Well it kinda does! Lets just say you are more outside the box than I am! I hope it works out and please post your results.
Rob 🙂
You haven't used Class D when you didn't try the newer Hypex amps. The modules with power supply are really easy to use.
The area from 2-5k is very often covered by one driver - the tweeter ;-) You just need a good model which is capable of doing so and a very good crossover. And I still like small distances between drivers.
Especially for that (and in combination with hard dome midranges) the Hypex amps are very good. When you want to hear the truth about the rest of the chain.
They based this curve on 11 (!!!) listeners? They even dare to publish this?
I confess to never having used hypex stuff aside from some DSP. They're probably on the same level as Icepower. I do like the bass from my SpeakerPower Ice modules. The low end is very transparent and without artifacts I've noticed with most higher power class AB pro audio. What makes a good plate amp even better is a quality input transformer ie. Jensen or Lundahl.
The bass levels in modern studio produced music is obscenely high, as is the lack of dynamics and outright distortion. I remember buying the LP vinyl version of Paul McCartney Memory Almosy Full. It sounded so bad that I never listened to it again and gave it to someone. The distortion was unbearable and the bass was just off the chart. Why PMC would approve of this type of sound is beyond my comprehension.
The reasoning a company like Harman is using to come up with this curve of theirs is just silly. No self respecting audio fan would tolerate this sound. Its so removed from any hint of reality. The problem is then how much worse it will sound on a system which doesn't hide anything rather than smear and melt it together in one big blob of noise.
The main reason people feel they can push the level so hard is the lack of gradual tape distortion onset when recording fully digital. The tape saturation is what actually helped the sound or tone of some instruments when applied judiciously. Drums can sound really good when recorded onto analog tape this way.
Digital clipping is however simply awful and has no place in music. The urge for getting the highest possible volume is greater than doing the quality of music justice and creating an overall great sounding album.
The Harman EQ curve is just another thing in our modern world which signals another step backwards in terms of audio quality. The people driving this trend are usually self taught engineers who don't understand the proper theory behind making high resolution, decent quality recordings. They basically know enough to be dangerous.
You used to require a decent amount of training and theory to record your own music, so you could use the existing analog equipment at most of its potential. Modern computer based DAWs don't require that much basic knowledge to operate, so there are few rules to getting basic results. That opened the door for any shmuck to butcher their own music and save a buck not needing to hire an engineer. The same amateur person wouldn't be able to run an 80 series Studer 24 track analog machine, so that would weed out the uneducated riff raff from using a hammer fisted recording technique that would otherwise ruin the sound.
Having such advanced conversion technology in a $200 box is astonishing these days. That just begs for instant gratification type use and fosters bad results.
The reasoning a company like Harman is using to come up with this curve of theirs is just silly. No self respecting audio fan would tolerate this sound. Its so removed from any hint of reality. The problem is then how much worse it will sound on a system which doesn't hide anything rather than smear and melt it together in one big blob of noise.
The main reason people feel they can push the level so hard is the lack of gradual tape distortion onset when recording fully digital. The tape saturation is what actually helped the sound or tone of some instruments when applied judiciously. Drums can sound really good when recorded onto analog tape this way.
Digital clipping is however simply awful and has no place in music. The urge for getting the highest possible volume is greater than doing the quality of music justice and creating an overall great sounding album.
The Harman EQ curve is just another thing in our modern world which signals another step backwards in terms of audio quality. The people driving this trend are usually self taught engineers who don't understand the proper theory behind making high resolution, decent quality recordings. They basically know enough to be dangerous.
You used to require a decent amount of training and theory to record your own music, so you could use the existing analog equipment at most of its potential. Modern computer based DAWs don't require that much basic knowledge to operate, so there are few rules to getting basic results. That opened the door for any shmuck to butcher their own music and save a buck not needing to hire an engineer. The same amateur person wouldn't be able to run an 80 series Studer 24 track analog machine, so that would weed out the uneducated riff raff from using a hammer fisted recording technique that would otherwise ruin the sound.
Having such advanced conversion technology in a $200 box is astonishing these days. That just begs for instant gratification type use and fosters bad results.
