Bose QuietComfort 35 II Review (Noise Cancelling Headphone)

[riker1384]

I tried a pair of the wireless version of their AE2. It had one big problem, there was noticeable hiss when sound was coming through, and when sound wasn't coming though it would shut the amp off entirely and it would be silent. So this happened quickly enough that the noise would pump on and off during pauses in a podcast. I found this unbearable and worse than if the noise had been constant.

Do the Quietcomfort models have noticeable noise and does it cut in and out like this?

 

[Helicopter]

I tried a pair of the wireless version of their AE2. It had one big problem, there was noticeable hiss when sound was coming through, and when sound wasn't coming though it would shut the amp off entirely and it would be silent. So this happened quickly enough that the noise would pump on and off during pauses in a podcast. I found this unbearable and worse than if the noise had been constant.

Do the Quietcomfort models have noticeable noise and does it cut in and out like this?

No, the noise cancellation stays on. You can play one song, fall asleep, and you still have NC five hours later. I detect no hiss from the NC, but it does have a sound I suppose... dufficult to describe since I perceive it as quieter than background noise.

 

[JoostE]

I don't see them as an audiophile headphone, they're just a very competent consumer headphone with great utility - bluetooth, noise cancelling, portability, comfort, etc. I almost never listen to music with them, but for podcasts or radio etc i'm often wearing them.

Yeah as a scientist who likes the scientific side of audio, but also heard multiple pairs of these. This review makes me scratch my head... I cannot imagine these graphs tell the whole story.

Yes they are fine, if people are looking for a noise cancelling headphone, this is the one to pick. (heck, if we are going to fly often again, I might pick up a pair). Enjoyable? yes, but the details in the music are simply not really there. What am I missing here?

 

[MayaTlab]

I cannot imagine these graphs tell the whole story.

What am I missing here?

IMO I don't really like the way the FR is shaped between, let's say 1khz and 4khz or so. It's quite uneven and in that range it's likely to manifest itself in a more or less similarly uneven way on your own head. Above that it probably gets increasingly variable between people but it's entirely possible that on your head they're quite uneven too, albeit at higher and higher frequencies from what I understand we are less and less sensitive to high Q variations in FR.
Basically, they might be fairly well balanced if your were to look at them at the "macro" level, which is perhaps the most important, but looking at exactly how the curve is shaped in more detail shows some shortcomings.
As a wireless, powered ANC headphones they also have other things to contend with that may be quite difficult to effectively measure, such as noise floor / ANC hiss (the only ANC headphones that I find tolerable in that regard are the most recent ones from companies fairly advanced in the ANC space, the Bose 700, XM4 and AirPods and at least the Bose 700 has some weird peculiarities in that aspect), the quality of their audio over bluetooth implementation (not necessarily a question of codec and bitrate, there a few little tests you can perform to assess that), etc. Perhaps also something related to how some ANC headphones (including this one) can use their inside mic to adjust the output in real time below 1khz to compensate for variations in clamp force / seal quality. This system can lower THD apparently (https://diyaudioheaven.wordpress.com/headphones/measurements/bose/qc35-ii/), but could it be susceptible to produce artefacts that wouldn't manifest themselves in that way ?

 

[JoostE]

Yeah I am going to try these with some EQ if one of my friends who have them visits. Coming from the self build audio world though, I cannot shake the feeling that just looking at the charts we are looking at, is not enough. We are missing information. I use a DSP in my stereo system, and while I can make a 2 speakers follow the exact same FR, that doesn't mean they sound the same.

For example, I build my stereo setup with 8 10" woofers for 2 reasons. First, to mimic a point source (all 4 are evenly spaced around the mid/high section), second to need as little movement of the speakers to create bass, making the movement faster, very audible in fast kick drum recordings. Both of these improved the sound quality and imaging of the speakers significantly.

Extending these two examples to headphones, I can imagine placement of the drivers inside the headphones, and strength of the magnets / rigidness of the surround could have the same effects. However, I don't think they would reflect in the graphs we keep staring at.

