GRIMM Audio LS1c & SB1 DSP Speaker Review

[Curvature]

For $25K I can get 5 X 8361

Maybe start with one and see if you like it.

 

[Curvature]

I am also a bit reserved when it comes to NFS results under very specific circumstances posing severe problems even under true anechoic conditions (such as active cardioid cancellation, very very big diaphragms or sound sources out-of-phase being far away from each other so at 1m I would not expect anything like an even wavefront). Whenever the d.i. is abruptly below 0dB, I take a closer look.

That is not the case here. Everything looks consistent. Ripples might be a side effect of lots of narrow-banded EQ corrections being applied based on a measurement at a different position or with different smoothing.

The robotic assembly vibrates. It is not clear to what extent that affects measurements, although I'm sure it introduces small resonances that affect FR. The way to check would be to look for correlated data across many speakers.

 

[Pearljam5000]

Maybe start with one and see if you like it.

I auditioned them and liked them

 

[Curvature]

I auditioned them and liked them

Doesn't count if you didn't buy

 

[Pearljam5000]

Doesn't count if you didn't buy

If I had the money I'd buy them
That's the closest you can get to actually buying them

 

[Littletycoon]

I don't know why you & some others in this thread have a problem pointing out flaws in the speaker, it's not perfect, and it's an expensive product - the bass is it's weakness, it's not a perfect speaker, and you could say the response is quite sawtooth from 350Hz - 2000Hz if you're being picky. So it's expensive and has a flaw or two, so why have a problem that people point out it's deficiencies. It's totally normal and not out of order to express perceived issues.

It's one of the best speakers measured on ASR, ever, one of very few. It is also sent in by the company who know pricing is exotic. What other company with such exotic priced speakers dared to send them to ASR for measurement, none!!!! .

And the main echo in the reactions is focus on the perceived weaknesses, even if there are virtually no examples on ASR that measure better. And then the kind of conclusion becomes that it must be a mediocre speaker at best. I don't find that normal. What if you could buy a set for less than €8K 2nd hand, would that change things?

 

[WaveCollapser]

For $25K I can get 5 X 8361

why would you want 5 x 8361 when you can buy 50 x Neumann KH 120 for the same price?

 

[Davey]

And the main echo in the reactions is focus on the perceived weaknesses, even if there are virtually no examples on ASR that measure better. And then the kind of conclusion becomes that it must be a mediocre speaker at best. I don't find that normal. What if you could buy a set for less than €8K 2nd hand, would that change things?

I don't see any of these for sale on the used market.....so that speaks to your point.
I'm not sure how many LS1 speakers have been sold. But seemingly the buyers like them and they're doing the job in the field. You can't argue with objective data like that.

 

[sceptical1]

I am also a bit reserved when it comes to NFS results under very specific circumstances posing severe problems even under true anechoic conditions (such as active cardioid cancellation, very very big diaphragms or sound sources out-of-phase being far away from each other so at 1m I would not expect anything like an even wavefront). Whenever the d.i. is abruptly below 0dB, I take a closer look.

That is not the case here. Everything looks consistent. Ripples might be a side effect of lots of narrow-banded EQ corrections being applied based on a measurement at a different position or with different smoothing.

Thanks for your contributions to the forum. Your input interests me, objective measurements and your interpretation of what it sounds like, especially.

If the ripple is low level and high Q who cares. It could even be related to the class D amp filtering depending on the resolution of the test that Amir uses.

 

[Pearljam5000]

why would you want 5 x 8361 when you can buy 50 x Neumann KH 120 for the same price?

They're too small

 

[Purité Audio]

I don't see any of these for sale on the used market.....so that speaks to your point.
I'm not sure how many LS1 speakers have been sold. But seemingly the buyers like them and they're doing the job in the field. You can't argue with objective data like that.

They weren’t an easy sell, customers by and large found the aesthetics ‘challenging’ and when the 8Cs were released well no comparison.
Keith

 

[beagleman]

It's one of the best speakers measured on ASR, ever, one of very few. It is also sent in by the company who know pricing is exotic. What other company with such exotic priced speakers dared to send them to ASR for measurement, none!!!! .