And now for something completely different. I have a set of Empire grenadiers. They have what appears to be a large dome mid-range. Three questions.
1. is it a dome mid range or a lens / cover?
2. Does it sound so bad because of the aged out caps or does it sound bad inherently?
3. No matter what the driver type, were they a good product when new? (the mid range.)
This is by no means an active project, just curiosity.
Jeremy
1. is it a dome mid range or a lens / cover?
2. Does it sound so bad because of the aged out caps or does it sound bad inherently?
3. No matter what the driver type, were they a good product when new? (the mid range.)
This is by no means an active project, just curiosity.
Jeremy
Quote below from here as it is freely available and has a lot more graph's and explanations that would be useful to anyone not familiar with this research
https://www.researchgate.net/public..._and_Calibration_of_Sound_Reproducing_Systems
"In the double-blind tests, listeners made bass and treble balance adjustments to a loudspeaker that had been equalized to a flat smooth steady-state room curve."
There is another graphic from one of Olive's headphone presentations that might be helpful to see what was changed vs the speakers native in room response without processing
All the listeners did with the tone controls was undo the EQ that had been applied and bump the bass a bit. Trained listeners don't go for quite as much boom and tizz as the untrained listeners.
I tested the same approach when I was equalizing my line arrays. I equalized the speakers to a flat in room target and then used a parametric EQ with shelving filters spaced an octave apart. I adjusted the gain of the shelving filters until the sound was balanced and then made some other slight modifications by ear until it sounded as good as I could get it. I preferred the bass to level out at 100Hz, I have the same dislike for rising sub bass in the Harman headphone target.
What is silly is to try and take the curve and apply it to all speakers with all different directivities and listening levels and expect it to work the same. It won't.
An approximate 1dB/octave fall in room for an ordinary cone and dome speaker with rising directivity at 80dB or below will sound good. Crank it 10 or 20dB and it won't sound so good, make the directivity much more constant and it won't sound so good.
@fluid >>"An approximate 1dB/octave fall in room for an ordinary cone and dome speaker with rising directivity at 80dB or below will sound good. Crank it 10 or 20dB and it won't sound so good, make the directivity much more constant and it won't sound so good,"
This is EXACTLY what I'm talking about in the context of perceived linearity. I can guarantee you once there is a mild 3.5k dip in that curve, a larger group of people would likely choose less low end bump. That little mid dip makes a massive difference in perceived volume and lowers listener fatigue.
The directivity error with most 2 way cone + dome speakers makes it necessary to apply some form of correction to make the speaker sound more neutral. Its not intended to just push the bass up and be happy. The mid dip is obviously volume dependent and confirms loosely to the modified Fletcher Munson (loudness) curve, but also includes a tiny bump around 1k. The issue is the over simplification of the proposed Harman correction curve, not considering the other discussed deviations, which while minor, are audibly substantial.
This is EXACTLY what I'm talking about in the context of perceived linearity. I can guarantee you once there is a mild 3.5k dip in that curve, a larger group of people would likely choose less low end bump. That little mid dip makes a massive difference in perceived volume and lowers listener fatigue.
The directivity error with most 2 way cone + dome speakers makes it necessary to apply some form of correction to make the speaker sound more neutral. Its not intended to just push the bass up and be happy. The mid dip is obviously volume dependent and confirms loosely to the modified Fletcher Munson (loudness) curve, but also includes a tiny bump around 1k. The issue is the over simplification of the proposed Harman correction curve, not considering the other discussed deviations, which while minor, are audibly substantial.
This is the reason why forcing a speaker to a specific in room curve without considering any directivity problems is a bad idea. In my own testing the difference between good and really good was not that much. Just a few little tweaks based on room and speaker issues. I have never read anything from Toole or Olive where they suggest using it as a target. Their point to me seems to be showing that the in room sound from a flat on axis speaker with rising directivity (without directivity problems) is close to the preference of most tested. Amount of bass preferred seems to be the most inconsistent variable in the testing. Program material, room and positioning all play a part in the amount of bass anyone might choose.
