• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

HD6XX bass

stevenswall

Major Contributor
Forum Donor
Joined
Dec 10, 2019
Messages
1,366
Likes
1,075
Location
Orem, UT
Of Sonarworks Truefi is still around, use that and it will extend the bass and improve things more than an amp.
 

bobbooo

Major Contributor
Joined
Aug 30, 2019
Messages
1,479
Likes
2,079
On a side note, I have an old NAD 705 and someone suggested it may have a nice dedicated headphone amplifier. Here are the schematics if anyone who know how to read them wouldn’t mind doing me the favor.

https://www.schematicsunlimited.com/n/nad/nad-705-receiver-integrated-amplifier-service-manual/download/MTkzNjQ=

Here's exactly how the 330 ohm output impedance of the NAD 705 will change the frequency response of your HD6XX (calculated using Reference Audio Analyzer's extremely useful tool here):

HD6XX.png

So, around a 1.5dB increase in mid/upper bass over your iPhone XS (which I'm presuming you're using with the Apple dongle which has negligible output impedance), if you volume match at around 1kHz. (Or conversely a 1.5dB decrease in mids/treble, depending on how your brain will perceive the change.) This may actually be less of a difference than using your Marantz SR5014, as AVRs are notorious for having ridiculously high output impedances on their headphone outs (my Denon's is a whopping 500 ohms!). One other thing such a high output impedance will do is reduce the electrical damping, which anecdotally may result in the bass sounding less controlled (although I'm not aware of any hard evidence this is effect is audible).
 
Last edited:

bobbooo

Major Contributor
Joined
Aug 30, 2019
Messages
1,479
Likes
2,079
Thanks, there’s an iPhone app, I’ll give it a try tonight

Alternatively you could try the EQ settings at the bottom left of this page. You'll need a music app with parametric EQ, such as Neutron Player (you can try it out for free here) on mobile, or you can use Equalizer APO together with Peace GUI on PC (both free). Those EQ settings should get you closer to the Harman target frequency response (the sound signature the majority of people prefer in blind tests) than Sonarworks, as 'Oratory', the acoustic engineer who came up with them, tries to closely match this target, and uses the same measuring equipment as the researchers who devised it in the first place. More details from Oratory can be fund in his FAQ on Reddit here.
 

solderdude

Grand Contributor
Joined
Jul 21, 2018
Messages
15,891
Likes
35,912
Location
The Neitherlands
In what way is it theoretically impossible?

You are talking about nulling... When you have ever seen an actual reproduced signal of any transducer and compared that with the original you will know.
It looks like your knowledge is based on info from the web and little to no practical experience with actual measurements of transducers.
 

bobbooo

Major Contributor
Joined
Aug 30, 2019
Messages
1,479
Likes
2,079
You are talking about nulling... When you have ever seen an actual reproduced signal of any transducer and compared that with the original you will know.
It looks like your knowledge is based on info from the web and little to no practical experience with actual measurements of transducers.

I specifically asked in what way it is theoretically impossible (as you said "not in theory"), a genuine attempt to learn. But you're still talking about practical problems. I'm fully aware it would be practically extremely difficult to get right. If you have no answer to my actual question other than condescension, I really don't see the point of you replying.
 

solderdude

Grand Contributor
Joined
Jul 21, 2018
Messages
15,891
Likes
35,912
Location
The Neitherlands
Phase, (timing) and and amplitude difference will be so huge compared to the input signal you would be lucky to get a Df of 3 to 5 and here, the same as with the Df of DAC's some differences might be 'pleasurable' others may not.
The practical difference between any measurement is so huge (many many factors) when leaving the electrical plane it is unbelievable.

Just look at impulse plots of speakers to see the timing issues alone. Plenty of those on the web.
Or look at my squarewave plots to get an idea. There is not even a Pinna and fake ear canal to mess the signal up (which in that case would only be partially corrected) for headphones.
Nah.. if that were remotely feasable I would have done this years ago already.
I have seen too much real world test signals.

The closest one can get is subjectively and 'score' preference ratings by people that are highly trained.

Just take a mic, record a speaker signal and try to match that with Paul's software and you will see.
With headphones different obstacles are present, let alone the measurement method and used compensation.

When trying to improve SQ about 99% of this is transducers and the rest is amp and DAC.

