GRIMM Audio LS1c & SB1 DSP Speaker Review

[Mort]

Nothing, but it's been discussed. I guess it depends on how many circles one wants to go in.

There is good info in threads on ASR, but it tends to be limited to the first few pages and after that is quickly degrades. Tough to find the good info in between all of that. Be cool if there was a sort of tagging system here to mark posts as being sort "best of the thread" so people can quickly wade through the noise and get to the signal.

Like reddit? no thanks.

 

[Robbo99999]

10 pages of people talking about the bass, 20 pages of people saying they don't like how the speaker looks. Really riveting stuff we have going on here.

It's called a speaker review & then people talking about the speaker & what they think to it - it's called ASR.

 

[BeeKay]

then people talking about the speaker & what they think to it - it's called ASR.

 

[Robbo99999]

View attachment 454532

Ha, we're a bit more civilised than that, give us all some credit, ha!

 

[MKR]

View attachment 454532

Yup, that’s about right

 

[MKR]

Ha, we're a bit more civilised than that, give us all some credit, ha!

Mostly yes, but sadly not always … often things devolve into personal attacks, which is where it definitely crosses the line, I myself have been target of this. Folks act like toddlers around here sometimes.

 

[Pearljam5000]

Mostly yes, but sadly not always … often things devolve into personal attacks, which is where it definitely crosses the line, I myself have been target of this. Folks act like toddlers around here sometimes.

You are one of the nicest members on ASR so don't pay attention to that

 

[MKR]

You are one of the nicest members on ASR so don't pay attention to that

You are too kind, right back at ya PJ! Thanks to you we have some of the most interesting threads on ASR, you help keep it fun around here, and I thank you for that, truly.

 

[RickS]

Like reddit? no thanks.

But an AI filter tuned to your prefs might be useful. Might have to train the AI though…

Maybe then we could replace the moderators too!

 

[Pearljam5000]

You are too kind, right back at ya PJ! Thanks to you we have some of the most interesting threads on ASR, you help keep it fun around here, and I thank you for that, truly.

Now I'm blushing
Thank you very much
Much appreciated.

 

[Robbo99999]

Mostly yes, but sadly not always … often things devolve into personal attacks, which is where it definitely crosses the line, I myself have been target of this. Folks act like toddlers around here sometimes.

Some of it's fun, but there's limits and it can be distracting.

 

[richard12511]

Nothing, but it's been discussed. I guess it depends on how many circles one wants to go in.

There is good info in threads on ASR, but it tends to be limited to the first few pages and after that is quickly degrades. Tough to find the good info in between all of that. Be cool if there was a sort of tagging system here to mark posts as being sort "best of the thread" so people can quickly wade through the noise and get to the signal.

I disagree with your point about the discussion in this thread(I enjoy reading it, and hope it continues), but I like your idea of tagging. We have that already to a certain degree with the bookmark system. You can add labels, which act sorta like tags, since you can filter by labels in the bookmark search. You can also add a message to help remind yourself why you originally wanted to save the post.

 

[kma100]

Ha, we're a bit more civilised than that, give us all some credit, ha!

can't go wrong with Asterix!

 

[Pearljam5000]

For about the same price much bigger monitors

 

[sigbergaudio]

You can get much cheaper if you're just buying for the size of the speakers. From a sound quality perspective I'm pretty sure the Grimm is the preferred speaker over both the ATC and Focal.

 

[Eelco Grimm]

Hello everybody,

First of all a big thank you to Amir for doing all the measurements and writing such a wonderfully positive review of our LS1c. And thanks to all people on this forum who care so much for high quality audio that they dedicate many pages of comment to this review. We share your feeling of being driven to improve.

I noted two topics that confused some people and received a lot of attention in the comments. So herewith I like to shine my light on these.

1. The low end of the LS1c three-way system seems to roll off in Amir's measurement. In practice this is not the case, the fact that the subwoofer is on the floor causes this error in the measurement. A sub on the floor (“2pi space”) shows a different behavior than in anechoic conditions, effectively it offers a 6 dB higher output. We measured the LS1 in the TU Delft anechoic room during its development and we saw the same result: in two way mode it measured linear, in three-way it showed a low frequency roll-off. To properly calibrate the three-way system we had to measure it outside on an empty parking lot. When we developed the SB1 motion feedback sub we did some outside measurements again, the result can be seen in the attached graph (“buiten drieweg” is Dutch for “outside three-way”). As you can see the system has a flat response with -3 dB @ 20 Hz, the slight ripple in the curve is caused by some distant walls. Note that the SB1 resonance frequency is app. 15 Hz, for this measurement the -3 dB point is put at 20 Hz via DSP.

