Dynaudio LYD 5 Studio Monitor Review

[amirm]

@amirm, any reason other than time for the multitone IMD measurements to go away?

It was more than time. It was super annoying for my wife (she was literally getting sick listening to test tones). My lab is in an open space overlooking the rest of the house and with the pandemic, my wife is most of the time home. I plan to move the test to the attic and close the door but that requires cleaning up the attic to make the space. And you know how much time I have for that. Come fall when the household chores reduce, I can resurrect them.

 

[q3cpma]

It was more than time. It was super annoying for my wife (she was literally getting sick listening to test tones). My lab is in an open space overlooking the rest of the house and with the pandemic, my wife is most of the time home. I plan to move the test to the attic and close the door but that requires cleaning up the attic to make the space. And you know how much time I have for that. Come fall when the household chores reduce, I can resurrect them.

I see, no problem then. I do wonder what the result would have been for this with the LF extension switch enabled, 5.2 kHz is no joke for a woofer; seen with the quite high THD already (compared to the 8030C, at least).

 

[BYRTT]

Based Amir's spindata Dynaudio LYD 5 radars with multiple overlays in good graph resolution, then in less graph resolution radar comparison LYD 5 verse KH 80 / 305P / 8030C, verticals then horizontals..

Comparison LYD 5 / KH 80 / 305P / 8030C, verticals then horizontals..

 

[testp]

Forgive another possibly silly question, but what's the difference between the LYD actives and the BM range, say here, a BM5A III? Are the LYDs supposed to be the next range up? I believe the BBC and other broadcasters seem to use BM5's for a lot of their general purpose broadcast studio use these days.

Their BM5 studio monitor is with longer successful history in production, that is now available as mark III, where class d amps are used.

LYD seems to be their new approach using higher crossover, that they wanted to try/ present to public.

Personally i would like to see BM5 mkIII test, where crossover is lower, also cheaper somewhat and 7" woofer i believe...

 

[watchnerd]

Forgive another possibly silly question, but what's the difference between the LYD actives and the BM range, say here, a BM5A III? Are the LYDs supposed to be the next range up? I believe the BBC and other broadcasters seem to use BM5's for a lot of their general purpose broadcast studio use these days.

The BM series are considered "legacy" at this point.

Compared to the older BM models, the LYD series have DSP based crossovers, Class D amplification, and different drivers with different crossover points. The "classic BM" models use class AB amps, active analog crossovers, and much lower tweeter crossovers.

The BM5 MKIII is an odd duck, as it has been updated to use LYD series electronics, but still has the old classic drivers.

I can't really figure out what it's purpose in life is.

The revisions of the old series to new maps to:

BM series => LYD series

AIR series => Core series

 

[watchnerd]

Anyway I'd bet many of those speakers are at least 2nd order actual roll-offs on woofer and 3rd or 4th on tweeter.

That's pretty correct.

I once took an old Audience 42 and disconnected the passive crossover and implemented an external version of the same using DSP.

On measuring, the drivers had very well controlled acoustic roll-offs that lead to higher effective net results.

 

[watchnerd]

Based Amir's spindata Dynaudio LYD 5 radars with multiple overlays in good graph resolution, then in less graph resolution radar comparison LYD 5 verse KH 80 / 305P / 8030C, verticals then horizontals..
View attachment 82838
View attachment 82839

Comparison LYD 5 / KH 80 / 305P / 8030C, verticals then horizontals..
View attachment 82841

I can't hear the pink.

 

[peanuts]

i dont understand the point of these high first order crossovers.. have a pair of audience 52 from the late 90s, i think they are 1800hz 2.order

 

[PuX]

nice to see a Dynaudio speaker tested, thanks!

a larger model (LYD 7?) would probably be better for more/better bass.

or better yet, one of their higher end hi-fi speakers (Contour, Confidence). this one is an entry-level professional monitor.

 

[dddenis]

i dont understand the point of these high first order crossovers.. have a pair of audience 52 from the late 90s, i think they are 1800hz 2.order

Dynaudio woofers have a smooth roll-off free of resonance peaks, as watchnerd has already said, which makes it possible to use such flat filters. Flat filters make a very wide transition region where both woofer and tweeter play in parallel.
The directivity of the woofer increases with frequency, especially above ~2 kHz. Then, a tweeter which is mounted in a 7" baffle without waveguide or strongly rounded edges has a very wide dispersion around 3 kHz (due to reflections at the baffle edges). The directivity of the tweeter starts to increase above ~6 kHz.
Now, those factors in combination with the high crossover frequency around ~4 kHz and the flat/smooth crossover filter make the woofer's narrow dispersion to compensate the tweeter's wide dispersion and vice versa. This gives a controlled horizontal directivity in the critical frequency range. As I have mentioned in my last post, all this is at the expense of vertical directivity and intermodulation distortion. That's why I called this a dirty trick.

