Skip to main content ↑ ↓

Notice

Please note that most of the software linked on this forum is likely to be safe to use. If you are unsure, feel free to ask in the relevant topics, or send a private message to an administrator or moderator. To help curb the problems of false positives, or in the event that you do find actual malware, you can contribute through the article linked here.
Topic: Toneboosters Morphit Headphone simulation accuracy? (Read 9325 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

Toneboosters Morphit Headphone simulation accuracy?

I've been playing around with the Headphone simulation feature in the Toneboosters Morphit VST plugin through Foobar2000.
I'm using Sennheiser HD280 PRO headphones and selected 'Sennheiser HD280Pro' as reference and scrolled through all the different headphone models that can be emulated.

When I selected the 'Shure SRH940' headphones I was amazed how much 'fun' the audio quality got. Now I am considering buying a pair, but i'm not sure how accurate this plugin is.
Some of the Stax headphones also sounded quite okay.

Anyone here know if this plugin is accurate in the headphone emulation?

Re: Toneboosters Morphit Headphone simulation accuracy?

Reply #1 –
Quote from: LithosZA on

Anyone here know if this plugin is accurate in the headphone emulation?

Be advised that there are audible differences based on acoustical and perceptual differences in how headphones and earphones interface with individual real world humans. 

People who fit hearing aids have a lot of wisdom about these things, and sometimes you can find them and read them. They are very sobering.

Bottom line is that every pair of headphones sounds different to every human that ever uses them. Simulations can't possibly be aware of how they work out for every individual.

What I take away from this is that if you don't use an equalizer that is tuned differently for every different person and every set of headphones, you are condemning yourself to suboptimal performance.  If you do, the sound quality of individual headphones becomes a lot less important to you because you then use the equalizer to get what you want.

Re: Toneboosters Morphit Headphone simulation accuracy?

Reply #2 –
I agree with Arnold that it is probably pretty hard to get this simulation accurate for everyone due to the effects of your ears on the response of the headphones.  I'd guess that overall its probably more accurate for IEMs than circumaural headphones, although that is just speculation. 

Re: Toneboosters Morphit Headphone simulation accuracy?

Reply #3 –
Quote
When I selected the 'Shure SRH940' headphones I was amazed how much 'fun' the audio quality got. Now I am considering buying a pair,
You might want to find a store that sells them and give them a listen.  It's also a good idea to try-out the fit/comfort, especially if you listen with headphones for extended periods.

Or continue to use the plug-in, or try to match the sound with regular 'ol EQ...   Or you might find an EQ setting that you like even more!

Re: Toneboosters Morphit Headphone simulation accuracy?

Reply #4 –
I have a question here. Doesn't the sound of headphones come down to frequency response differences (assuming relative absence of distortions/resonances)? I sometimes use the Sonarworks Plugin for mastering. When it linearizes the frequency reponses of the HD-600 or the ATH-M50x, they both sound pretty much alike to my ears.
And didn't Sean Olive use just one headphone to simulate others? The only difference being frequency responses? How accurate does any simulation need to be to reach the desired effect?
marlene-d.blogspot.com

Re: Toneboosters Morphit Headphone simulation accuracy?

Reply #5 –
Quote
I have a question here. Doesn't the sound of headphones come down to frequency response differences
Yes.
Quote
Morphit is the result of a painstaking and meticulous measurement effort to capture the frequency response of headphones, and correcting them using scientifically-proven target curves.

Of course, the trick is measuring all of the headphones under identical conditions.      And as Arny & Saratoga pointed out, there are unknown differences between listeners so matched/corrected headphones may not sound exactly alike to all listeners.

Re: Toneboosters Morphit Headphone simulation accuracy?

Reply #6 –
Quote from: DVDdoug on
Quote
I have a question here. Doesn't the sound of headphones come down to frequency response differences
Yes.
Quote
Morphit is the result of a painstaking and meticulous measurement effort to capture the frequency response of headphones, and correcting them using scientifically-proven target curves.

Of course, the trick is measuring all of the headphones under identical conditions.      And as Arny & Saratoga pointed out, there are unknown differences between listeners so matched/corrected headphones may not sound exactly alike to all listeners.

It may sound klike hair-splitting defensivenesss, but I think that calling these differences unknown is very misleading.

