Custom Target Response Scoring?
I’ve been loving the “My Ratings” feature for creating custom lists of headphones to see what I might like, however the biggest issue that I’ve come across is that the target response curve score doesn’t match my preference, and so the scores that might considered lower by Rtings would actually be my preference for sound.
I would love to be able to organize the headphones by which one best fits my target response preference, instead of which one just fits the standard Harman curve the best.
(I understand that this would probably be quite the task to implement, so I would understand if this never happens. But I still think the idea is incredibly useful for making it easier to find new headphones.)
- 32120
Hi danegraphics,
Never apologize for adding more to the suggestion pile! As I’ve mentioned before, we really do appreciate user feedback and suggestions! Especially when you’re putting a noticeable amount of thought and effort into your posts! Messages like these really do help us continue to improve our reviews and are testing, so just know that it doesn’t go unnoticed! I can totally understand why you would be advocating for a test like that, I’ll definitely add it to our test bench suggestions for the test devs to consider in a future update! :)
- 21010
Dynamic range compression in passive headphone drivers? Unheard of. Is there any published study or is it more audiophile BS?
The amount of power used in headphones isn’t even 0.25 mW at normal listening levels. Although, DR compression is employed in Bluetooth headphones and automobile head units to protect the speakers.
- 21010
Dynamic range compression in passive headphone drivers? Unheard of. Is there any published study or is it more audiophile BS?
There are a few studies and online discussion about how it applies to both loudspeakers and headphones. (Example with loudspeakers: http://www.klippel.de/fileadmin/_migrated/content_uploads/Loudspeaker_Nonlinearities–Causes_Parameters_Symptoms_01.pdf Discussion of headphones on a forum: https://www.audiosciencereview.com/forum/index.php?threads/an-attempt-to-measure-headphone-non-linearity.18937/ ) Sadly though, it doesn’t appear to be much looked into.
The word that would be used would be “nonlinearities” which actually applies to a few different aspects of how drivers function, but the one I’m interested in is how the amplitude’s relationship to the power is non-linear, resulting in slight compression in many cases, especially at higher volumes, with lower frequencies and transients.
Now, it could be completely meaningless when it comes to most headphones and I could be imagining the whole thing, but I would not be surprised if cheaper headphone drivers displayed more nonlinearities in amplitude vs power compared to more expensive headphones, and I think its still something worth looking into.
Edited 3 years ago: fixed sentence - 21010
There are a few studies and online discussion about how it applies to both loudspeakers and headphones. (Example with loudspeakers: http://www.klippel.de/fileadmin/_migrated/content_uploads/Loudspeaker_Nonlinearities–Causes_Parameters_Symptoms_01.pdf Discussion of headphones on a forum: https://www.audiosciencereview.com/forum/index.php?threads/an-attempt-to-measure-headphone-non-linearity.18937/ ) Sadly though, it doesn’t appear to be much looked into. The word that would be used would be “nonlinearities” which actually applies to a few different aspects of how drivers function, but the one I’m interested in is how the amplitude’s relationship to the power is non-linear, resulting in slight compression in many cases, especially at higher volumes, with lower frequencies and transients. Now, it could be completely meaningless when it comes to most headphones and I could be imagining the whole thing, but I would not be surprised if cheaper headphone drivers displayed more nonlinearities in amplitude vs power compared to more expensive headphones, and I think its still something worth looking into.
Actuallty, Rtings already test headphones for non lineareties. Because of a bunch of laws about orthogonalety between frequencies, non linear distortion appear at a multiple of the frequency being played. These are refered to as harmonic distortions and is something Rtings measure and show.
- 21010
Actuallty, Rtings already test headphones for non lineareties. Because of a bunch of laws about orthogonalety between frequencies, non linear distortion appear at a multiple of the frequency being played. These are refered to as harmonic distortions and is something Rtings measure and show.
Again, that’s not the information I’m looking for. There are multiple types of non-linearities, and that is only one of them. I’m looking specifically at the amplitude to power relationship, or the output amplitude to input amplitude relationship.
