Mark Levinson No 5909 Headphone Review

[amirm]

In “digital mode”, as Amir listened, is the ANC ever off?

Yes. You can cycle between pass-through, environment aware and ANC on. I tested with the first. I did turn on the ANC but it made no difference in my setup as the cups seal enough as to not matter (with respect to fan noise next to me).

 

[wwenze]

40-50ms for the digital interface (normal) and another 30 for the noise-cancelling...

Individually these are acceptable latency but together... delicate...

 

[Maiky76]

This is a review and detailed measurements of the Mark Levinson No 5909 noise cancelling headphone. It is on kind loan from a member and costs US $999.
View attachment 215166

I must say, I expected something larger and more substantial. Other than the logo, not much oozes luxury befitting of the brand. Expanding the headband causes creaking sound. And overall feel -- whether made so -- is somewhat plasticky. The cups are naturally small and feel that way. Mind you, it is not bad overall. Just doesn't match expectation.

The buttons are cryptic and "overloaded" as far as usage with multiple modes. Power on is press for 1 second. But pairing is pushing longer. Some little LED blinks this color or that and I was not about to memorize what was what. Fortunately there is pleasant voice feedback which helps a ton compared to others that lack it.

Analog input oddly, is provided via a special USB to 3.5mm cable. I appreciated that it was there, not forcing me to use Bluetooth. Plug in an ordinary USB-C cable and the headphone acts like a DAC which is very nice. What is not so nice is that it turns off Bluetooth so the accompanied app no longer works. The app is where you change the EQ settings and such. Why or why is this done? Worse yet, at least on Windows, the volume buttons did not work. I had to use the sound control panel in Windows which made me quite grumpy.

Disclaimer: our company Madrona Digital is a dealer for Harman products in custom install industry. So while we theoretically could source this headphone, it is not something we remotely sell. Still, if you want to read bias into this review, by all means, do so.

Mark Levinson No 5909 Measurements
The small cup and their tendency to lose seal made fitting the headphone on measurement rig extremely difficult. That is, until I realized some of it is inherent to the design and is corrected in digital input mode! Here is analog:

View attachment 215171

Now as DAC device on Windows:
View attachment 215172

Notice the much improved bass and matching of channels! Shoot, I could have saved half hour if I knew this. There was another major impact and that was in Group Delay:

View attachment 215173

Fascinating!

EDIT: my mistake for not equalizing the vertical axis. That may show both having similar issue.

Back to our frequency response, compliance with our target is excellent as we would expect. There are some kinks to be sure and some resonance in higher frequencies but overall, this is the second best fit I have seen in all the headphones I have tested.

View attachment 215176

Distortion was the same in both modes:
View attachment 215177
View attachment 215180

As you see, in pass-through analog mode, the headphone went out of control at 114 dBSPL and with digital, it would simply not get that loud. Otherwise, distortion is extremely low. Considering that we don't need to boost anything, this is even better result than it seems. Here is the absolute level distortion:
View attachment 215181

There is one bad peak which we are going to correct anyway. It is bad enough that it is contributing to extra output at that frequency.

Passive mode input has very high impedance:

View attachment 215183

Which tells me that it is a buffered input of sorts. Sensitivity in this mode is average:
View attachment 215184

Mark Levinson No 5909 Listening Tests
I started listening before I measured the headphone and it immediately produced that familiar tonality I expect from neutral headphones. Post measurements, I started listening in digital mode that I thought it was a bit cleaner. Same nice tonality although with a hint of brightness. So I pulled out the EQ:
View attachment 215185

I tried a second filter at 12 kHz and was not sure if I liked it better. It seemed to be more "accurate" but made the sound more closed. Speaking of which, that is a major problem with this headphone. It has almost no spatial effects. With the small drivers, the sound is coupled claustrophobically inside your head. I switched to my DC Stealth headphone and what a revelation that was in this front.

Another problem in digital mode was lack of volume. Yes, I could get it loud but not enough for bass notes to have any impact. Switching to analog fixed this, allowing me to turn up the volume much more. Still, bass notes lacked impact that Stealth brings. I just couldn't get excited over the experience with the 5909.

Conclusions
Nice to see Harman bringing more headphones out that comply with their own research. Tonality of the No 5909 is right on the money, sans a bit of resonance brightness (which may be fine with others). The issue is that it comes in a small form factor that while good for portability, misses the mark hugely to provide a statement kind of experience. I am sure a Mark Levinson user would not mind schlepping a large headphone on the plane that sounded better. As is, from looks to experience, it is that of a $300 headphone, albeit with very correct sound.