Since the Klippel NFS system is designed to derive a completely anechoic response from its scan, it shows a 6 dB lower output of the LS1 sub, just like in the anechoic chamber. That is why you see a rolloff below the crossover frequency of 70 Hz in Amir’s measurement. But as you can see from our outside measurements: when the sub is placed on the floor and the main speaker at normal height, the frequency response is straight. And this is the intended use case.

When placing the LS1 system in a room, the boundary effects of nearby surfaces of course need to be taken into account, as well as room modes. Since the sub is on the floor, positioning works slightly different than with a full range speaker with elevated woofer. Like with any speaker, to get a reasonably flat in-room response down to 20 Hz one needs quite some careful placement adjustments but since the sub can be taken off the footplate it allows for an independent optimization of the sub and main speaker positioning when needed. In my experience the linear free-field calibration with the sub in 2pi space usually leads to an optimal alignment of the sub and main levels in-room. Every time I touch the EQ or sub gain of the LS1 control software for a correction, after a while I feel that the balance is off and careful repositioning of the system solves a lack of certain bass frequencies much better. Some customers prefer a +1 or +2 dB nudge in the low end but a +6 dB gain on the sub is for sure totally out of balance.

2. Another point of discussion was the distortion shown in the “THD” graphs. It seemed exceedingly high for a subwoofer with motion feedback control, that attenuates distortion by 25 dB or more. We understand that Amir measured the 3-way system from a single mic position 33cm in front of the main cabinet. In other words, the distortion graph was not derived from the NFS measurement but from a single mic position (including room mode effects). It was then pasted into an NFS derived “in room” frequency response graph.

Because of this setup, the measurement shows too much contribution of the distortion of the main woofer since it is much closer to the mic than the SB1 subwoofer. Compared to the mic position results, at the listening seat the distance factor attenuation (20 log r, ie 6 dB softer per doubling of the distance) affects the main woofer a lot more than the sub. If the main woofer was at 33cm distance of the measurement mic, the sub was at 85cm. That means that at a listening distance of 3m the main woofer distortion will be attenuated by 19 dB and the sub’s distortion by 11 dB. And thus the absolute distortion as seen in the “86 dBSPL @ 1 meter” and “96 dBSPL @ 1 meter” graphs would be app 8 dB lower in the range below the crossover frequency of 70 Hz. For your reference, the measured distortion of the SB1 motion feedback sub was already posted by one of the readers (at the bottom of page 21).

Next, please note that in the two ‘THD Percentage’ graphs the distortion is plotted as a percentage of the (NFS derived) amplitude graph. That percentage is therefore affected by two extra factors: a) the floor boundary effect referred to in point 1 above, so the LF part of the amplitude graph is not drawn straight as it should have been. And b) the distance factor correction in the distortion plot. This means that at for instance 40 Hz the percentage plot needs a correction of 6 dB (for the amplitude curve) + 8 dB (for the distance factor) = 14 dB, which means the percentage needs to be divided by 5 (ie it is 0.8% in stead of 4%).

Since the SB1 has such unusually low distortion, it may need some getting used to when listening. Compared to loudspeakers with higher time domain and harmonic distortion it may seem as if there is less bass, until you discover the beauty of resonance free and undistorted bass. Acoustic instruments like upright bass or church organs form a great reference for this. It’s not for nothing that the SB1 designer, TU Delft professor Rob Munnig Schmidt, is a fanatic organ music lover.

In closing, I like to add that even though we measure a lot, it can only be the starting point. To obtain the best sound quality the question how you achieve your goals is of equal importance as what you aim for. For example, we developed a software tool that calculates and compares thousands variants of EQ settings that offer a similarly linear frequency response for the LS1. Its target is to find the most simple solution with the least amount of filter bands and the lowest Q’s etc. We discovered that a simpler solution almost always sounds better than a more complex one. We use this software during production; every single LS1 is measured and calibrated individually. Yes, I agree that this amount of attention to detail does increase the price of our products. But everybody in our team lives for achieving the best possible result, not for just finding a nice compromise. And that quality is what we love to offer our customers.