This can give systems with subs a bit of an edge. You can easily adjust as required for the best overall balance.
Rob 🙂
Having separate sub to balance and adjust is a great benefit. Once you've had that flexibility, its hard to live without it from then on. I do have to mention that I generally never touch my sub levels once they're dialed in. Some people want to accommodate differences of various recordings and sources by varying the sub level, but I find that more of a nuisance than benefit, unless you're an OCD type that needs to fiddle with everything constantly.
I used to be of the opinion that having separate level controls for mids and tweeters in a 3 way passive crossover speaker was essential. I could see it being very needed on the tweeter but a midrange pad was only useful if the mid converted a specific bandwidth. I abandoned that after having dealt with intermittent problems of flaky, scratchy and noisy lpads. Most of the currently available lpads on the market are of poor consistency and build quality. I now only use rotary switches with set increments, but only if the driver has crossover points which are suited to the purpose ie. HF high pass of 3 - 4k.
The directivity mismatch of most 3 way systems with a large dome mid almost requires a midrange level control, unless the directivity doesn't vary that greatly. Its therefore considered more of a crutch than a solution to the problem.
I used to be of the opinion that having separate level controls for mids and tweeters in a 3 way passive crossover speaker was essential. I could see it being very needed on the tweeter but a midrange pad was only useful if the mid converted a specific bandwidth. I abandoned that after having dealt with intermittent problems of flaky, scratchy and noisy lpads. Most of the currently available lpads on the market are of poor consistency and build quality. I now only use rotary switches with set increments, but only if the driver has crossover points which are suited to the purpose ie. HF high pass of 3 - 4k.
The directivity mismatch of most 3 way systems with a large dome mid almost requires a midrange level control, unless the directivity doesn't vary that greatly. Its therefore considered more of a crutch than a solution to the problem.
The Hypex DSPs ... are the worst thing from Hypex ;-)
The power amps are the next evolution step over IcePower - definitely better. Not sure if one can hear that but technically they are significant better.
"Very transparent without artefacts" is what Hypex delivers broadband, not only at low frequencies like older switched power amps.
Still doesn't mean that's what everybody likes cause it shows everything but it's the perfect start/base line for a development.
@IamJF In most cases I'd agree on the higher end class D being in line with quite a few class AB and even some class A. I did alot of listening and evaluating when I bought my JC5 amp. Not many other class AB or even pure class A can touch the purity of this amp, being its a high bias AB design, operating at up to 25W in class A.
I don't think I'll ever personally use class D for full range operation, at least not after my experiences with a broad spread of other amps I've critically listened to. Every example of class D I've heard so far is behind my flavor of A and AB, but I'm sure there are a few exceptions when considering other parameters, like dampening factor linearity and IMD.
As a whole, it may be true that some higher end class D can approach the pinnacle of current A and certainly AB, at least in the audible frequency band at medium power levels. Even though we can't hear much past 16k, the harmonics are still there in the case of some better analog recordings, if you can find them.
In a professional studio setting, I've been spoiled by having experienced 30 ips analog tape with the fullest practical bandwidth signal available (direct to tape stereo recordings), which blows away anything up to 96k/24bit resolution wise. Almost all commercially available music doesn't have any content past 20k, so it won't fully expose the harmonics of a full bandwidth HD signal comimg from of a superior pair of small diaphragm condensers. This is through class A preamps, going directly onto 30 ips of Studer 80 series tape without a console in between. I've done quite a bit of classical direct to tape recording with full orchestras, solo piano and other acoustic instruments. This material has harmonic content well past 20k, so it will expose anything sub par in the chain. Recording at 96k through higher end converters is close to the quality of analog I'm referring to.
So for anything under 100 hz I haven't heard much of a difference in the examples I've had in front of me between class AB and class D. With typical 20 kHz bandwidth music from digital source, there likely isn't much difference, at least not to my ears. I will still defend the notion that good class A will put any other type of amplification to shame, given that other parameters are equivalent. Class AB is another story, as there are quite a few examples which aren't as good as people claim, at least not with specific master grade recordings.