The differences are knowable by means of doing something like a hearing test.  You measure the same ears with different headphones and come up with different frequency response curves. You measure them by some reasonable means that does not directly involve human hearing. You again come up with different frequency response curves. The differences between the responses of the headphones when evaluated by ears and by standard measurements is the difference due to the differences in the ears.

These differences are known to exist, and the technical means for addressing them are known and shared  by experienced practitioners in the art of fitting hearing aids.

IME these differences are so strong that ignoring them leads people on an inherently futile search for their ideal headphones or earphones.  This situation is partially responsible for the fact that many people have many pairs of headphones.

The solution to this problem is finding some means whereby people can tailor their own headphone's response in a way that includes these differences.

Re: Toneboosters Morphit Headphone simulation accuracy?

Reply #7 –
Quote from: Cavaille on
Doesn't the sound of headphones come down to frequency response differences (assuming relative absence of distortions/resonances)?
If this assumption is correct, the answer must almost be yes because the frequency response is probably the only difference.
If I look at headphone measurements there are differences on many aspects be it thd, impulse, etc
https://www.innerfidelity.com/images/AudezeLCD4.pdf
https://www.innerfidelity.com/images/SennheiserHD800.pdf
Maybe in  well-designed headphones the frequency response is the one that really makes the audible difference.


Anyway, had some fun with Morphit

First I let it correct the response of the HD800
Makes a slight difference.
The I used the compare functionality by emulating a Etymotic ERP4P
Indeed the sound changes and get that lean and mean characteristic of the ERP4P
I had the feeling it is a bit overdone.
Than I did the opposite, using the ERP4P, I let Morphit emulate the HD800
Indeed you get the wider soundstage, a bit warmer sound but I’m not really convinced  this sounds exactly like the HD800.
But I do think it gives you at least a taste of how other headphones sound.
Nice toy.
TheWellTemperedComputer.com

Re: Toneboosters Morphit Headphone simulation accuracy?

Reply #8 –
Quote from: Arnold B. Krueger on
The differences are knowable by means of doing something like a hearing test.  You measure the same ears with different headphones and come up with different frequency response curves. You measure them by some reasonable means that does not directly involve human hearing. You again come up with different frequency response curves. The differences between the responses of the headphones when evaluated by ears and by standard measurements is the difference due to the differences in the ears.

But are these differences really that huge? If I assume an imagined "100%" frequency response, then a listener that deviates at max -20% and another listener deviating at max +20%, both must be well within the parameters of a desired headphone sound. And when it comes down to it, don't distortions effectively change the actual or perceived frequency response? So if a headphone distorts, this would surely show up on the frequency response plot. Or am I wrong?

Quote from: Arnold B. Krueger on
These differences are known to exist, and the technical means for addressing them are known and shared  by experienced practitioners in the art of fitting hearing aids.

I don't doubt that. But isn't a hearing aid fundamentally different to a headphone that is only aiming at music reproduction? If I remember correctly, a hearing aid concentrates on frequencies around the spectrum of the human voice and pretty much ignores very low and high frequencies. If that is wrong, please correct me.

Quote from: Arnold B. Krueger on
IME these differences are so strong that ignoring them leads people on an inherently futile search for their ideal headphones or earphones.  This situation is partially responsible for the fact that many people have many pairs of headphones.

I have, too. But not for that reason. It's just that I'm stupid and like to collect unnecessary things. My most used headphone is a compromise. It sits extremely comfortable, gets very loud with little gain but sounds fairly unbalanced so I use an EQ on my smartphone to linearize it. Works very well and turns a mediocre headphone into a good one.
marlene-d.blogspot.com

Re: Toneboosters Morphit Headphone simulation accuracy?

Reply #9 –
Quote from: Roseval on
Than I did the opposite, using the ERP4P, I let Morphit emulate the HD800
Indeed you get the wider soundstage, a bit warmer sound but I’m not really convinced  this sounds exactly like the HD800.

A wider soundstage? So this plugin does more than adapting frequency responses? It certainly sounds like it includes a stereo widening function. That sounds more like a simple effect tool to "make-music-sound-'better'".
marlene-d.blogspot.com

Re: Toneboosters Morphit Headphone simulation accuracy?