I was taken a bit by surprise.. you appear to have accumulated a lot of info and appeared to me as someone with actual measurement experience so I was a bit baffled you suggested it could be used for this. So above my more to the point reply.
 
Last edited:
  • Like
Reactions: PSO

MRC01

Major Contributor
Joined
Feb 5, 2019
Messages
3,424
Likes
4,030
Location
Pacific Northwest
One of the challenges in boosting attenuated low bass up to neutral levels is that it increases driver displacement, which is a contributing factor to distortion. If the bass is significantly attenuated (say, more than 6 dB), then boosting it all the way up to neutral can make it sound loose or bloated. For example if you boost it by 6 dB you're making the driver (and amp) put out 4x more power at that frequency, relative to the rest of the spectrum. And it's not uncommon for EQ to apply even more boost than that. With headphones having a single driver (like the HD600, many others), this can increase midrange and treble distortion too, because the driver is being forced to travel closer to its extremes, where it's less linear. This relates to @solderdude 's comment earlier about keeping volume levels down.

Personally, I find what sounds best is typically a compromise. If low bass is attenuated, boost it closer to neutral, but limit the amount of boost to avoid making the cure worse than the disease.
 

KaiserSoze

Addicted to Fun and Learning
Joined
Jun 8, 2020
Messages
699
Likes
592
The veil isn't really due to a lack of treble - it's mostly down to the slightly raised upper bass / lower midrange a lot of Sennheiser headphones have, which perceptually masks higher frequencies, making them harder to distinguish, leading to a slightly veiled subjective impression (simultaneously providing a 'warmer' sound signature). (The HD650 does also have rolled off treble however, which exacerbates this problem further.) And yes, both the HD600 and HD650 have poor bass extension. So both deficiencies exist, and are clearly shown in their measured frequency responses.

I can't even imagine how you went about proving that this was the cause of a phenomenon that isn't even well-defined.
 

Kouioui

Active Member
Forum Donor
Joined
Nov 9, 2019
Messages
164
Likes
185
Location
Central FL
With headphones having a single driver (like the HD600, many others), this can increase midrange and treble distortion too, because the driver is being forced to travel closer to its extremes, where it's less linear. This relates to @solderdude 's comment earlier about keeping volume levels down.
The 650/6XX doesn't seem to suffer too badly according to Oratory when using EQ. Not much increased mid/treble distortion shown here:

1596147758021.png
 

bobbooo

Major Contributor
Joined
Aug 30, 2019
Messages
1,479
Likes
2,079
One of the challenges in boosting attenuated low bass up to neutral levels is that it increases driver displacement, which is a contributing factor to distortion. If the bass is significantly attenuated (say, more than 6 dB), then boosting it all the way up to neutral can make it sound loose or bloated. For example if you boost it by 6 dB you're making the driver (and amp) put out 4x more power at that frequency, relative to the rest of the spectrum. And it's not uncommon for EQ to apply even more boost than that. With headphones having a single driver (like the HD600, many others), this can increase midrange and treble distortion too, because the driver is being forced to travel closer to its extremes, where it's less linear. This relates to @solderdude 's comment earlier about keeping volume levels down.

Personally, I find what sounds best is typically a compromise. If low bass is attenuated, boost it closer to neutral, but limit the amount of boost to avoid making the cure worse than the disease.

Most (decent) software EQ has a digital pre-amp gain setting. Just set that to be at least minus whatever the peak of the total EQ curve is and you should be fine. Obviously your amp needs enough headroom for this in order to turn up the volume to satisfactory levels.
 
Last edited:

MRC01

Major Contributor
Joined
Feb 5, 2019
Messages
3,424
Likes
4,030
Location
Pacific Northwest
Most (decent) software EQ has a digital pre-amp gain setting. Just set that to be at least minus whatever the largest positive EQ filter gain is and you're fine. Obviously your amp needs enough headroom for this in order to turn up the volume to satisfactory levels.
Of course one always has to do that to prevent clipping, but it doesn't solve the problem I'm talking about.

In the above graph, bass distortion increased by about 10 dB. But that distortion didn't leak into the mids & treble, because the overall SPL at which they measured it was low enough. The point is that when you boost the bass with a single driver headphone, 2 things happen. (1) bass distortion increases at all levels. This can be benign, especially at low to moderate levels, since our hearing isn't as sensitive to distortion in the bass. (2) the SPL at which distortion starts to become audible in the midrange and treble, is reduced.

