• 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!

Sennheiser HD650 Review (Headphone)

Robbo99999

Master Contributor
Forum Donor
Joined
Jan 23, 2020
Messages
6,875
Likes
6,673
Location
UK
Just get the permit to do that and here's my measurement
L+R total
View attachment 105785
Green= L, Red= R
View attachment 105786
That in room response is not too far off a Harman Curve in itself, it's about a 10dB difference from bass through to end of treble, so your speakers probably sound quite "Harman Like" so you shouldn't really dislike the Harman Headphone Curve in terms of tonality unless of course your personal head & ear anatomy (HRTF & also headphone transfer function) is quite different to the average mannequin that was used to create the Headphone Harman Curve and the measurements of the headphones......and also your speakers are not producing much low bass so a headphone that tracks the Harman Curve into the low bass will sound different to what you're used to.
 

Coach_Kaarlo

Active Member
Joined
Apr 27, 2020
Messages
196
Likes
222
Location
Sydney
Hi,

If anyone is interested.
based on the data from the first page, I did not go through the entire thread to check if the raw data was available...

Score no EQ: 72.3
Score with Amirm EQ: 79.4
Score with attached EQ: 87.8

Code:
HD650 APO EQ Score 96000Hz
January122021-123909

Preamp: -7.9 dB

Filter 1: ON PK Fc 31.8 Hz Gain 8 dB Q 0.5
Filter 2: ON PK Fc 178.3 Hz Gain -2.84 dB Q 0.94
Filter 3: ON PK Fc 9492 Hz Gain 6 dB Q 1.2
Filter 4: ON PK Fc 1257 Hz Gain -1.56 dB Q 3.26
Filter 5: ON PK Fc 2207 Hz Gain 0.81 dB Q 4.25
Filter 6: ON PK Fc 3295 Hz Gain -2.32 dB Q 3.39
Filter 7: ON PK Fc 5284 Hz Gain -4.25 dB Q 5.75
Filter 8: ON PK Fc 8496 Hz Gain -2.57 dB Q 8
Filter 9: ON PK Fc 13409 Hz Gain -7.2 dB Q 6

View attachment 105510

The Oratory EQ (attached) sounds and looks quite different from this despite having the same Harman target.


I am wondering why? I have an old pair of HD650's (circa 2007) driven off a Benchmark DAC2-HGC with factory headphone gain settings.


After a few hours with each EQ filter set (Amrim's default 4 band EQ | @Maiky76's January122021-123909 9 band EQ | the attached Oratory HD650 10 band EQ).

My personal subjective preference is the Oratory EQ - ok no big deal.

However, what I am confused about is the shear difference between 1.Oratory and 2.Amrim / January122021. If the Harman target is the goal how are such huge differences part of any subsequent EQ attempt? Have others found these EQ's to be similar or different? I have attached the results of my Earful hearing test - but nothing stands out which would explain the gap between EQ's.

Apologies if this was answered or discussed in the previous 26 pages.....just very curious where the big difference is occurring.
 

Attachments

  • Sennheiser HD650 (10 Band Graphic EQ).pdf
    318.4 KB · Views: 260
  • TEST 1_REFINED.txt
    433 bytes · Views: 155
Last edited:

Kuma

Member
Joined
Dec 6, 2020
Messages
45
Likes
124
Location
USA
I’ve no idea what I’m doing with EQ, HD650s, ADI-2...whatever.

I did my best to replicate Amir’s EQ settings in the ADI-2. Well, it sounds really good.
 

Robbo99999

Master Contributor
Forum Donor
Joined
Jan 23, 2020
Messages
6,875
Likes
6,673
Location
UK
The Oratory EQ (attached) sounds and looks quite different from this despite having the same Harman target.


I am wondering why? I have an old pair of HD650's (circa 2007) driven off a Benchmark DAC2-HGC with factory headphone gain settings.


After a few hours with each EQ filter set (Amrim's default 4 band EQ | @Maiky76's January122021-123909 9 band EQ | the attached Oratory HD650 10 band EQ).

My personal subjective preference is the Oratory EQ - ok no big deal.

However, what I am confused about is the shear difference between 1.Oratory and 2.Amrim / January122021. If the Harman target is the goal how are such huge differences part of any subsequent EQ attempt? Have others found these EQ's to be similar or different? I have attached the results of my Earful hearing test - but nothing stands out which would explain the gap between EQ's.