Reply #10 –
http://asa.scitation.org/doi/10.1121/1.4984044
If I understand this article correctly, a PCA analyses yielded 6 eigenvalues. All of them frequency related.
I see no mentioning of other effects.
Now I must admit that I don’t understand PCA nor eigenvalues but it gives me the feeling it is based on frequency response only.
TheWellTemperedComputer.com

Re: Toneboosters Morphit Headphone simulation accuracy?

Reply #11 –
I've tried Morphit, I've used it so far to simulate headphones I own / owned. The result is that the simulated headphones do not sound like their real-world counterpart. Headphones currently in use must naturally be corrected first to have a flat frequency response so that they can be used to simulate other headphones. The target curve seems to be the problem, it sounds as if they used Diffuse Field EQ instead of a more recent approach (Olive). They don't say what they use (or I couldn't find it).

Another problem is manufacturers changing models during production. The HD-380 as measured by them does have a large bump around 2500 Hz and the simulation sounds like it. The one I own apparently lacks this. The same happens when the ATH-M50x is simulated. The unit Tonebooster measured lacks an audible peak around 10 kHz that is present on my unit.

From what I could find, Morphit only uses an EQ to correct/simulate headphones. It also contains a selectable limiter. I couldn't see any stereo widening going on.
marlene-d.blogspot.com

Re: Toneboosters Morphit Headphone simulation accuracy?

Reply #12 –
Thanks for the responses. So it seems that I can't really use the tool to figure out if I will like certain headphones or not.
I wonder if they have the frequency response correct for the Sennheiser HD280 PRO, because if I simulate most of the headphones then the bass is boosted unnaturally high that drowns out the music.

Re: Toneboosters Morphit Headphone simulation accuracy?

Reply #13 –
Quote from: Cavaille on
I've tried Morphit, I've used it so far to simulate headphones I own / owned. The result is that the simulated headphones do not sound like their real-world counterpart. Headphones currently in use must naturally be corrected first to have a flat frequency response so that they can be used to simulate other headphones. The target curve seems to be the problem, it sounds as if they used Diffuse Field EQ instead of a more recent approach (Olive). They don't say what they use (or I couldn't find it).

Another problem is manufacturers changing models during production. The HD-380 as measured by them does have a large bump around 2500 Hz and the simulation sounds like it. The one I own apparently lacks this. The same happens when the ATH-M50x is simulated. The unit Tonebooster measured lacks an audible peak around 10 kHz that is present on my unit.

From what I could find, Morphit only uses an EQ to correct/simulate headphones. It also contains a selectable limiter. I couldn't see any stereo widening going on.
I don't see why the reference signature should matter? we're dealing with relative measurements/corrections anyway and a variation can be registered properly no matter if there is an initial compensation applied or not(as long as all measurements are done the same way).
now how their choice of neutral isn't yours, well that's the logical result of headphones. as said by others, different people will need a different response. it's not like speakers.

about morphit using EQ only, that would probably mean they relied on third party measurements of FR. because if I was doing measurements myself, I would make impulses and use convolution. it doesn't take any more efforts and contains a good deal more than just FR. I can't think of a reason to measure headphones and decide to only apply basic EQ as correction.

about stereo widening, you've misunderstood Roseval's post. it's not like the hd800 has a stereo DSP. ^_^

as for manufacturing and measurement stability, sadly there is little we can do about it. it just adds up with personal fit, size of the head, and stuff like that.  also IDK how they decided to set their corrections:
-apply a correction based on FR graph and hope for the best(neglecting resonance and disto)?
-measure the compensated result and adjust until the FR is right, no matter how much distortions that could create?
-maybe EQ only a limited range instead of the all audible range to avoid the rolled off part most of the time?
-maybe it's a case by case response checking a given distortion level or some subjective preference from the guy doing the work?
any choice will have pros&cons and more or less success depending on the gear used. but all can result in some audible extra variations.


Quote from: LithosZA on
Thanks for the responses. So it seems that I can't really use the tool to figure out if I will like certain headphones or not.
I wonder if they have the frequency response correct for the Sennheiser HD280 PRO, because if I simulate most of the headphones then the bass is boosted unnaturally high that drowns out the music.
of course you can, but you have to keep in mind that like Roseval mentioned, you will probably get a taste more than the exact actual experience of using a given headphone yourself.  Morphit is trying to give a general idea of the signature, which for a user might but more relevant than the poetic ramblings of amateur owners and sponsored reviewers. they too are rarely 100% accurate, and that's saying it nicely. ^_^