When EQ raises the amplitude of bass frequencies, it causes the driver to move with increased displacement/amplitude at the same overall SPL as before. This causes distortion to rise in the mids and treble earlier (at a lower volume) than it would without the EQ.
 

Kouioui

Active Member
Forum Donor
Joined
Nov 9, 2019
Messages
164
Likes
185
Location
Central FL
In the above graph, bass distortion increased by about 10 dB. But that distortion didn't leak into the mids & treble, because the overall SPL at which they measured it was low enough.
If you're listening at the 85dB shown (which is plenty to overcome the ear's sensitivity) and above for significant periods of time, the distortion you'll hear will be from your loss of hearing ability. Keep the volume reasonable people if you want to appreciate good audio for a lifetime.
 

KaiserSoze

Addicted to Fun and Learning
Joined
Jun 8, 2020
Messages
699
Likes
592
The headphone out is simply the speaker out via a 330 Ohm resistor so for today's standard a VERY high output resistance.
For those days quite normal.
Works well for high efficiency planar headphones and 600 Ohm headphones.
Not really suited for a lot of low impedance headphones. The HD6xx will sound noticeably 'darker' on this amp.
So there is no nice dedicated headphone section inside

View attachment 75867


The potential problem introduced by placing a resistance in series with the headphone output is that the voltage split is in accordance with the ratio of two series impedances. For example the "nominal" impedance of the HD580, HD600, HD650 and HDxx headphones is roughly the same as the 330 Ohm resistor. This is roughly the headphone's impedance at very low frequency and from about 800 Hz on up to about 8 kHz. But at the resonance, about 90 Hz, the impedance increases by about 40%. In the vicinity of this frequency the signal voltage seen by the headphone increases from roughly 50% (1 divided by 1 + 1) of the amplifier output voltage to about 58% (1.4 divided by 1 + 1.4) of the amplifier output voltage. The ratio of .58 V to .5 V is 1.17, and the base 10 log of 1.17 is .067, which means there will be mild peak of roughly + .7 dB centered at about 90 Hz. The mild peak will start from very low frequency and will have declined half way back to the nominal level by about 200 Hz. I doubt if this would be audible unless maybe you were making an effort to hear it.


(EDIT: I did the above scratch calculation before I saw the post from bobbooo, showing the change predicted by RAA. In trying to reconcile the difference, I will note that the 330 Ohm resistor is not the full output impedance of the amplifier, and all I considered was the 330 Ohm resistor. Also, my calculation of 1.17 as the voltage gain at 90 Hz, which corresponds to +.7 dB, is for voltage. To calculate the corresponding gain in acoustic power, the square of 1.17 is about 1.4, or 1.4 to 1.5 dB. However the RAA graph indicates about +3 dB, so there is a discrepancy, possibly explained by my not having considered the full output impedance of the amplifier.)

If I were concerned about it, I would maybe consider placing a parallel notch filter in the common (chassis ground) lead to the headphone jack. The combined, parallel impedance of the resistor, capacitor and inductor should be about 140 Ohms at 90 Hz, such that when this impedance is combined with the 330 Ohm resistor this combined series impedance will be about 470 Ohms, which is the impedance of the headphone driver at 90 Hz. (The reason for wanting the total series impedance of the notch filter and the 330 Ohm resistor to be equal to the impedance of the headphone driver is only that this is how it is elsewhere, away from the impedance peak.) If you don't want to solve three equations in three unknowns you could maybe just assign half of the 140 Ohms to the resistor, 1/4 to the capacitor and 1/4 to the inductor, calculating both of these impedances for 90 Hz. But if you were to do this, you would likely discover that you cannot hear any difference.
 
Last edited:

bobbooo

Major Contributor
Joined
Aug 30, 2019
Messages
1,479
Likes
2,079
I can't even imagine how you went about proving that this was the cause of a phenomenon that isn't even well-defined.

Not proof, just the most likely explanation given the data, see below.

HD600:
HD600a.png


HD650:
HD650a.png


HD58X:
HD58Xa.png


HD598:
HD598a.png


etc. etc.