I don't think I'll ever personally use class D for full range operation, at least not after my experiences with a broad spread of other amps I've critically listened to. Every example of class D I've heard so far is behind my flavor of A and AB, but I'm sure there are a few exceptions when considering other parameters, like dampening factor linearity and IMD.
As a whole, it may be true that some higher end class D can approach the pinnacle of current A and certainly AB, at least in the audible frequency band at medium power levels. Even though we can't hear much past 16k, the harmonics are still there in the case of some better analog recordings, if you can find them.
In a professional studio setting, I've been spoiled by having experienced 30 ips analog tape with the fullest practical bandwidth signal available (direct to tape stereo recordings), which blows away anything up to 96k/24bit resolution wise. Almost all commercially available music doesn't have any content past 20k, so it won't fully expose the harmonics of a full bandwidth HD signal comimg from of a superior pair of small diaphragm condensers. This is through class A preamps, going directly onto 30 ips of Studer 80 series tape without a console in between. I've done quite a bit of classical direct to tape recording with full orchestras, solo piano and other acoustic instruments. This material has harmonic content well past 20k, so it will expose anything sub par in the chain. Recording at 96k through higher end converters is close to the quality of analog I'm referring to.
So for anything under 100 hz I haven't heard much of a difference in the examples I've had in front of me between class AB and class D. With typical 20 kHz bandwidth music from digital source, there likely isn't much difference, at least not to my ears. I will still defend the notion that good class A will put any other type of amplification to shame, given that other parameters are equivalent. Class AB is another story, as there are quite a few examples which aren't as good as people claim, at least not with specific master grade recordings.
Was it here where someone wanted to see the inside of the MD60?
Interesting, that MD60N. It looks a little different than the ones I had. They clearly tried to get right in behind the dome with the dampening material. That's the key to making a bid dome sound good in the lower mids.
The cavity resonances you can get with a large airspace can make or break performance. The conventional dampening setup with the open center pole cavity is less than optimum, unfortunately typical for cheaper designs like the budget hivi 2" soft dome.
The secondary chamber formed by the airspace behind the VC gap is the other area which can cause problems around 2k on most 2" domes and 1.5k on most 3'. You'll observe a second impedance peak at that point and a noticeable CSD trail. In most cases the main symptom is just a small spike in 3rd order HD, often not being that audible. The only way to tame this issue is with a cross drilled pole, nulling the resonance. You can't really get any dampening material in that gap, so its often an accepted drawback to live with this second peak.
The motor design on that MD60N is typical of most larger neo domes.
The cavity resonances you can get with a large airspace can make or break performance. The conventional dampening setup with the open center pole cavity is less than optimum, unfortunately typical for cheaper designs like the budget hivi 2" soft dome.
The secondary chamber formed by the airspace behind the VC gap is the other area which can cause problems around 2k on most 2" domes and 1.5k on most 3'. You'll observe a second impedance peak at that point and a noticeable CSD trail. In most cases the main symptom is just a small spike in 3rd order HD, often not being that audible. The only way to tame this issue is with a cross drilled pole, nulling the resonance. You can't really get any dampening material in that gap, so its often an accepted drawback to live with this second peak.
The motor design on that MD60N is typical of most larger neo domes.
@IamJF How old is that MD60N of yours? I suspect its a first gen design being you could open it up. I couldn't get the chamber to come off of the ones I had. Does the VC former have any vent holes in it? Curious if they changed the design from the first production run. Please take a pic of the VC former and windings if you can.
That's why you should try to listen to a really good Class D 😎
These are especially in IMD/THD and noise way better es every "normal" analog design. NC500 has an output impedance of 1,5mR at 16kHz and is linear below.
When you look through the Hypex brochure you see the Difference - it starts with the older generations and NCx500 is the newest one.
https://www.hypex.nl/oem-brochure/Hypex_OEM-brochure.pdf
This is the level of what you heared (UcD400, that's about the level of Pascal and B&O)
Output impedance 11mR at 1kHz and 35mR at 20kHz. That's all together not bad and sounds good!
This is what the newest Generation does:
There is a 0 more in the THD scale ... about 130dB S/N, easy to use plug and play.