Eelco Grimm, for the Grimm Audio team

 

[Lars Risbo]

Hi Eelco, great that you explain here. Question: where is the mic placed in this buiten drieweg measurement? I am wondering if the usual floor bounce cancellation is avoided by having the sub so close to the floor (asphalt or concrete in this case ) whilst crossing above the cancellation frequency?

cheers

Lars

 

[Ilkless]

Cool to see this speaker reviewed here. I can certainly relate to Amir's opinions on it subjectively and do feel speakers with these macro aspects (large baffle with large round overs) are very much in a league of their of their own as stated in the review. A sort of clone of these is on my to do list and some fairly extensive experimentation has led me value many of the aspects the LS1 exhibits. I have yet to hear a speaker sound as good as my prototypes that have 3" radius round overs.

I've always been a proponent of large round overs, the larger the better, one simply needs to hear a speaker with effectively no edge diffraction down to 1khz and often lower to understand. There is a clarity that I do not hear in other speakers which is doubly apparent when the music tends to get busy, which makes a lot of sense (at least to me). The more busy and sustained the signal gets, the greater the benefit of the round over as the content starts to resemble something closer to noise which is a scenario that I find edge diffraction to be most apparent.

Here are some quick in room measurements of my last prototype, about 0-60 in room. I don't think you need me to tell you this speaker sounds incredible.

View attachment 454276

If I ever get around to making my LS1 clones, I'll be sure to share them here.

Congrats to Grimm on making an speaker I have no doubt is incredible, shame they priced it out of the hands of regular humans. At least they are not hard to DIY. I have 3d print of the necessary round overs that I may post up somewhere for others to use, and the cabinet construction is pretty well documented and very simple.

I have always thought many underestimate the impact of a wide baffle pushing down baffle step and allowing directivity control to lower frequencies -- without using brute-force resistive or destructive interference-based approaches that raise HD as a ratio of the fundamental -- on listening.

It's probably why speakers like the Devore O/96 still capture imaginations for all their horrendous crossover flaws -- simply because the tweako audiosphere rarely has exposure to speakers with lower baffle step. The reduced room contribution to the midrange likely yields an uncommon clarity there notwithstanding all the other linear and nonlinear distortion in something like the Devore.

And well, we see in this Grimm an all-out attack on this wide-baffle approach in domestic sizes.

 

[Pearljam5000]

Hello everybody,

First of all a big thank you to Amir for doing all the measurements and writing such a wonderfully positive review of our LS1c. And thanks to all people on this forum who care so much for high quality audio that they dedicate many pages of comment to this review. We share your feeling of being driven to improve.

I noted two topics that confused some people and received a lot of attention in the comments. So herewith I like to shine my light on these.

1. The low end of the LS1c three-way system seems to roll off in Amir's measurement. In practice this is not the case, the fact that the subwoofer is on the floor causes this error in the measurement. A sub on the floor (“2pi space”) shows a different behavior than in anechoic conditions, effectively it offers a 6 dB higher output. We measured the LS1 in the TU Delft anechoic room during its development and we saw the same result: in two way mode it measured linear, in three-way it showed a low frequency roll-off. To properly calibrate the three-way system we had to measure it outside on an empty parking lot. When we developed the SB1 motion feedback sub we did some outside measurements again, the result can be seen in the attached graph (“buiten drieweg” is Dutch for “outside three-way”). As you can see the system has a flat response with -3 dB @ 20 Hz, the slight ripple in the curve is caused by some distant walls. Note that the SB1 resonance frequency is app. 15 Hz, for this measurement the -3 dB point is put at 20 Hz via DSP.

Since the Klippel NFS system is designed to derive a completely anechoic response from its scan, it shows a 6 dB lower output of the LS1 sub, just like in the anechoic chamber. That is why you see a rolloff below the crossover frequency of 70 Hz in Amir’s measurement. But as you can see from our outside measurements: when the sub is placed on the floor and the main speaker at normal height, the frequency response is straight. And this is the intended use case.