But I admire how good this trick works and how effective it is: A classic speaker design with 2-order filters and a crossover frequency of 2 kHz would feature an increasing directivity until 2 kHz, whereas it becomes wide at 3 kHz and increases again above 5...6 kHz. To overcome this effect and to provide a smoothly increasing directivity vs frequency is why so many speakers with controlled directivity have waveguides. Also, a waveguide and high-order analog/passive filters (which is not the case here, but in other Dynaudio designs) would increase cost. So, all you need is a woofer with smooth roll-off.

 

[watchnerd]

nice to see a Dynaudio speaker tested, thanks!

a larger model (LYD 7?) would probably be better for more/better bass.

or better yet, one of their higher end hi-fi speakers (Contour, Confidence). this one is an entry-level professional monitor.

Core 7 is the step up line in the pro monitors from here.

 

[watchnerd]

Dynaudio woofers have a smooth roll-off free of resonance peaks, as watchnerd has already said, which makes it possible to use such flat filters. Flat filters make a very wide transition region where both woofer and tweeter play in parallel.
The directivity of the woofer increases with frequency, especially above ~2 kHz. Then, a tweeter which is mounted in a 7" baffle without waveguide or strongly rounded edges has a very wide dispersion around 3 kHz (due to reflections at the baffle edges). The directivity of the tweeter starts to increase above ~6 kHz.
Now, those factors in combination with the high crossover frequency around ~4 kHz and the flat/smooth crossover filter make the woofer's narrow dispersion to compensate the tweeter's wide dispersion and vice versa. This gives a controlled horizontal directivity in the critical frequency range. As I have mentioned in my last post, all this is at the expense of vertical directivity and intermodulation distortion. That's why I called this a dirty trick.

But I admire how good this trick works and how effective it is: A classic speaker design with 2-order filters and a crossover frequency of 2 kHz would feature an increasing directivity until 2 kHz, whereas it becomes wide at 3 kHz and increases again above 5...6 kHz. To overcome this effect and to provide a smoothly increasing directivity vs frequency is why so many speakers with controlled directivity have waveguides. Also, a waveguide and high-order analog/passive filters (which is not the case here, but in other Dynaudio designs) would increase cost. So, all you need is a woofer with smooth roll-off.

All of that, plus phase benefits of 1st order.

 

[KaiserSoze]

Those are interesting results. Dynaudio sets a very high crossover frequency to overcome the need for a waveguide for controlled tweeter directivity. That's a dirty trick which comes along with worse vertical directivity and IMD at 2...4 kHz. The estimated in-room response looks so smooth but the vertical directivity is critical. It looks like the listening position must be vertically centered within much less than +-10°. Pink noise and moving your head up and down would easily reveal this weakness.

Essentially true except that the frequency that is affected is the crossover frequency which is in the vicinity of 5 kHz. Also I don't think that the reason for the high crossover point is for avoidance of the need for the waveguide. The tweeter would only begin to get increasingly directional at this frequency, where wavelength is still more than 2.5 x the diameter of the tweeter. But the woofer will have much higher directivity at 5 kHz compared to 2.5 kHz, even given its small diameter.

The tradeoff that Dynaudio has chosen here is interesting. They have clearly made an effort to keep the woofer and tweeter closer together, which would generally mean a wider listening window in the vertical plane, but the listening window in the vertical plane ends up being not a lot better than it is with most other small 2-way speakers that use a waveguide and a lower crossover point. From the standpoint of the listening window in the vertical plane it's kind of a wash. The explanation is most likely the one presented by Rooskie:

Quote from soundonsound, "Dynaudio say that the primary motivation is to get the crossover frequency, with its potential response discontinuities and phase changes, out of the region where the human ear is at its most sensitive. "

For reasons that made sense to them (but necessarily to everyone else) one of the early decisions was to use a high crossover point so that any crossover-related effects would be kept above 5 kHz. Given this high crossover point, the listening window in the vertical plane would have been very narrow if they had used a waveguide, or if the vertical separation between the woofer and tweeter had been as large as is typical, regardless of the reason why it was that great. Yet the listening window in the vertical plane would have been a good deal wider if they had used a lower crossover point, so clearly their philosophical views about crossovers are strongly held, since otherwise they would have set the crossover point lower, toward achieving a wider listening window in the vertical plane.

My guess is that notwithstanding that this is an active speaker, Dynaudio still chose to use low-order filters on the two drivers. This is evidenced by the way the "eyes" in the vertical dispersion are spread over an unusually wide range of frequency. The woofer's rolloff looks to be rather steep, which may be due more to the woofer's natural rolloff than to the LPF. The only way that Amir can take measurements of the individual drivers is by placing the microphone close to one or the other, which means that these measurements are affected by the output of the other driver. Nevertheless, the plot he shows for the tweeter is unusually flat below the crossover point, suggesting a gradual rolloff. This no doubt contributes to the good dispersion of the woofer, or rather makes up for the woofer's poor dispersion in the octaves immediately below the crossover point.

