Official Rythmik Audio Subwoofer thread - Page 216

I make subs, so my job is easier. All I do is make sure the woofer output is linear to the input signal. And I leave the rest to the best point tool. But even that, I am not so adamant about which "flat" response should be used as reference and that is why I put in multiple extension and damping factor settings. Therefore I expressed my personal opinion that flat frequency response is the trailing indicator, not leading indicator, just like the subtle difference between correlation and causation that a statistical study cannot differentiate. But for engineers and scientist, the correlation and causation are two different things.

Sensitivity is 0.3VRMS. If you can use RCA, using Y cable to plug in both channels increases the sensitivity to 0.15VRMS.

Sanjay, that is why I made comment that the answer is our subs make statistically more coherent sound(or lack of a better word because some also use coherent to mean linear phase). "Coherent" here means the both the phase and frequency response is consistent over time, not changing time over time. Our brain analyzes sound more than just by frequency response. If you blind-fold a person, and lead him to a room, he can do a gross guess of the dimension of the room and the type of floor (such as carpet vs tiles vs hardwood) if he is properly trained. That is right, it is a learning process. Our brain keeps comparing what we hear now to what we have heard before. It is like searching through a huge database and some do it better than others. The word "accurate" is used because it triggers our brain to return a "hit" in the database more frequently. If the sound cannot maintain correct relative amplitude and phase with respect between the original sound and echo, our brain will have hard time doing estimate. The same thing applies to other quality of sound. Some recognized that we never heard bloated sound in real life, so they are particularly turn off by boomy sound. It will not click them as real sound and time domain response is related to both phase response and frequency response. So why the frequency/phase response would change over time? It is because: 1) the voice coil changes over time and the frequency response change with it, and 2) the components used in speakers possess memory effect with some worse than others. And servo addresses both issues. For the first point, it was demonstrated in objective frequency sweep and all nonservo subs will have a progressively large dip at a frequency point between 60hz and 100hz as we move the SPL higher. And worse of all, it is a memory effect meaning that change in frequenc/phase response will stay for a short while until the temperature drops back. And as for point 2), we all know spider and surround has hysteresis. Do a google search on the subject. There has been effort in the past to get rid of spiders altogether. But it wasn't successful. Our servo also addresses that. So in short, we completely solve the memory effect introduced by the changing voice coil resistance, and reduce the memory effect from speaker components by a factor of 3x or more. That is the type of sound our customers hear. You can also conclude that the same mechanism offering these improvements also ensures we have a much better controlled unit-to-unit variation in terms of sound quality and frequency/phase response. That is another plus for quality control. Customers notice it, but they don't know the reason. The reason is the servo mechanism reduces the variation arised from T/S parameter value change. This reduction does not depend on if it is a short term effect (as in distortion), or permanent effect (as in unit to unit variation). Last point, our subs are much easy sells to veteran audiophiles than younger listeners. That is a harsh reality and I am well aware of that. What is why I replied to HD330 about what to listen to. It is a learning process. But he misunderstood the intention.

Kyle, #8 for drivers and #6 for amps.

In terms of sensitivity, we design the gear to be more sensitive than 0.9Vrms. All designs need some margin. So it takes only 0.3VRMS input to drive the sub to full power.

They are in the pipeline. We need to get the A370/H600XLR2/3 out in June and HX800 will follow after that.

My explanation is that the clear and coherent reproduction of the original source and its echos is the reason. The microphones at the recording site will pick up the direct sound from the instrument (let us call that source) and its echos from near-by objects. If that object where the echo is reflected from is 4 ft away from the instrument plus if the echo signal is an inverse (meaning it has negative magnitude), the lowest frequency being affected has a half wavelength (180 degrees shift) of 4 ft. So the frequency is 1000/(2*4)=120hz. The effect is best demonstrated as an impulse response. In this case, we will see the source signal at time 0 (which looks like a normal impulse waveform), and then at 4/1000 sec (or 4ms) later, there is another blip with negative shape (which is similar in shape to the source, except it is going to the negative amplitude direction) and with smaller amplitude. When you do FFT (fast Fourier transformation), we get the frequency response of the composite signals. So you can see for frequency response, both source and echos contributes to it. But our brain does not treat them that way. In addition, if the echo is further delayed (because it is from farther objects), the affected frequency goes lower.

In short what the sub plays is the low frequency source plus its multiple echos. If the sub is "coherent" over time, our brain can still recognize the music rhythm of the source out of complex composite sound. At the same time, this complex composite sound makes it sound like the instruments is boxed-in in a vaguely large box (the echos define the box). In a less coherent sub, the bad distortion (arisen from memory effect) makes the source and echos hard to be individually recognized and therefore just seems to be some sort of rumble in the background. But in spectral analysis, the two have very similar. Crucial information is lost in analysis.

As we switch the extension filter setting on our plate amps, on good recording, the sound effect is the sound stage seems to become larger, deeper, and taller when we switch from 28hz to 14hz. In addition, as we do that, the sound lost its one note characteristic where some describes it as punchy. If some prefer to hear that type of sound, he should try 28hz/low damping and then boost signal around 80hz to 150hz.

But in the same track, we can also hear the gusty sea wind noise and the detailed sound of the waves hitting on the hulk of ship as the battleship charged forward and broke the waves. And then there was this splash sound when the barrels were dropped into the sea and the sound that one woud have heard from under-water from a dropped object. All of these are something we can experience in life. We can make judgement how detail and truthful the sound is. Yes, the sound of the depth charge itself is unknown to most of us, but almost everything else is what we can experience in real life. So percentage wise, we can conclude we heard more accurate sound, isn't it? U571 remains to be one of my favorate because the sound effects are very natural.

The low frequency rumble (say below 20hz) is most likely air-borne and the only way get around that is turn on rumble filter. But frequency band above 20hz, up to 40hz is more like the resonance of your house structure. You want to do a sweep when your neighbors are away and find out what frequency causes your walls to resonate the most and have the PEQ to trim down that frequency specifically.

We have very limited supply and we release them to those who need the driver immediately. If you got to have it by a deadline, let me know.