I guess what it comes down to is that this review, with a pair of headphones I am very familiar with, gives me a better understanding of how to use measurement data in my personal audio journey. If something measures bad, it is probably bad, but if something measures good, it doesn't mean it sounds good.

 

[MayaTlab]

Yeah I am going to try these with some EQ if one of my friends who have them visits. Coming from the self build audio world though, I cannot shake the feeling that just looking at the charts we are looking at, is not enough. We are missing information.

For passive, wired headphones, the main information you're missing is the FR curve at your own eardrum. We can learn a tremendous lot from test rigs measurements but particularly at the two extremes the FR can deviate quite significantly from the test rig.
I think this paper is a good illustration of that : https://www.aes.org/e-lib/browse.cfm?elib=16877
If you want to know they used that : https://www.etymotic.com/product/er-7c/
Mere mortals can use small electrets placed somewhere in the concha or near the ear canal but the results you'll get are only truly somewhat valid up to 1khz... if you're careful and methodical.
This is what Rtings uses to measure bass response on five humans and then merge it above a few hundred Hz with their dummy head measurements.
For the QC35II this is a moot point as they use their ANC circuit and inwards facing mic to ensure a constant FR below 1Khz (cf Rtings, DIYaudio).

Then there's the question of the interpretation of the graphs, ie in my view the audibility of some features of a headphones' FR curve and hierarchising the problematic features in terms of importance (which may sometimes depend on how you intend to use the headphones, for example whether you plan to use them with EQ or not).

I use a DSP in my stereo system, and while I can make a 2 speakers follow the exact same FR, that doesn't mean they sound the same.
For example, I build my stereo setup with 8 10" woofers for 2 reasons. First, to mimic a point source (all 4 are evenly spaced around the mid/high section), second to need as little movement of the speakers to create bass, making the movement faster, very audible in fast kick drum recordings. Both of these improved the sound quality and imaging of the speakers significantly.
Extending these two examples to headphones, I can imagine placement of the drivers inside the headphones, and strength of the magnets / rigidness of the surround could have the same effects.

I'll leave that to people with more knowledge in acoustics, but I'd be very, very wary of equating one's experience with speakers with headphones. For starters from what I understand headphones mostly operate in pressure conditions up to a few kHz unlike speakers.

If something measures bad, it is probably bad, but if something measures good, it doesn't mean it sounds good.

That's mostly what I look for with measurements on a test rig : is it good enough to be worth trying ?

 

[JoostE]

Yeah I am just trying to say that just a frequency response, THD and phase response is not enough to completely cover the quality of a headphone. I already had that feeling, but this review confirms that for me.

 

[WickedInsignia]

Bose reduced the ANC performance of these with a firmware update when the 700 came out….and the 700 requires you to download their app to be able to use the headphones at all.

That’s grounds enough for me to avoid Bose, no matter how good they sound. I’m completely out of patience for anti-consumer practices.
The Razer Opus measures slightly better while also being cheaper and less scummy, so maybe I’ll check those out.

 

[infinitesymphony]

Bose reduced the ANC performance of these with a firmware update when the 700 came out….

This sounds like a juicy tidbit, but was it ever confirmed with measurements?

"In a blog post that chronicles the entire saga, Bose states how the murmurs initially started back in January 2017 when the firmware version 1.2.9 did cause an issue with audio quality with the QC35. Upon close inspection, Bose learned that this was due to the headphones using the wrong audio profile and the issue was fixed in the subsequent 1.2.10 update. However, some customers still kept complaining about the sound, even though Bose found no evidence to support those claims.

Later in June 2019, Bose pushed out update 4.5.2, which is when the complaining reached an all time high. Many customers claimed how the update made the active noise cancellation feature worse, that it was buggy or broken and that all of this was done to promote the sales of the new Bose Headphones 700.

To deal with this, Bose did an extensive evaluation of the firmware update, both internally and using an unbiased outside contractor. The company reviewed and repeated manual and automated firmware QA testing and spent dozens of hours replicating customer issues. During this process, the company found no evidence of any change made in the update that would directly or indirectly affect the noise canceling performance."

https://www.gsmarena.com/bose_dispr...erformance_with_qc_35_ii_updat-news-42506.php

 

[DeepFried]

This sounds like a juicy tidbit, but was it ever confirmed with measurements?