Apologies if this was answered or discussed in the previous 26 pages.....just very curious where the big difference is occurring.
The reasons for the differences are a couple of reasons. The first one is that the measurements are slightly different between Amir & Oratory for the HD650, especially in the treble, I took a look just now. Second reason is that Amir is eyeballing the frequency response and applying more approximate corrections to match the Harman Curve, whereas Oratory is trying to match the curve closer using I guess graphical/mathematical EQ software like REW for example, so they'd both end up with slightly different EQ's even if they EQ'd the same measurement. Both Oratory & Amir are correcting EQ also based on their listening tests, so both are not blindly matching the Harman Curve at all costs. I have one headphone that Amir has tested, the NAD HP50, and whilst I didn't compare Amir's EQ with Oratory's I did compare Amir's EQ vs my own EQ based on Oratory's measurement which I did using REW to match the Harman Curve accurately....and in my listening tests flipping between the two my EQ had far greater clarity and less muddiness - it boils down to the same two reasons, the measurement used and the EQ approach.

EDIT: there's also the variable of differences from unit to unit, but that's a relatively small random difference, so you'd think that wouldn't sway you in one direction or another when comparing EQ's from two sources.
 
Last edited:

svenz

Member
Joined
Dec 24, 2018
Messages
82
Likes
60
Location
London
The Oratory EQ (attached) sounds and looks quite different from this despite having the same Harman target.


I am wondering why? I have an old pair of HD650's (circa 2007) driven off a Benchmark DAC2-HGC with factory headphone gain settings.


After a few hours with each EQ filter set (Amrim's default 4 band EQ | @Maiky76's January122021-123909 9 band EQ | the attached Oratory HD650 10 band EQ).

My personal subjective preference is the Oratory EQ - ok no big deal.

However, what I am confused about is the shear difference between 1.Oratory and 2.Amrim / January122021. If the Harman target is the goal how are such huge differences part of any subsequent EQ attempt? Have others found these EQ's to be similar or different? I have attached the results of my Earful hearing test - but nothing stands out which would explain the gap between EQ's.

Apologies if this was answered or discussed in the previous 26 pages.....just very curious where the big difference is occurring.

Welcome to headphone calibration! It's almost an art. A slight movement of the headphone on a calibration rig can modify the EQ substantially. New vs worn out pads can change the EQ substantially. Substantial variance across headphones of the same model. Different shapes of ears can lead to different sound. There are so many variables. Why do you think pros use speakers for mixing? It's a lot more straightforward to get them to a reference sound - just build a 100k-1mil studio :).

Based on this, I trust the EQs made by those with the most experience. I think oratory's profiles are really good, and he has loads of experience measuring and correcting 100s of pairs of headphones. You can also learn yourself, then you can calibrate to your own ears - but as an amateur, your EQ may be less than ideal. But ultimately you'd have to tailor to your own ears to get reference sound. Or just don't worry about it and enjoy your headphones - ignorance is bliss as they say. If there isn't anything too major then it won't matter much unless you're producing music.
 

Maiky76

Senior Member
Joined
May 28, 2020
Messages
441
Likes
3,708
Location
French, living in China
The Oratory EQ (attached) sounds and looks quite different from this despite having the same Harman target.


I am wondering why? I have an old pair of HD650's (circa 2007) driven off a Benchmark DAC2-HGC with factory headphone gain settings.


After a few hours with each EQ filter set (Amrim's default 4 band EQ | @Maiky76's January122021-123909 9 band EQ | the attached Oratory HD650 10 band EQ).

My personal subjective preference is the Oratory EQ - ok no big deal.

However, what I am confused about is the shear difference between 1.Oratory and 2.Amrim / January122021. If the Harman target is the goal how are such huge differences part of any subsequent EQ attempt? Have others found these EQ's to be similar or different? I have attached the results of my Earful hearing test - but nothing stands out which would explain the gap between EQ's.

Apologies if this was answered or discussed in the previous 26 pages.....just very curious where the big difference is occurring.

Here is the comparison of the ASR and the Oratory data as well as the corresponding EQs.

The targets I used and Oratory's are identical, could have been stupid but easy explanation...

There are some significant differences which are difficult to explain but can probably be due the following factors:
- sample to sample variation
- measurement apparatus
- measurement procedure
Oratory_vs_ASR.png




For the EQed versions most of the deviation is caused difference around the 10k trough which is much more pronounced on the ASR data.
Depending on your very unit one E or the other might work better...

EDIT:

The graph presented shows the impact of the Oratory EQ in the bottom left and the comparison of the EQed version on the bottom right.
My calculation based on the Oratory data (scanned so there will be a small difference) and EQ are matching the data from the pdf attached by @Coach_Kaarlo which seems to be a graphical EQ with constant Q:
- Raw Oratory: 78 mine: 78.8
- EQed Oratory: 94 mine: 94.9

The 10k trough does impact the EQ through the score it produces (slope of the regression and standard deviation) because I do score optimization as well as fitting the target which is close but not equivalent to solely fitting the curve.
The "handicap" of the original data is carried over in the EQ score.
 

