Rythmik L12 Subwoofer Review

[amirm]

@amirm I was wondering whether the Klippel distortion report for a subwoofer can be usefully compared to the Klippel distortion report for other typical (full range) speakers?

More specifically, in my testing, I would not expect to be able to compare the distortion values unless the test were done under the same conditions and using the same type of test. Does this still hold true or not?

I am using identical protocol for both. So yes, you can compare the two. Where it becomes tricky is that I did not match SPL levels.

One thing Klippel NFS can do by the way is to get rid of room modes. The modes are liable to screw up distortion measurements.

 

[amirm]

So something like this:

https://www.amazon.com/Champion-880...ds=electric+garage+lift&qid=1584835078&sr=8-8

Yes. I bought one of these a couple of years ago. Thinking of getting it out of the box and using it. Question is how much faith one can put on one of these not failing with someone's speaker!

 

[igufi]

REL is *far* from the best for music. They make mediocre subs at best. Just very good at marketing to audiophiles, so they have garnered that reputation. Just as B&W has an incredibly strong audiophile reputation as a great speaker, while most actually are quite poor. Any comparably priced Rhythmic sub is far superior to any REL in every metric.

I’ve been wondering about this - I am on the market for a sub and I was thinking of getting a REL 212/SE but there seems to be absolutely zero measurements on it on the web. Same goes to the other modern REL models, which is suspicions, especially as the manufacturer provides quite sparse measurements of their own.

Are there any subwoofer manufacturers that provide comparable fit&finish in 2-3k€ range?

 

[Francis Vaughan]

I believe Rhythmic’s site says the reason they use the servo system they do is that it minimizes the delay between sensing the cone action and generating a reaction. Is that assertion flawed? Is a system using accelerometers providing feedback to a controller just as fast or faster (less lag/latency)?

The feedback loop is attempting to linearise the relationship between cone velocity and input voltage. So, in order to use an accelerometer you need to integrate the acceleration. This of necessity means a lag over a direct velocity sensor (such as magnetic coil.) Where this matters is in the design of the feedback loop. The additional lag means your stability margin is lessened, and you cannot use as much feedback.

The patent on the sub design explicitly references using either direct velocity of accelerometers, so is covering the bases. But there is clear value if you can get a good direct velocity signal. Things are never clear cut, and I can imagine difficulties in avoiding cross coupling between the voice coil and velocity sensing coil making the choice less clear in achieving a good outcome. Engineering is always a mix of such questions.

 

[Tom C]

The feedback loop is attempting to linearise the relationship between cone velocity and input voltage. So, in order to use an accelerometer you need to integrate the acceleration. This of necessity means a lag over a direct velocity sensor (such as magnetic coil.) Where this matters is in the design of the feedback loop. The additional lag means your stability margin is lessened, and you cannot use as much feedback.

The patent on the sub design explicitly references using either direct velocity of accelerometers, so is covering the bases. But there is clear value if you can get a good direct velocity signal. Things are never clear cut, and I can imagine difficulties in avoiding cross coupling between the voice coil and velocity sensing coil making the choice less clear in achieving a good outcome. Engineering is always a mix of such questions.

I appreciate the reply, Mr. Vaughn, thank you.
So, if I’m understanding, it’s possible the Rythmic system could have some advantage, but whether or not the advantage is realized depends on the implementation. The devil’s in the details.

 

[DS23MAN]

The feedback loop is attempting to linearise the relationship between cone velocity and input voltage. So, in order to use an accelerometer you need to integrate the acceleration. This of necessity means a lag over a direct velocity sensor (such as magnetic coil.) Where this matters is in the design of the feedback loop. The additional lag means your stability margin is lessened, and you cannot use as much feedback.

The patent on the sub design explicitly references using either direct velocity of accelerometers, so is covering the bases. But there is clear value if you can get a good direct velocity signal. Things are never clear cut, and I can imagine difficulties in avoiding cross coupling between the voice coil and velocity sensing coil making the choice less clear in achieving a good outcome. Engineering is always a mix of such questions.

https://www.rmsacoustics.nl/audiodesign.html

 

[Bear123]

I’ve been wondering about this - I am on the market for a sub and I was thinking of getting a REL 212/SE but there seems to be absolutely zero measurements on it on the web. Same goes to the other modern REL models, which is suspicions, especially as the manufacturer provides quite sparse measurements of their own.

Are there any subwoofer manufacturers that provide comparable fit&finish in 2-3k€ range?

