Sennheiser HD800S Review (Headphone)

[MacClintock]

Looks good. Here it is compared to Oratory’s with the horizontal and vertical scales adjusted.

View attachment 289764

Very similar indeed. I have also a version where I boost the bass shelf addionally (i.e. even above Harman) by a few dB and for some music I have the impression that audible distortion kicks in, which would be my only complaint of the HD800, but which is explainable by Amir's measurements.

 

[Phoney]

Thinking shouty mids, but can you give some numbers, please?

It's pretty similar to the oratory one, but leaning more towards crinacles measurements as he had 2 units that measured similarly to eachother. It's still based on the harman curve.

 

[Phoney]

Very similar indeed. I have also a version where I boost the bass shelf addionally (i.e. even above Harman) by a few dB and for some music I have the impression that audible distortion kicks in, which would be my only complaint of the HD800, but which is explainable by Amir's measurements.

Mine boosts the bass by the same (or maybe 0,5db more) than the oratory one in reality, it's just that his EQ pulls everything down by around a couple of db more and therefore use less pregain. Effectively it's pretty similar if you compare them post pregain.

 

[Phoney]

Looks good. Here it is compared to Oratory’s with the horizontal and vertical scales adjusted.

View attachment 289764

Cool! Mind if you make a new one with the exact numbers that i posted? It would be cool to see the exact comparison.

 

[Presently42]

EQ any headphone to the original frequency response of the HD 800 and I am pretty sure you will not replicate the soundstage of it. So specific frequencies certainly have an effect, but it is minor in comparison to the rest.

You are quite right. I have indeed performed this experiment - and also found there to be not much difference in spatial effects. However, I did notice a fairly large difference when I removed the spatial effects bands from the HD800. I suspected (and still do, at least in part), that spatial effects were correlated with reflections in the headphone-ear system - which explains why spatial effects can't easily be created by altering frequency response; but can be removed by altering the frequency response - and could therefore be predicted by group delay, especially in certain frequency regions. Sadly, the group delay seems also to not predict spatial effects. The search for quantifiable measurements which predict and explain spatial effects continues!

 

[Chagall]

It's pretty similar to the oratory one, but leaning more towards crinacles measurements as he had 2 units that measured similarly to eachother. It's still based on the harman curve.

View attachment 289800

The scale threw me off. will try it, tnx

 

[Zensō]

Cool! Mind if you make a new one with the exact numbers that i posted? It would be cool to see the exact comparison.

Here you go...

 

[Presently42]

In my experience, using Oratory’s EQ through Pro-Q3 running in SoundSource sounds great, with little to no change in stereo imaging or soundstage.

Just in case you're not aware, ProQ's Q values are not the same as those provided by oratory1990: oratory1990's must be equated to ProQ's with an annoying equation, dependant on frequency, bitrate and Q value: Qp = Qo*(2*π*(f/b))/sin(2*π*(f/b))*√(2) , where Qp is ProQ's Q value, Qo is oratory1990's Q value, f is frequency and b is bitrate. This equation is seemingly only valid for peaking filters: I've not yet discovered the equation for shelving filters

 

[Zensō]

Just in case you're not aware, ProQ's Q values are not the same as those provided by oratory1990: oratory1990's must be equated to ProQ's with an annoying equation, dependant on frequency, bitrate and Q value: Qp = Qo*(2*π*(f/b))/sin(2*π*(f/b))*√(2) , where Qp is ProQ's Q value, Qo is oratory1990's Q value, f is frequency and b is bitrate. This equation is seemingly only valid for peaking filters: I've not yet discovered the equation for shelving filters

Thanks, yes I'm aware of the need to convert the Q values. Multiplying by 1.414 gets you very close, close enough to not matter for the imprecise art of headphone EQ.

 

[MacClintock]

You are quite right. I have indeed performed this experiment - and also found there to be not much difference in spatial effects. However, I did notice a fairly large difference when I removed the spatial effects bands from the HD800. I suspected (and still do, at least in part), that spatial effects were correlated with reflections in the headphone-ear system - which explains why spatial effects can't easily be created by altering frequency response; but can be removed by altering the frequency response - and could therefore be predicted by group delay, especially in certain frequency regions. Sadly, the group delay seems also to not predict spatial effects. The search for quantifiable measurements which predict and explain spatial effects continues!

I agree. Amir also originally suspected the group delay revealing anything relevant about spatial effects but abandoned that hypotheses after many measured headphones. So, the search continues.

 

[skam22]

Senheiser claims the Frequency Response is: 4 - 51,000 Hz.
I wonder how they can measure anything below 20Hz?

 

[staticV3]

I wonder how they can measure anything below 20Hz?

With a microphone. Those can easily measure down to 4Hz.

Edit: "4 - 51,000 Hz" is still an idiotic spec though.

 

[MacClintock]

With a microphone. Those can easily measure down to 4Hz.

Edit: "4 - 51,000 Hz" is still an idiotic spec though.

Yes, the Earthworks M50 does this. Not cheap, though.

 

[solderdude]

Even some cheap electret capsules can measure down to a few Hz.

The Earthworks can do the 50kHz, cheap electrets and even expensive HATS/couplers can not.

The 50kHz is indeed idiotic, the 4Hz is easily measurable, the -10dB is questionable though (depends on the target)

 

[staticV3]

The Earthworks can do the 50kHz, cheap electrets and even expensive HATS/couplers can not.

This one goes to 80kHz: https://micbooster.com/ultrasonic-microphones/146-ultrasonic-mic-board.html

 

[solderdude]

Yes... +/- 10dB

Strangely the preliminary specs say 20Hz - 10kHz +/- 2dB but is already -2dB at 100Hz (can't imagine 10Hz) and 24kHz +10dB and 80kHz -6dB
But... I have to get one now for the 30-100kHz range... we all know how important that range is to us MP3 playing BT headphone guys

 

[MacClintock]

With a microphone. Those can easily measure down to 4Hz.

Edit: "4 - 51,000 Hz" is still an idiotic spec though.

Wait, but what if some alien with extended hearing wants a headphone recommendation? Or your dog?

 

[DualTriode]

I am aware of the fact that Jens Blauert did write one of the most relevant books on psychoacoustics and is (or was, now retired) a leading expert in the field. But still, simply make the opposite experiment. EQ any headphone to the original frequency response of the HD 800 and I am pretty sure you will not replicate the soundstage of it. So specific frequencies certainly have an effect, but it is minor in comparison to the rest.

You Have it backwards.

According to the Harman research frequency response comes first, other variables are secondary.

Thanks DT

 

[IAtaman]

You Have it backwards.

According to the Harman research frequency response comes first, other variables are secondary.

Thanks DT

Is there a sound stage research from Harman?

 

[Robbo99999]

Is there a sound stage research from Harman?

Nope, I don't think so, we're on our own with that, hence I support your rhetorical question in defence of @MacClintock 's post. (It does exist, it's not defined).