Attachments

  • Sennheiser HD650 (10 Band Graphic EQ).pdf
    318.4 KB · Views: 171
Last edited:

Robbo99999

Master Contributor
Forum Donor
Joined
Jan 23, 2020
Messages
6,875
Likes
6,673
Location
UK
Here is the comparison of the ASR and the Oratory data as well as the corresponding EQs.

The targets I used and Oratory's are identical, could have been stupid but easy explanation...

There are some significant differences which are difficult to explain but can probably be due the following factors:
- sample to sample variation
- measurement apparatus
- measurement procedure
View attachment 106875



For the EQed versions most of the deviation is caused difference around the 10k trough which is much more pronounced on the ASR data.
Depending on your very unit one E or the other might work better...
I think there's a mistake in your graphs, you're not showing Oratory's EQ result curve in any of your curves?

For me, the main takeaway from your graphs is the difference between the Oratory Measurement Curve and the Amir Measurement - the difference in the treble I mentioned earlier. And hence one of the reasons for the differences in the EQ's.

I don't think you need to worry about the 10kHz trough, it's supposed to be there in all measurements to a greater or lesser degree, I don't think it would have any impact on your EQ choices.
 

ishouldbeking

Active Member
Joined
Oct 8, 2019
Messages
119
Likes
179
Location
DC-adjacent
As a casual observer and new user of PEQ (I just got a Qudelix 5k for this purpose, as I can't currently use EQ on my computer), it's been fascinating to pore over this and similar threads and try different EQ profiles. I've now compared approaches from Oratory, Amir, AutoEQ, Solderdude, and metal571 (per his twitter, he prefers basic 2-band adjustments at 30hz and 3500k). Still need to try Maiky76's latest entrant.

Just for the sake of subjective commentary and nothing more: in the end I very much preferred Amir's settings, although I made a minor tweak to the low mids by ear, which were otherwise a tad bloated (though quite fun). Settled on cutting 2 db at 175hz, 1.414 Q; looking back at some of the other suggested profiles, this was pretty close to what others have done, too. If I had to describe the sound of this profile vs the others, I'd say the bass feels quite strong (in the most satisfying way) and the treble feels present and the most natural of any of the various EQ profiles.

One thing I noticed was that the profiles with more surgical EQ to the treble frequencies were quite interesting, but ultimately didn't feel as natural, at least to my ears, on my particular unit (which is a 2019 6xx made in Romania). I'm guessing variations in listening position, ear shape, and maybe measurement idiosyncrasies might lead us to create narrowly tailored EQ that works for a particular unit or listener but could introduce some slight weirdness in someone else's setup. For me, going "less is more" and using fewer bands and less aggressive tweaks overall led to the most natural and satisfying result. Either way, looking forward to experimenting and learning more.

EDIT: reading this back I realized I keep saying "natural", which isn't particularly descriptive. What I mean is the treble sounded clearer, with a seemingly smoother response overall, without sacrificing energy or air. Some of the other profiles seemed to introduce a bit of chaos--either a touch too much of something, which registers as bristling my inner ear, or the feeling of dead air where a frequency is zapped out of existence. But of course that's a subjective impression based on my ears, cans, source gear, and music, so YMMV.
 
Last edited:

infinitesymphony

Major Contributor
Joined
Nov 21, 2018
Messages
1,072
Likes
1,806
One thing I noticed was that the profiles with more surgical EQ to the treble frequencies were quite interesting, but ultimately didn't feel as natural, at least to my ears, on my particular unit (which is a 2019 6xx made in Romania). I'm guessing variations in listening position, ear shape, and maybe measurement idiosyncrasies might lead us to create narrowly tailored EQ that works for a particular unit or listener but could introduce some slight weirdness in someone else's setup. For me, going "less is more" and using fewer bands and less aggressive tweaks overall led to the most natural and satisfying result. Either way, looking forward to experimenting and learning more.
That's where I am as well regarding full-system EQ. Broad sculpting seems okay, trying to force a system into a particular box with overlapping EQs less so. Maybe it's the equipment I've tried it on, which has tended toward the lower end of the spectrum and thus has needed more EQ to match a target at the expense of headroom and potential distortion.

Your second point is well-taken, too; all of these corrections are good starting points at best, with production run and unit-to-unit variations, personal geometry, and preference to take into account for each individual user.
 