Overseas shipping really changes the cost *alot* compared to the prices available here in the states. 2-3K per sub or total? Maybe take a look at Power Sound Audio S1512...they are available in a variety of wood veneer finishes if you want a furniture grade sub. They use exceptionally high quality B&C Neo pro audio woofers. This gives very very high quality sound in the mid and upper bass with very low distortion, but also quite strong deep bass performance with 16.5mm Xmax.

https://www.powersoundaudio.com/products/S1512

https://www.bcspeakers.com/en/products/lf-driver/15-0/4/15ds115-4

Otherwise, Rythmik does offer some options that cost less with their 12" options. Hsu makes a nice sub as well with their ULS15 MKII.

Not familiar with European based options.

 

[Poseidons Voice]

To augment on what Bear123 stated, it bears saying that a subwoofer driver that is optimized for excellent midbass is quite different from one that is optimized for frequencies less than 50Hz. What Rythmik has done is try to have drivers that work well in both cases due to the use of Brian Ding's servo design. However, the caveat is that it only works up to a maximum SPL level (judging by the THD and SPL compression curves). This is obvious when perusing Josh Ricci's Databass site.

I realized some of this a long while back. As such, my multisub setup tries to get the best of both worlds (maybe all?). I use B&C and BMS drivers along with Funk Audio drivers. The only EQ I have had to implement is below 30Hz for extension purposes. Other than that, the asymmetric distribution of subwoofers substantially cancels the need for EQ above 30Hz. The crossover points used, phase, as well as slopes were all dictated by one thing. My microphone, not my ears!

I did not use Acourate. I use a Behringer DCX2496, and tons of measurements with a calibrated microphone.

 

[Tom C]

To augment on what Bear123 stated, it bears saying that a subwoofer driver that is optimized for excellent midbass is quite different from one that is optimized for frequencies less than 50Hz. What Rythmik has done is try to have drivers that work well in both cases due to the use of Brian Ding's servo design. However, the caveat is that it only works up to a maximum SPL level (judging by the THD and SPL compression curves). This is obvious when perusing Josh Ricci's Databass site.

I realized some of this a long while back. As such, my multisub setup tries to get the best of both worlds (maybe all?). I use B&C and BMS drivers along with Funk Audio drivers. The only EQ I have had to implement is below 30Hz for extension purposes. Other than that, the asymmetric distribution of subwoofers substantially cancels the need for EQ above 30Hz. The crossover points used, phase, as well as slopes were all dictated by one thing. My microphone, not my ears!

I did not use Acourate. I use a Behringer DCX2496, and tons of measurements with a calibrated microphone.

View attachment 55413View attachment 55415View attachment 55416View attachment 55417View attachment 55418View attachment 55419

Are all those driven by a single LFE/x.1 channel?

 

[Poseidons Voice]

I have dual XLR outputs on my preamp. You can do the same with a DAC, just use an XLR splitter. Or with an AVP, you can use an XLR splitter as well. I take the L/R XLR outputs which plug into the DCX2496. The DCX 2496 has a “SUM” feature which converts it into mono.

Others use a Minidsp instead of the Behringer DCX2496.

Tomato-tomato...

Best,
Anand.

 

[Ron Texas]

Some around here are ignoring that the L12 cost $559 shipped (plus sales tax in the US) with $50 off the pair if you buy two. So I don't know what one member has in mind about this being badly designed, but you can have 2 (or more) of these for whatever he has in mind. Remember, the competition here is the SVS SB-1000.

 

[MZKM]

115db peak at the listening point is THX reference. Remember this is for movies, where you're more likely to have big peaky low-frequency sound effects, and again that's after room gain and EQ effects on headroom.

That's for just the sub though. If you have a 7.1, under the most strenuous scenario, the sub would need to output 117dB at the MLP. For shits and giggles, for a 10.1.24 Atmos setup (max allowed for residential use), the sub would need 121dB.

 

[Dj7675]

Some around here are ignoring that the L12 cost $559 shipped (plus sales tax in the US) with $50 off the pair if you buy two. So I don't know what one member has in mind about this being badly designed, but you can have 2 (or more) of these for whatever he has in mind. Remember, the competition here is the SVS SB-1000.