The tradeoffs Dynaudio has chosen are unconventional by today's norms and provide much food for thought.

 

[dddenis]

crossover frequency which is in the vicinity of 5 kHz.

As it can be seen in the near field measurements, it seems that the crossover frequency is maybe rather around 4 kHz. But this is hard to tell, as well as how the real curve of each driver looks like, because both woofer and tweeter active in this measurement. For a clean measurement of one driver, the other should be switched off.

 

[KaiserSoze]

I think the simplest way to understand it is that, when it is in the far field, the sounds from all the individual drivers will integrate well.

That's what I was thinking, that you have to be a certain distance from the speaker in order for the two independent wavefronts to have blended into a single wavefront. It is not the least bit apparent to me whether there is any meaningful equivalence between this notion and the technical definitions in the Klippel document.

 

[KaiserSoze]

As it can be seen in the near field measurements, it seems that the crossover frequency is maybe rather around 4 kHz. But this is hard to tell, as well as how the real curve of each driver looks like, because both woofer and tweeter active in this measurement. For a clean measurement of one driver, the other should be switched off.

The best and only good indication of the crossover point is in the vertical directivity graph. The crossover location is revealed by the location, on the frequency axis, of the "eyes" in the vertical directivity graph.

With speakers where steep filters are used the overlap region is small, such that the eyes are well-defined and give a good indication of the crossover frequency. If the slopes are less step the overlap region will be greater and the eyes will be less well defined, making it more difficult to pinpoint the exact location. The eye is better defined in the lower half of the graph and indicates a crossover point located at about 4.5 kHz.

As you suggest, it is difficult to identify the crossover point from looking at the "Driver Components Nearfield" graph. There is no way for Amir to take "clean" measurements of the individual drivers because he would have to open up the speakers that people send to him.

 

[NTK]

That's what I was thinking, that you have to be a certain distance from the speaker in order for the two independent wavefronts to have blended into a single wavefront. It is not the least bit apparent to me whether there is any meaningful equivalence between this notion and the technical definitions in the Klippel document.

Hahaha... Mathematicians have the unique talent of transforming properties that have nice physical meanings into something else that are completely abstract.

Let's see if I can explain it here: Another defining characteristic of farfield is that the source will look like a point source [see note below]. When the source is a point source, many nice properties follow -- sound pressure level attenuates by 6 dB every doubling of distance, the sound waves behave like a plane wave (in the spherical sense), pressure and velocity are in phase, etc. Of course, when the source behaves like a point source, that also means all the disparate sources are integrated into one.

Thus, if we can use apparent sound power curves to determine where far field begins, all the other characteristics of far field (which include integration of the drivers) will apply.

[Edit:] Note: Before someone asks, the point source is not a generic point source, but a special one that has the exact same sound radiating directivity characteristics as the loudspeaker. If anyone still finds it difficult to conceptualize, then imagine it as a spherical source instead of a point source, but a very very small one.

 

[BYRTT]

All of that, plus phase benefits of 1st order.

Hope not to sound harsh but where is that phase benefit coming from i cant see it other than in a electric domain or for coxials, because of center to center distance plus the huge >10 decade XO region and always around 90º out of phase its a mess of new step responses for every few milimeter microphone is moved around, also because woofer cant keep that wide slope off axis and main lope in XO region tilt down for WT configuration and up for TW configuration and probably reason why famous Dunlavy made his speakers as WTW or WMTMW, agree subjetive 1st order acoustics can sound fantastic or heaven for some its just myself that cant live with the objective step responses because that signature is not inherent and coming from the track material.

Regarding LYD 5 sorry to say i dont think it use any 1st order slopes because vertical radar in XO region should then tilt downwards and as seen below lobe is pointing strait forward at 4695Hz, use of 4695Hz is because CAD software bench LYD 5 slope to be in area 4500Hz-4800Hz..

Another one its not 1st order slope in LYD 5 is analyze Amir's nearfield sweeps, to be more right or a real model they should be corected for a baffle loss/diffraction curve but i'm too lazy do that now, nearfield is not precise method over a certain frequency relative to piston diameter and distance to microhone but very precise in the low end so ignore low pass slope of woofer and tweeter in they not presenting real world because of physics, did find two slopes that overlay quite nicely with tweeter in animation below and they both of 4th order..

 

[watchnerd]

Regarding LYD 5 sorry to say i dont think it use any 1st order slopes

I think you misunderstood my comment.

I wasn't referring to the LYD series.

I was saying that, back in the old days, >20 years ago, one of the reasons some designers preferred 1st order crossovers was the claim that it gave better phase coherence.

My old Audience 42 had a 1st order electrical network.

 

[tktran303]

^^^ (@post 98)

Exactly. The benefits, real or perceived, of 1st order filters... well what matters is the acoustic roll off.

And although Dynaudio have often touted 1st order filters in the past, I haven’t seen one example Of their speaker where their crossover filters ARE first order electrical, or the driver responses are 1st order acoustic.