Ron Texas

Master Contributor
Joined
Jun 10, 2018
Messages
6,078
Likes
8,916
Here are EQ parameters for the HD650 transcribed from one of Oratory1990's PDF's into a form which can be read by Equalizer APO. They are a little different from our host's parameters:

-------------------------------------------------

Preamp: -8.5 dB
Filter 1: ON LS Fc 65 Hz Gain 8.5 dB Q 0.7
Filter 2: ON PK Fc 210 Hz Gain -2.3 dB Q 1.0
Filter 3: ON PK Fc 2150 Hz Gain 1.5 dB Q 2.2
Filter 4: ON PK Fc 3000 Hz Gain -2.2 dB Q 3.0
Filter 5: ON PK Fc 16000 Hz Gain -5.0 dB Q 1.0
Filter 6: ON HS Fc 3000 Hz Gain 3.0 dB Q 0.7
 

ishouldbeking

Active Member
Joined
Oct 8, 2019
Messages
119
Likes
179
Location
DC-adjacent
Here are EQ parameters for the HD650 transcribed from one of Oratory1990's PDF's into a form which can be read by Equalizer APO. They are a little different from our host's parameters:

-------------------------------------------------

Preamp: -8.5 dB
Filter 1: ON LS Fc 65 Hz Gain 8.5 dB Q 0.7
Filter 2: ON PK Fc 210 Hz Gain -2.3 dB Q 1.0
Filter 3: ON PK Fc 2150 Hz Gain 1.5 dB Q 2.2
Filter 4: ON PK Fc 3000 Hz Gain -2.2 dB Q 3.0
Filter 5: ON PK Fc 16000 Hz Gain -5.0 dB Q 1.0
Filter 6: ON HS Fc 3000 Hz Gain 3.0 dB Q 0.7

I think Oratory's profile has been updated a bit. I have a 10-band PEQ version of (what I believe to be his latest) 650 profile and it looks like this:

Pre: -10 db
1: Peak 23hz, 2.6db, Q 1.6
2: LShelf 70hz, 7.5db, Q .7
3: Peak 210hz, -2.2db, Q 1.0
4: Peak 2.15k, 2.0db, Q 2.2
5: Peak 3.1k, -1.1db, Q 3.4
6: Peak 4.27k, 3.0db, Q 4.0
7: Peak 5k, -0.8db, Q 6.0
8: Peak 5.7k, 1.8db, Q 3.0
9: Peak 6.6k, 0.9db, Q 4.0
10: Peak 11k, 3.0db, Q 1.3

See here: https://www.dropbox.com/s/zr5tqw0qojom9uh/Sennheiser HD650.pdf?dl=0
 

Ron Texas

Master Contributor
Joined
Jun 10, 2018
Messages
6,078
Likes
8,916
I think Oratory's profile has been updated a bit. I have a 10-band PEQ version of (what I believe to be his latest) 650 profile and it looks like this:

Pre: -10 db
1: Peak 23hz, 2.6db, Q 1.6
2: LShelf 70hz, 7.5db, Q .7
3: Peak 210hz, -2.2db, Q 1.0
4: Peak 2.15k, 2.0db, Q 2.2
5: Peak 3.1k, -1.1db, Q 3.4
6: Peak 4.27k, 3.0db, Q 4.0
7: Peak 5k, -0.8db, Q 6.0
8: Peak 5.7k, 1.8db, Q 3.0
9: Peak 6.6k, 0.9db, Q 4.0
10: Peak 11k, 3.0db, Q 1.3

See here: https://www.dropbox.com/s/zr5tqw0qojom9uh/Sennheiser HD650.pdf?dl=0

Both are updated. I looked at the 10 band as well but went with the 6 band due to less typing and a smaller preamp adjustment.
 

Docteur Poulet

Member
Forum Donor
Joined
Dec 3, 2020
Messages
35
Likes
59
Location
Switzerland
Hi folks,
I would like to buy a new open headphone.
I currently own a Sennheiser HD25 that I use daily, and I'm planning to had a new headphone for the living room.
I already own a RME ADI-2, are there any others valid options in Europe in the price range of the HD650?
 

Daverz

Major Contributor
Joined
Mar 17, 2019
Messages
1,294
Likes
1,451
Here are EQ parameters for the HD650 transcribed from one of Oratory1990's PDF's into a form which can be read by Equalizer APO. They are a little different from our host's parameters:

-------------------------------------------------

Preamp: -8.5 dB
Filter 1: ON LS Fc 65 Hz Gain 8.5 dB Q 0.7
Filter 2: ON PK Fc 210 Hz Gain -2.3 dB Q 1.0
Filter 3: ON PK Fc 2150 Hz Gain 1.5 dB Q 2.2
Filter 4: ON PK Fc 3000 Hz Gain -2.2 dB Q 3.0
Filter 5: ON PK Fc 16000 Hz Gain -5.0 dB Q 1.0
Filter 6: ON HS Fc 3000 Hz Gain 3.0 dB Q 0.7

I'm trying to reproduce these various EQs in rePhase. Should I use constant or proportional Q? How about minimum vs linear phase?