The major commonality between all these is the raised upper-bass / lower midrange (seen clearly in the upper red curves), and far more Sennheiser headphones have this feature than don't. The psychoacoustics of lower frequencies better masking higher ones is consistent with this feature causing a slightly veiled sound. What the veil is not likely due to is a lack of treble, as several of Sennheiser's headphones do not lack treble, yet some say still sound slightly 'veiled', such as the HD600 above.
 
Last edited:

MRC01

Major Contributor
Joined
Feb 5, 2019
Messages
3,424
Likes
4,030
Location
Pacific Northwest
If you're listening at the 85dB shown (which is plenty to overcome the ear's sensitivity) and above for significant periods of time, the distortion you'll hear will be from your loss of hearing ability. Keep the volume reasonable people if you want to appreciate good audio for a lifetime.
I too am a proponent of listening at moderate volumes. And I consider the 85 dB measured to be "loud" not "moderate". Here, the above graph shows that the EQ increased bass distortion by about 10 dB. My point is that lifting attenuated bass always increases distortion, so be aware and do it with care. The increased distortion may or may not be audible, depending on how much the EQ boosts it, the headphone, the music one listens to, how loud one listens, and one's personal hearing sensitivity or acuity.
 

maverickronin

Major Contributor
Forum Donor
Joined
Jul 19, 2018
Messages
2,527
Likes
3,308
Location
Midwest, USA
If you're listening at the 85dB shown (which is plenty to overcome the ear's sensitivity) and above for significant periods of time, the distortion you'll hear will be from your loss of hearing ability. Keep the volume reasonable people if you want to appreciate good audio for a lifetime.

Quite a significant period of time. 85dB is perfectly safe for several hours daily. That's where US federal occupational noise exposure regulation begins.

Unless you listen at at that level all day long there's not really much to worry about.
 

Racheski

Major Contributor
Forum Donor
Joined
Apr 20, 2020
Messages
1,116
Likes
1,699
Location
Chicago

bobbooo

Major Contributor
Joined
Aug 30, 2019
Messages
1,479
Likes
2,079
Phase, (timing) and and amplitude difference will be so huge compared to the input signal you would be lucky to get a Df of 3 to 5 and here, the same as with the Df of DAC's some differences might be 'pleasurable' others may not.
The practical difference between any measurement is so huge (many many factors) when leaving the electrical plane it is unbelievable.

Just look at impulse plots of speakers to see the timing issues alone. Plenty of those on the web.
Or look at my squarewave plots to get an idea. There is not even a Pinna and fake ear canal to mess the signal up (which in that case would only be partially corrected) for headphones.
Nah.. if that were remotely feasable I would have done this years ago already.
I have seen too much real world test signals.

The closest one can get is subjectively and 'score' preference ratings by people that are highly trained.

Just take a mic, record a speaker signal and try to match that with Paul's software and you will see.
With headphones different obstacles are present, let alone the measurement method and used compensation.

When trying to improve SQ about 99% of this is transducers and the rest is amp and DAC.

I was taken a bit by surprise.. you appear to have accumulated a lot of info and appeared to me as someone with actual measurement experience so I was a bit baffled you suggested it could be used for this. So above my more to the point reply.

Thanks for the more useful reply. It still seems to me that electroacoustic null difference measurements would face pretty much the same types of challenges as electrical ones, just more so, and these are still all practical challenges, not theoretical barriers. To be clear, the full idea was to measure a high standard 'reference' headphone's acoustic output of some test signal (e.g. the Program Simulation Noise) using a high-quality mic (and ADC, and from a high quality DAC) via a flatplate set-up (no need for HATs as the effects of that would be nulled out anyway between headphones), then do the same for the headphone under test, then find the difference signal between the two recordings using e.g. DeltaWave (not the difference signal between the tested headphone's recorded acoustic output and the original test file - that of course would be way too high to be useful.) The test headphone could even be EQed to the reference headphone's frequency response to narrow the test to only focus on other deviations / degradations e.g. THD etc. Of course the precision of the measurements, environmental noise, exact positioning of the headphones, the added need for a high-quality mic (over just a high-quality ADC for electrical null tests) etc. would be enormous practical obstacles to overcome that may make this in practice impossible (with current technology, resources etc.), but as a thought experiment, it still seems like a theoretical possibility. Anyway, this is way off-topic, so I'll stop there.
 
Last edited:
Top Bottom