I think the delivered cost is often overlooked as well the size. I original had 2 Dayton Ultimax UM18 in a 4cu ft sealed enclosure in my HT powered by a Behringer NU6000. It is a significant DIY project to undertake. I ordered the 2 Rythmik L12’s to replace the 2 UM18s and there has been a big improvement. The L12’s are so small I can put them about anywhere, where the UM18s I was limited to putting them on the front wall. Measured frequency response is much much better. Output is fine for my listening levels. Will probably pick up a 3rd at some point.

 

[MZKM]

Most of the SVS sealed subs do just fine up to 250hz or 300hz, I merely use them as an example because I'm most familiar with them. But it's certainly not an uncommon or expensive capability.

This Rythmik subs can go just as high, it’s just a different setting:

 

[Prana Ferox]

Some around here are ignoring that the L12 cost $559 shipped (plus sales tax in the US) with $50 off the pair if you buy two. So I don't know what one member has in mind about this being badly designed, but you can have 2 (or more) of these for whatever he has in mind. Remember, the competition here is the SVS SB-1000.

In this price range I'd be looking at the Monoprice Monoliths.

 

[Tom C]

In this price range I'd be looking at the Monoprice Monoliths.

In what way or ways would they be preferable to Rhythmik?

 

[Ron Texas]

In what way or ways would they be preferable to Rhythmik?

I looked at them too. At the moment I can't remember exactly what I didn't like.

 

[Hidde]

A small value series/sensing resistor doesn't make the amplifier current-drive. From what I gleaned from Brian Ding's vague explanation regarding this, the series resistor used is only 0.1 ohms.

The amplifiers used by Rythmik here are voltage sources.

Dave.

Assuming the voltage source works like any other differential amplifier and operates according to (OpenloopGain)*(NONIN-IN). The amplifier tries to keep the voltage difference on it's input equal to 0.

Knowing that, we can put a sense resistor between the load and ground (very small, order of 0.1Ohms as you mentioned) and feed it back to the input of the amplifier. Since ohms law comes witht the resistor and we control the voltage over the resistor now. You could also say that you control the current! Hence a Transconductance amplifier is made!

 

[Deleted member 2944]

Assuming the voltage source works like any other differential amplifier and operates according to (OpenloopGain)*(NONIN-IN). The amplifier tries to keep the voltage difference on it's input equal to 0.

Knowing that, we can put a sense resistor between the load and ground (very small, order of 0.1Ohms as you mentioned) and feed it back to the input of the amplifier. Since ohms law comes witht the resistor and we control the voltage over the resistor now. You could also say that you control the current! Hence a Transconductance amplifier is made!

I gathered from his description that voltage feedback from the small transducer coil was the primary input to the servo control. So, no, this is not a transconductance amplifier.

You seem to be interested in redesigning his scheme to what it could be.....I'm just relating to you what it IS. If you have some suggestions to improve his product, I suggest to contact Brian Ding directly.

Dave.

 

[CumSum]

I think the delivered cost is often overlooked as well the size. I original had 2 Dayton Ultimax UM18 in a 4cu ft sealed enclosure in my HT powered by a Behringer NU6000. It is a significant DIY project to undertake. I ordered the 2 Rythmik L12’s to replace the 2 UM18s and there has been a big improvement. The L12’s are so small I can put them about anywhere, where the UM18s I was limited to putting them on the front wall. Measured frequency response is much much better. Output is fine for my listening levels. Will probably pick up a 3rd at some point.

I completely disagree. I own a Rythmik L12 for my bedroom system. And I own 2 of the same DIY 18" Sealed Dayton Ultimax's powered by an iNuke6000 for my large HT.

The 18" Ultimax is in a different league, especially for high SPL listening. It easily competes with the likes of the SVS SB-16 and Rythmik F18. Now in my relatively small bedroom it would be way overkill and the Rythmik L12 is a wonderful sub to listen to for that space. In terms of bass quality, I prefer the Ultimax's in my large mid-field setup over my Rythmik in a small near-field setup. And the Dayton has the output equivalent of 8 Rythmik L12's. For $1400 to build two of them, there is no other product that can come close to the Ultimax's.

I am going to guess that you had no room correction running with the Ultimax's. It is the only way the Rythmik would sound like an improvement.

Every sub need room correction, but especially a DIY sub like the Ultimax. The Ultimax does not have a flat frequency response. The Rythmik L12 has a flat frequency response only because it is running DSP inside of its plate amp.

Both my HT and bedroom setups are running DIRAC through MiniDSP's. Both setups have 99% of their bass performance extracted from the room. And the Ultimax is the undisputed champ.