And sure, it’s partially encoded in some of the other measurements, but only partially, and in an incredibly obfuscated way.
I’m looking for a graph similar to this, where I can see at what volume we start to see amplitude compression.
Edited 3 years ago: removed a period - 10-11-1
There are a few studies and online discussion about how it applies to both loudspeakers and headphones. (Example with loudspeakers: http://www.klippel.de/fileadmin/_migrated/content_uploads/Loudspeaker_Nonlinearities–Causes_Parameters_Symptoms_01.pdf Discussion of headphones on a forum: https://www.audiosciencereview.com/forum/index.php?threads/an-attempt-to-measure-headphone-non-linearity.18937/ ) Sadly though, it doesn’t appear to be much looked into. The word that would be used would be “nonlinearities” which actually applies to a few different aspects of how drivers function, but the one I’m interested in is how the amplitude’s relationship to the power is non-linear, resulting in slight compression in many cases, especially at higher volumes, with lower frequencies and transients. Now, it could be completely meaningless when it comes to most headphones and I could be imagining the whole thing, but I would not be surprised if cheaper headphone drivers displayed more nonlinearities in amplitude vs power compared to more expensive headphones, and I think its still something worth looking into.
OK. The first paper from Klippel is a technical note not something published in a journal. I don’t have a degree in electroacoustics, so impossible to comment on the content. The only thing I can infer from it is that the paper discusses the driver parameters that might lead to non linearities.
The forum discussion is inconclusive. There isn’t anything meaningful there.
The most important question is whether these non-linearities are audible. We already measure THD though IMD isn’t measured for transducers as it is not known to vary much AFAIK. For THD, we know the levels at which it becomes audible and at normal listening levels, it isn’t an issue. This is well researched.
- 10-11-1
Again, that’s not the information I’m looking for. There are multiple types of non-linearities, and that is only one of them. I’m looking specifically at the amplitude to power relationship, or the output amplitude to input amplitude relationship. And sure, it’s partially encoded in some of the other measurements, but only partially, and in an incredibly obfuscated way. I’m looking for a graph similar to this, where I can see at what volume we start to see amplitude compression.
Amplitude to power relationship is nothing but sensitivity which is already published for all headphones. The FR graph also shows the sensitivity with respect to frequency. Isn’t it?
- 21010
Amplitude to power relationship is nothing but sensitivity which is already published for all headphones.
No. That’s a single number. It tells us nothing about how higher amplitudes are effected as the power increases.
The FR graph also shows the sensitivity with respect to frequency. Isn’t it?
No. I’m not talking about volume, THD, frequency, or sensitivity. I’m talking about the displacement of the diaphragm of the driver relative to the power of the input signal.
Here’s a more mechanical explanation:
Every driver has limitations to how far forward or back the diaphragm can move. It has restoring forces that pull it back toward a neutral position. Those forces increase as the diaphragm’s position approaches the limits of its suspension system.
This means that as the input signal increases linearly, the diaphragm will follow the input signal less and less as the restoring force of the suspension system pulls it back toward neutral. Once the diaphragm hits the limit of the suspension system, it stops, even if the input signal continues to increase.
What I’m asking for in physical terms is what the displacement of the diaphragm looks like compared to a range of power in the signal. A well designed driver will displace almost perfectly linearly for the entire range of amplitudes that correspond to standard listening volumes (even up to 100db and higher). A lower quality system will be non-linear within standard listening volumes, and in that case, the audio is compressed (not in the data compression sense, but the audio compression sense), meaning that the peaks of the signal are brought down, which reduces detail and dynamic range, and also causes the instruments to be difficult to discern individually. This effect can also differ depending on the frequency.
This happens mechanically one way or another with every driver type, whether on speakers or on headphones. My question is to what extent on different headphones.