When placing the LS1 system in a room, the boundary effects of nearby surfaces of course need to be taken into account, as well as room modes. Since the sub is on the floor, positioning works slightly different than with a full range speaker with elevated woofer. Like with any speaker, to get a reasonably flat in-room response down to 20 Hz one needs quite some careful placement adjustments but since the sub can be taken off the footplate it allows for an independent optimization of the sub and main speaker positioning when needed. In my experience the linear free-field calibration with the sub in 2pi space usually leads to an optimal alignment of the sub and main levels in-room. Every time I touch the EQ or sub gain of the LS1 control software for a correction, after a while I feel that the balance is off and careful repositioning of the system solves a lack of certain bass frequencies much better. Some customers prefer a +1 or +2 dB nudge in the low end but a +6 dB gain on the sub is for sure totally out of balance.

2. Another point of discussion was the distortion shown in the “THD” graphs. It seemed exceedingly high for a subwoofer with motion feedback control, that attenuates distortion by 25 dB or more. We understand that Amir measured the 3-way system from a single mic position 33cm in front of the main cabinet. In other words, the distortion graph was not derived from the NFS measurement but from a single mic position (including room mode effects). It was then pasted into an NFS derived “in room” frequency response graph.

Because of this setup, the measurement shows too much contribution of the distortion of the main woofer since it is much closer to the mic than the SB1 subwoofer. Compared to the mic position results, at the listening seat the distance factor attenuation (20 log r, ie 6 dB softer per doubling of the distance) affects the main woofer a lot more than the sub. If the main woofer was at 33cm distance of the measurement mic, the sub was at 85cm. That means that at a listening distance of 3m the main woofer distortion will be attenuated by 19 dB and the sub’s distortion by 11 dB. And thus the absolute distortion as seen in the “86 dBSPL @ 1 meter” and “96 dBSPL @ 1 meter” graphs would be app 8 dB lower in the range below the crossover frequency of 70 Hz. For your reference, the measured distortion of the SB1 motion feedback sub was already posted by one of the readers (at the bottom of page 21).

Next, please note that in the two ‘THD Percentage’ graphs the distortion is plotted as a percentage of the (NFS derived) amplitude graph. That percentage is therefore affected by two extra factors: a) the floor boundary effect referred to in point 1 above, so the LF part of the amplitude graph is not drawn straight as it should have been. And b) the distance factor correction in the distortion plot. This means that at for instance 40 Hz the percentage plot needs a correction of 6 dB (for the amplitude curve) + 8 dB (for the distance factor) = 14 dB, which means the percentage needs to be divided by 5 (ie it is 0.8% in stead of 4%).

Since the SB1 has such unusually low distortion, it may need some getting used to when listening. Compared to loudspeakers with higher time domain and harmonic distortion it may seem as if there is less bass, until you discover the beauty of resonance free and undistorted bass. Acoustic instruments like upright bass or church organs form a great reference for this. It’s not for nothing that the SB1 designer, TU Delft professor Rob Munnig Schmidt, is a fanatic organ music lover.

In closing, I like to add that even though we measure a lot, it can only be the starting point. To obtain the best sound quality the question how you achieve your goals is of equal importance as what you aim for. For example, we developed a software tool that calculates and compares thousands variants of EQ settings that offer a similarly linear frequency response for the LS1. Its target is to find the most simple solution with the least amount of filter bands and the lowest Q’s etc. We discovered that a simpler solution almost always sounds better than a more complex one. We use this software during production; every single LS1 is measured and calibrated individually. Yes, I agree that this amount of attention to detail does increase the price of our products. But everybody in our team lives for achieving the best possible result, not for just finding a nice compromise. And that quality is what we love to offer our customers.

Eelco Grimm, for the Grimm Audio team

Can you explain also about the price? Thanks

 

[Arindal]

a wide baffle pushing down baffle step and allowing directivity control to lower frequencies -- without using brute-force resistive or destructive interference-based approaches

Could you please explain what you mean by ´brute-force resistive´ directivity control? Horns, waveguides?

If I understand you correctly, I would intuitively file a wide baffle under ´brute-force resistive´ as well, as it is doing exactly that: forcing soundwaves into a certain three-dimensional dispersion pattern. It is just a flat one and the resulting pattern comes closer to half-spherical shape. Which is indeed an advantage in terms of evenness of dispersion over frequency bands, but can also be a disadvantage if you take the listening room, its side walls and resulting discrete reflections into account.

An interesting question evolving from that is how sufficient and even midrange directivity can be achieved at all, without using ´resistive´ or ´destructive´ approaches. In my understanding, this would be a call for large diaphragm area, no-baffle concepts such as line sources.