I hope I am not being too insulting but I wish Harman would just OEM a headphone from Dan Clark similar to Stealth line. Price it at $4,000 and it would sell similarly to what this is doing. FYI the number of downloads for the 5909 App was "100+" so not that much sale anyway.

It pains me to not recommend a headphone that hits the magical tonality curve but here we are. I want the headphone experience to do things that even good speakers can't. And we simply are not there with Mark Levinson No 5909 headphone.

-----------
As always, questions, comments, recommendations, etc. are welcome.

Any donations are much appreciated using: https://www.audiosciencereview.com/forum/index.php?threads/how-to-support-audio-science-review.8150/

Here are some thoughts about the EQ.

Notes about the EQ design:

• The average L/R is used to calculate the score.
• The resolution is 12 points per octave interpolated from the raw data (provided by @amirm)
• A Genetic Algorithm is used to optimize the EQ.
• The EQ Score is designed to MAXIMIZE the Score WHILE fitting the Harman target curve (and other constrains) with a fixed complexity.
  This will avoid weird results if one only optimizes for the Score.
  It will probably flatten the Error regression doing so, the tonal balance should be therefore more neutral.
• The EQs are starting point and may require tuning (certainly at LF and maybe at HF).
• The range around and above 10kHz is usually not EQed unless smooth enough to do so.
• I am using PEQ (PK) as from my experience the definition is more consistent across different DSP/platform implementations than shelves.
• With some HP/amp combo, the boosts and preamp gain (loss of Dynamic range) need to be carefully considered to avoid issues with, amongst other things, too low a Max SPL or damaging your device. You have beed warned.
• Not all units of the same product are made equal. The EQ is based on the measurements of a single unit. YMMV with regards to the very unit you are trying this EQ on.
• I sometimes use variations of the Harman curve for some reasons. See rational here: https://www.audiosciencereview.com/...pro-review-headphone.28244/page-5#post-989169
• https://www.audiosciencereview.com/...pro-review-headphone.28244/page-6#post-992119
• NOTE: the score then calculated is not comparable to the scores derived from the default Harman target curve if not otherwise noted.

Fair L/R match.

I have generated two EQs, the APO config files are attached.
Score and Fit seem to yield rather different results this time.

this Headphone probably does not EQ, listening to some podcast (voices in general) may help determine if the HF peak is detrimental or not.

Score no EQ: 94.2
Score Amirm: 92.2
Score with Score EQ: 102.4
Score with Fit EQ: 99.8

Mark Levinson No 5909 Score EQ Flat@HF 96000Hz
June282022-104515

Preamp: -2.7 dB

Filter 1: ON PK Fc 66.66 Hz Gain 2.47 dB Q 1.47
Filter 2: ON PK Fc 100.75 Hz Gain -2.90 dB Q 2.34
Filter 3: ON PK Fc 1009.00 Hz Gain -1.06 dB Q 3.04
Filter 4: ON PK Fc 4235.02 Hz Gain -2.55 dB Q 4.28
Filter 5: ON PK Fc 5723.27 Hz Gain 4.18 dB Q 1.14
Filter 6: ON PK Fc 7004.97 Hz Gain -9.92 dB Q 6.00

Mark Levinson No 5909 Fit EQ Flat@HF 96000Hz
June282022-104616

Preamp: -2.9 dB

Filter 1: ON PK Fc 67.16 Hz Gain 2.47 dB Q 1.25
Filter 2: ON PK Fc 102.52 Hz Gain -3.26 dB Q 2.09
Filter 3: ON PK Fc 703.78 Hz Gain 0.93 dB Q 0.20
Filter 4: ON PK Fc 951.23 Hz Gain -1.88 dB Q 1.65
Filter 5: ON PK Fc 4356.05 Hz Gain -1.96 dB Q 6.14
Filter 6: ON PK Fc 5709.27 Hz Gain 4.25 dB Q 1.95
Filter 7: ON PK Fc 6920.99 Hz Gain -8.17 dB Q 5.09

 

[PeteL]

40-50ms for the digital interface (normal) and another 30 for the noise-cancelling...

Individually these are acceptable latency but together... delicate...

You mean for video? I think it’s quite ok, maybe I’m not very sensitive to these lipsync small delays. If it was for recording music in sync with a track This would be absolutely unusable, but I’m not that fussy for the bi-yearly airplane movie watching really.