 

[infinitesymphony]

Thank you for that. If someone at home with a limited setup can confirm this difference it makes Bose's response even more puzzling. It sounds like at least they allow a firmware downgrade for anyone who wants to go back.

 

[WickedInsignia]

This sounds like a juicy tidbit, but was it ever confirmed with measurements?

Yeah, Rtings measured it before and after and they found it was worse.
Considering the timing and Bose’s penchant for this sort of thing I would say it’s deliberate.
Bose usually have all their headphones on display next to each other in the store, so anything that would give the new model an edge I guess.

 

[solderdude]

Ratings had this to say about this:

Update 08/22/2019:
We've re-tested the noise isolation of the Bose QC35 II after noticing a flaw in our testing methodology. While the 4.5.2 firmware update reduced the ANC performance slightly, it's less than we initially reported. This review has been updated to reflect these changes.

While the ANC doesn't work quite as well as it did before the 4.5.2 firmware update, it's still outstanding and is among the best we've ever tested. These headphones can block out the low rumble of bus or plane engines as well as background chatter or the hum of nearby AC units.

The latest FW is 4.8.1. which is not tested by Rtings it seems.
Talk on the Reddit Bose forum states 4.8.1. is 'better' than the 4.5.2. update.
Seeing as the video was posted (almost) 2 years ago and the 4.8.1. update is from 0.5 year ago is it possible the issue is resolved now ?

 

[ksaugrin]

I have purchased the QC35 II after the review but I am a little disappointed, don't get me wrong this is a good headphone and it sound right but I find it little bit dull sounding and not as enjoyable as my Etymotic ER4XR. I found the ANC doesn't compete with ER4XR amazing passive noise isolation.

 

[JohnYang1997]

I have purchased the QC35 II after the review but I am a little disappointed, don't get me wrong this is a good headphone and it sound right but I find it little bit dull sounding and not as enjoyable as my Etymotic ER4XR. I found the ANC doesn't compete with ER4XR amazing passive noise isolation.

Because the etymotics are too good for this world.

 

[ksaugrin]

Because the etymotics are too good for this world.

After being on Etymotic, it's hard to return to something other. I have pre-ordered the EVO I am so excited to test Etymotic new beast

 

[WickedInsignia]

I have purchased the QC35 II after the review but I am a little disappointed

Try the PXC 550 ii. I agree with you on the Bose and in my short listening it was very dulled in comparison to the Sennheiser.
The PXC is close to Harman with some flavorful additions (mainly in the low bass and treble) that make it very dynamic and lively along with that beautiful Senny midrange. At least that was my experience with the original, I hope to get the new version soon but they're meant to sound pretty much identical.

 

[solderdude]

The 5kHz peak is not appreciated by many owners in the PCX550(II) though.

I measured it and thought it sounded better than it measured. Never wrote a review about it though. Would take the PCX550 over the QC35 myself.

Below PCX550 FR in all 3 'positions' off, on, adaptive

and CSD showing the resonance.

compare to QC35:

It also has a resonance but and the treble is rolled off considerably compared to the clearer sounding PCX550.

FR of the QC35 for comparison on the same rig (on and off):

 

[WickedInsignia]

The 5kHz peak is not appreciated by many owners in the PCX550(II) though.

On the original the sound is thin with that prominent peak with ANC off. The Dynamic Loudness setting via the app mostly resolved it for me. With ANC on, I didn't find it to be an issue (bass masking maybe?)
I could have become accustomed to it, or the ii has a more prominent peak. The channel inconsistency in Rting's measurements bothers me a bit too.

Thanks for providing your measurements in the different settings! The sound signature is indeed different depending on whether ANC is set to off, adaptive or full.

 

[WickedInsignia]

I guess the other option is the Razer Opus, which is meant to be very close to Harman and overall a solid choice. The only thing that concerns me is reports of ANC hissing.