You can get a sense of the effect of this by looking up how audio compression works. You’ll probably see a graph like this which shows that past a certain point in the input level, the output level is brought down so that the peaks of the sound are flattened. This is usually done so that the perceived volume of a track or section of a track can be increased without clipping the audio, and is used in “normalizing” audio tracks for things like radio broadcast, podcasts, and so on where the recordings of voices can have a wide dynamic range but you want it all to play at the same perceived volume. (a basic google search will bring up some good resources to start learning about audio compression and normalization).
Edited 3 years ago: forgot ly - 21010
I understand what you are saying, but why is that a concern as a consumer? The designers of the driver worry about non-linearity and the driver is rated for a range of power inputs under which the driver exhibits linear relationship with the input power.
Also, most music has a maximum dynamic range of about 30 dB. Even 30 dB is very rare.
Even if we assume that the quietest part of music is placed at 70 dB SPL and a DR of 30 dB, the driver reproduces sound in the range of 70 to 100 dB SPL which is normal operating range for a driver.
If we the place the quietest part of music at 80 dB SPL, then the peak hits 110 dB at which a driver might exhibit non-linearity, but it isn’t concern as peaks in music are only for fraction of a second. Due to loudness war, the DR of modern music is even less that DR compression isn’t an issue.
- 21010
I understand what you are saying, but why is that a concern as a consumer?
My concern is that it sounds, to my ears at least, like many lower quality drivers are non-linear within standard listening volumes, and that that has a major impact on the quality of the sound. As I described in my first post, the dynamic range of the SHP9500 sounds compressed compared to higher quality headphones.
Also, more than half of the music I listen to has a wide dynamic range (especially classical music). I’d rather that range be accurate to some degree rather than all at the same volume.
Without measurements to prove my ears wrong (which they just might be), I can’t simply assume that all headphone drivers are perfectly linear at standard listening volumes, especially when it doesn’t sound like it.
Edited 3 years ago: word - 10-11-1
SHP9500 sounds compressed
Actually, Philips had a very competent team(in Netherlands) that designed the drivers for SHP9500 and its predecessors. I remember reading some article where they explained how they used a laser to track the surface of the diaphragm to see how it responds.
SHP9500 is an amazing headphone and I do use them regularly. If you are below the age 30, EQ the 4-8kHz range, above 40+, even that EQ might be unnecessary.
If the headphone was designed by a major manufacturer, you could be rest assured that these non-linearities have already been taken care of.
It is important to enjoy the music rather than worry about some unknown technical detail. For example, what would you do about thermal compression effect due to heat generated in the voice coils? Listen for 10 minutes, then cool it and listen again? 😂
- 21010
Actually, Philips had a very competent team(in Netherlands) that designed the drivers for SHP9500 and its predecessors. I remember reading some article where they explained how they used a laser to track the surface of the diaphragm to see how it responds.
Then who knows what I’m hearing. The lack of dynamic range could all be in my head. But measurements would be nice, and I have never been able to find any.
It would be great if someone could show some, even if they are completely uninteresting.
If the headphone was designed by a major manufacturer, you could be rest assured that these non-linearities have already been taken care of.
There are no guarantees of that. In fact, major manufacturers are more likely to cheap out on the cheaper products.
Engineering is all about getting within certain tolerances, and they may care about some non-linearities more than others, and they care about those tolerances less when the product they’re making is cheaper too.
SHP9500 is an amazing headphone and I do use them regularly.
I wholeheartedly agree. They are fantastic headphones for their price. But there is something about them that makes them significantly lower quality than HD600’s or Sundara’s, and it’s definitely not the frequency response or the harmonic distortion or anything else that Rtings currently measures.
Rtings has them rating higher than the HD600’s in the neutral sound quality category, which is obviously untrue with any A/B listening. So there is some measurement of quality that they are missing.
It is important to enjoy the music rather than worry about some unknown technical detail.
Sadly, I am cursed with an obsession for unknown technical details, especially when they make a noticeable difference in the quality. My enjoyment of music is forever hindered by my nitpicking, and it may never be unhindered again. Truly a tragedy.