 

[DanTheMan]

This side of research really interests me, as so far I haven't seen (and I'm speaking for myself) much HRTF research that can correlate these weird effects like those from the HD 800S' . Harman is godsend for tonality, but I personally can't see it going too much into the importance (or not) of driver size, angle, openness, etc. for what audiophiles usually call "technicalities".

Some serious p2p research would be very welcomed, as the anedoctal side of these tricks in headphone discussion/lore is heavily contaminated with cognitive bias, improper testing methods and general snake oilery.

There has been research that alludes to how these things work. http://dtmblabber.blogspot.com/2011/05/binaural-hearing-and-direction.html

I imagine a psychoacoustically optimized way to look at the impulse response may prove useful as well, but alas we are without research.

 

[Maiky76]

40-50ms for the digital interface (normal) and another 30 for the noise-cancelling...

Individually these are acceptable latency but together... delicate...

Noise cancelling does NOT add 30ms but at most 30us (rather 5-20us)

 

[respice finem]

999$, 99 for the product and 900 for bragging rights

That said, if ML listens and fixes the criticised points, "ver. 2.0" could become really worth it.

 

[Nango]

Wish you had tested ANC feature.

 

[Hephaestus]

I have had these since launch and have only 2 complaints:

1. Treble peak should be fixed via firmware update
2. Head on detection is flaky

Other than those I have been happy with mine.

 

[3125b]

Very much what I expected. Kind of similar to the K371 just wireless and a bit more (albeit still not overly) nicely made.
I don‘t have too much faith in the durability of this thing.

 

[amirm]

Wish you had tested ANC feature.

I don't have the setup to measure it. I did comment on testing it by ear earlier.

 

[solderdude]

@amirm might be interesting to find out what the clipping level is. Is it at say 107dB or 112dB ?
(This is the max of the internal amp).
We know it is above 104dB and well below 114dB. 10dB is quite a difference.
Maybe ramp up 20Hz and listen/measure for distortion and when it occurs measure SPL ?
Does it respond well to broken seal ?

I hope FW is upgradeable and batteries as well.

Also have a look here

Mark Levinson 5909, new ANC wireless cans that advertise use of the Harman curve

Just say this article here. Can't wait to read a review.

www.audiosciencereview.com

Sean Olive is also active in this thread.

 

[amirm]

€ 4k is a lot for an active headphone.

It is $1K in US. I have it packed and ready to be shipped tomorrow as the owner needs to listen to it. I can say that I cranked it up some but didn't seem to break up.

 

[PeteL]

€ 4k is a lot for an active headphone.

999

 

[GXAlan]

Which spatial features? The micro level ones that create sense of space? If so, sure, I have done that with many headphones and even speakers. The macro ones are not impacted a lot in that manner.

The ones you were describing as being better with the Sennheiser and Dan Clarks and are missing from these ML.

I honestly don’t use headphones except when I need to be quiet and use IEMs most of the time when I need to be quiet

Is it as simple as having big drivers?

 

[solderdude]

999

Hit the 4 instead of the 1 on the numpad....

The most interesting part of this headphone may well be that ANC is fully digital and with ANC off (passive mode) the tonal balance is the same.

 

[MayaTlab]

Passive mode input has very high impedance:
Which tells me that it is a buffered input of sorts. Sensitivity in this mode is average:

Some of USB-C's Audio Adapter Accessory Mode's design details in this PDF : https://www.ti.com/lit/ug/tidub66/tidub66.pdf?ts=1656314756284&ref_url=https%3A%2F%2Fwww.google.com%2F

Can this explain some of the interesting findings in "passive" mode (latency, impedance, etc.) ?

Do you recall what the firmware was ?

 

[MayaTlab]

The most interesting part of this headphone may well be that ANC is fully digital

Wasn't it already the case that most current ANC implementations are ?

and with ANC off (passive mode) the tonal balance is the same.

That's quite unlikely to happen for most people, for ANC headphones in general. Unless the feedback mechanism still operates when it's turned off, but to my knowledge only Apple products can do so (and the ML5909 can't).
That certainly wasn't the case for me with two different samples.

 

[amirm]

Do you recall what the firmware was ?

No but the app said it was up to date. This was a new headphone drop shipped to me a few weeks ago.

 

[solderdude]

Wasn't it already the case that most current ANC implementations are ?

This is the first one where I see little difference in tonal balance between passive and active.

Certainly not most ANC headphones do this digitally.

The ANC can be turned off in the 5909 (acc to Sean O.) and will still have the same tonal balance.
AFAIK this is the only headphone that does this (not change tonal balance without ANC).

Of course it will be seal dependent wihout ANC for sure.