I did try Amir's filters (minimum phase, constant Q, 16384 taps), and it sounds pretty good. I was afraid the bass would be overwhelming (knowing what a bass head Amir is), but it's not at all.
 

ishouldbeking

Active Member
Joined
Oct 8, 2019
Messages
119
Likes
179
Location
DC-adjacent
Hi folks,
I would like to buy a new open headphone.
I currently own a Sennheiser HD25 that I use daily, and I'm planning to had a new headphone for the living room.
I already own a RME ADI-2, are there any others valid options in Europe in the price range of the HD650?
Well besides the sibling headphones the HD600 and 660... I'm not sure about EU pricing specifically, but perhaps something like Hifiman Sundara, Beyerdynamic DT880 (probably cheaper) or DT1990 (probably more expensive), or AKG K712.
 

Ron Texas

Master Contributor
Joined
Jun 10, 2018
Messages
6,078
Likes
8,916
I'm trying to reproduce these various EQs in rePhase. Should I use constant or proportional Q? How about minimum vs linear phase?

I did try Amir's filters (minimum phase, constant Q, 16384 taps), and it sounds pretty good. I was afraid the bass would be overwhelming (knowing what a bass head Amir is), but it's not at all.

I always use constant Q and minimum phase. There are long discussions about whether to use minimum phase or linear dealing with ringing. To use shelf filters in Rephase it's necessary to have more than one bank of filters. I have tried both the FIR filters and just using Equalizer APO with Foobar2000 and can't say there is a difference, sighted testing. Actually, the Oratory1990 parameters give a bit more boost around 100 hz.

Enjoy the music.
 

Coach_Kaarlo

Active Member
Joined
Apr 27, 2020
Messages
196
Likes
222
Location
Sydney
Thanks for the insights.

The explanations make sense, but I am left wondering at the value of providing a single EQ filter set - which differs in application and philosophy so much, to be used on ears and hearing which also differ so much, to satisfy the preferences of the wider listening community which differ so much.

Surely if neutral and clear and accurate are the goals we would be closer with the targets?

https://drive.google.com/file/d/1cAFHrvUMkMUf78qZyalxEEGbj1ToX2Cb/view?usp=sharing

So, to try and remove some of my own circle of confusion I want to test a few things.

I want to end up with an EQ which addresses the errors in the HD650 frequency response and tonality. AND also corrects particular frequencies for my own hearing loss. What is the best approach;
  1. Using speakers in nearfield or free field conditions to run the Earful app (as it seems using headphones with EQ introduces another factor between signal and hearing).
  2. Or do I EQ the headphones and then do the hearing test and then adjust the EQ to achieve the desired response?

Seems I have much to learn about this, apologies if these are questions answered elsewhere.
 

Coach_Kaarlo

Active Member
Joined
Apr 27, 2020
Messages
196
Likes
222
Location
Sydney
I think Oratory's profile has been updated a bit. I have a 10-band PEQ version of (what I believe to be his latest) 650 profile and it looks like this:

Pre: -10 db
1: Peak 23hz, 2.6db, Q 1.6
2: LShelf 70hz, 7.5db, Q .7
3: Peak 210hz, -2.2db, Q 1.0
4: Peak 2.15k, 2.0db, Q 2.2
5: Peak 3.1k, -1.1db, Q 3.4
6: Peak 4.27k, 3.0db, Q 4.0
7: Peak 5k, -0.8db, Q 6.0
8: Peak 5.7k, 1.8db, Q 3.0
9: Peak 6.6k, 0.9db, Q 4.0
10: Peak 11k, 3.0db, Q 1.3

See here: https://www.dropbox.com/s/zr5tqw0qojom9uh/Sennheiser HD650.pdf?dl=0

Actually.....yours is from Feb 2019, the one I have is from October 2019.


Seems like he changed the bass correction to get rid of that distorted flab which some of the HD650 EQ's have.
 

Attachments

  • Sennheiser HD650.pdf
    300.2 KB · Views: 204
  • Sennheiser HD650 (6 Band EQ).pdf
    292.6 KB · Views: 196
  • Sennheiser HD650 (10 Band Graphic EQ).pdf
    318.4 KB · Views: 137
Top Bottom