
Would have sounded/measured better
I have a weakness for berillyum

 

[Andreas007]

I just learned this sub has DSP high-pass filter setting, What high-pass filter setting did you setup for the sub in this case?

I also can't imagine that this speaker will not match it's spec (30 Hz +/- 0.5 dB).

 

[beefkabob]

The performance below 100 hz, in volume and in distortion, kills otherwise perfect speakers. Triple the volume of that sub, port, and then match them, and you'll have appropriate performance. Even then, it will be overpriced, but at least it will be perfect.

 

[Puddingbuks]

What a speaker!

Nice design also.

 

[feitaishi]

why not add another sub on the top of this system(LS1c + SB1) and cross around 40hz?

 

[beefkabob]

why not add another sub on the top of this system(LS1c + SB1) and cross around 40hz?

If you're paying car prices for speakers, you want them to get it right all the way to 15 hz. YOu don't want to have to do it yourself.

 

[Pearljam5000]

What's the difference between LS1C and this model?