And the main echo in the reactions is focus on the perceived weaknesses, even if there are virtually no examples on ASR that measure better. And then the kind of conclusion becomes that it must be a mediocre speaker at best. I don't find that normal. What if you could buy a set for less than €8K 2nd hand, would that change things?

I think while for sure, some aspects of it are great, the lack of bass, or at least "measured" bass is mostly what a lot of us are curious about.

 

[Lars Risbo]

I assume you mean the original LS1 'whitepaper'. If so, given that one of the design requirements was to be as close to linear-phase as possible, then of course passive can not meet that requirement.

That said. and switching gears, the paper has always felt to me to be more of a marketing piece, than a technical paper.
i think while it makes very valid points about obvious FIR misuses, it also uses very overblown examples ('trainwrecks') to support their LS1 processing solution.
I question why present such overblown, bad-practice examples, other than for making splash.

Further, the technique in the paper of using an inverse-all pass via FIR to eliminate the phase rotation of a standard LR 24 crossover surely works.
But it is no different than using a linear-phase crossover to begin with, on top of the same IIR filtering needed either way for the drivers. A linear-phase crossover and reasonable driver corrections could of been used in "non-trainwreck" examples. (which would have made a much more valid basis for technical comparison to their electrical filter phase correction solution).

To the extent either one of these techniques doesn't have fully acoustic complementary response through crossover on and off-axis, they both have the same pre-ring potential.
No difference.
The paper makes it sound like the electrical filter linearization is superior, with no pre-ring potential ....(like is often feared with linear-phase crossovers lol.)
That's clever technical marketing, imo.

Pls don't get me wrong...I think the Grimm paper has some excellent and accurate info..
But the way the info is presented, seems to be about inflated techno-selling. Which seems to be central to the marketing and pricing strategy.

pre ring for phase linear is only an issue if you limit the bandwidth (high or low pass). If you only make the xovers phase linear then you can end up with a minimum phase band pass, eg from 20-20k. This is also reflected in the step response (sharp coherent rise and then slowly relaxing back to zero). This has no pre ring but is not phase linear around the band edges but phase linear in between. The big question is if there is an advantage to also phase correcting the high pass (eg20Hz corner) and if the pre ring causes issues. interesting if someone has experience here.

 

[Davey]

They weren’t an easy sell, customers by and large found the aesthetics ‘challenging’ and when the 8Cs were released well no comparison.
Keith

People want box speakers. Or speakers that look like box speakers. This is where the market is, unfortunately.

 

[bo_knows]

People want box speakers. Or speakers that look like box speakers. This is where the market is, unfortunately.

I would like for my speakers to look like two blades .

 

[Robbo99999]

It's one of the best speakers measured on ASR, ever, one of very few. It is also sent in by the company who know pricing is exotic. What other company with such exotic priced speakers dared to send them to ASR for measurement, none!!!! .

And the main echo in the reactions is focus on the perceived weaknesses, even if there are virtually no examples on ASR that measure better. And then the kind of conclusion becomes that it must be a mediocre speaker at best. I don't find that normal. What if you could buy a set for less than €8K 2nd hand, would that change things?

I don't really see it as one of the best speakers measured on ASR. The most impressive thing is the Horizontal Beam Width though. Now if it could do the bass better, then yes I'd say it's one of the best, but no. Yes, and overall you take the price into consideration, and would have voted it higher if it wasn't so expensive for instance.

 

[gnarly]

pre ring for phase linear is only an issue if you limit the bandwidth (high or low pass). If you only make the xovers phase linear then you can end up with a minimum phase band pass, eg from 20-20k. This is also reflected in the step response (sharp coherent rise and then slowly relaxing back to zero). This has no pre ring but is not phase linear around the band edges but phase linear in between. The big question is if there is an advantage to also phase correcting the high pass (eg20Hz corner) and if the pre ring causes issues. interesting if someone has experience here.

Yes sir, that's certainly been my understanding and experience, on both the linear-phase crossovers and linear-phase system high-pass or low pass.

I've made many tunings on 5-6 way MEH designs (including sub as a way) using complementary linear-phase crossovers with every order from 12dB/oct to 96dB/oct.
Can't say I've ever heard pre-ring unless phase linearizing system high pass.
My MEHs are well behaved on and off-axis, but not as well behaved as the LS1...I just can't imagine hearing any pre-ring on the 2-way LS1.

Also agree the big question about phase is whether to correct the system (sub) high pass. I've done a boat load of experimenting with this, mainly because I run large reflex subs at high SPL. As you know, reflex natural 4th order roll-off and the need for an electrical high-pass to prevent driver unloading, can make for some pretty rapidly rising group delay.

I've ended up splitting the difference, linearizing about half-the group delay away. But I sure can't claim it's the best technique...even outdoor listening is pretty tough to discern differences.

Question: Looking at the LS1+sub step response that Amir posted, it looks like you guys might be using a little bottom end phase linearization too.
Yes/ no? If so, can you comment on the degree?


Fwiw, the only time I think I've definitely hear pre-ring, is when using a large FIR filter across the full system bandwidth and linearizing the system high-pass. Global FIR correction I call it.
I can't say I've heard pre-ring for sure, when the sub is the only section with a linear high-pass.
I prefer using linear-phase crossovers directly on each multi-way section, to be able to isolate the sub's linear phase high-pass. (and also to be able to FIR filters of decreasing length, as each pass-band covers higher and high frequency.)

 

[KSTR]

The big question is if there is an advantage to also phase correcting the high pass (eg20Hz corner) and if the pre ring causes issues. interesting if someone has experience here.

I did this several times and when the corner frequency is low enough and the slope is shallow (2nd or 3rd order), it can work quite well, even though the improvement is not that audible.
When you correct for, say, a 6th order highpass from 4th order ported box with 2nd order subsonic filter tuned at ~40Hz this can give very audible and distracting artifacts. Like kick drums having a noise-like "fade-in attack" section before the "main event" which isn't sufficiently masked by this main event. While the improvement in "bass speed" is tremendous and working well for bass notes with a rather soft attack (eg plucked upright bass), the artifacts with step-like transients spoil the fun.

The noise-like (non-tonal) attack sound was sort of unexpected but my working theory is similar to what I feel is the sound perception for very low frequency sines: you don't hear the actual frequency as such, rather the alternating sound pressure seems to amplitude-modulate any background noises.

When reducing the phase rollback to about 3rd order equivalent high-pass phase I often got a good compromise for those bookshelf-type of speakers, faster than without phase rollback and still low tendency to produce artifacts.

Personally, I think the most important thing to correct in a speaker setup is the sub-to-woofer XO (often a LR4 @80Hz) followed by the woofer-to-mid XO. When the main speaker is a two-way as in the case of the LS1 the benefit from having a linear-phase XO at several kHz is not that big.
I found a rock solid phase coherence between ways at higher XO's -- like a stable 0degree offset in case of LR-types -- is more important than whether the whole thing is linear phase or allpass phase. If it is not coherent, imaging tends to suffer. Linear phase is then the icing on the cake. But for low frequency XOs I find it mandatory.

 

[gnarly]

Personally, I think the most important thing to correct in a speaker setup is the sub-to-woofer XO (often a LR4 @80Hz) followed by the woofer-to-mid XO. When the main speaker is a two-way as in the case of the LS1 the benefit from having a linear-phase XO at several kHz is not that big.
I found a rock solid phase coherence between ways at higher XO's -- like a stable 0degree offset in case of LR-types -- is more important than whether the whole thing is linear phase or allpass phase. If it is not coherent, imaging tends to suffer. Linear phase is then the icing on the cake. But for low frequency XOs I find it mandatory.

I agree with this. I think linear-phase phase alignments increase in importance as frequency lowers.

 

[Lars Risbo]

Agree too. the group delay is inverse to the xover frequency. The best setup would be sealed box woofers in a sealed room. Below the lowest room mode, this goes flat to DC in theory at least (air tight doors and windows ). EQing flat to 10Hz or below is possible. no need for phase linear when the minimum phase response is flat well below 20Hz