Official Rythmik Audio Subwoofer thread - Page 270

That is not quite true. A lot of techno sound goes below that. But there is more to it. Reverberation from the surrounding changes the frequency bandwdith. For instance, let us say we have an impulse response at time t=0. If there is an echo at time 50ms and this reflection occurs with the same phase as the orginal impulse, there will be an emphasis at 20hz on FFT (Fast Fourier Transform) result. If, in addition to that, there is also an echo at 25ms, but with inverse phase of the original impulse, that will further enhance the frequency component at 20hz (as the impulse has made it more like a waveform resembling a sine wave). We all know the sound of the announcer in a large heavy reverberate stadium sound like it has a lot of bass. Now in the movie, how do you create that type of effect? We first record the announcer voice in the studio and put it through DSP with an impulse response that injects many of these tiny time delayed echoes and out comes the special effect. In the early days of room EQ, it was a tough sell because engineers quickly realize this type of reverberation (similar to what happen in room) is not same as just manipulating the frequency domain response.

Now most of the musically recording is still done naturally. For instance, in the 1980's, labels like Proprius started recording in huge churches. The sound effect is just amazing. When you have a sub with good extension, it makes you feel you are actually in there. On the other hand, once it is played in the elevator with limited low end extension, all the sense of being in the church is gone. So one can argue it is easier to demonstrate the sound quality with music that is naturally recorded than with HT riddled a lot of DSP processing. Personally I feel there is a push from movie studios to dismiss sound quality difference ever exists in good amps vs bad amps so that they can continue to feed us artificial sounds.

BTW, I started working on these servo subwoofers when I started really enjoy those recordings. "Jazz at the Pawnshop" is just one example. I found the "hall effect" is greatly enhanced when I set the extension filter to 14hz.

-
Edited by Rythmik - 1/26/13 at 8:51am

As for LFE vs line in, they are almost identical except the phase alignment will be a bit different. For instance, if one swaps same signal from LFE in to LINE in, he needs to add about 2 ft distance to the subwoofer distance to compensate for the phase difference. Since these two sources can all be used up to 100hz, I have implemented such that they have a bit of different sonic characteristic. I encourage everyone try both (with the correct distance compensation of course).

BTW, I have seen some companies boasting their subwoofer bandwidth are up to blah blah blah. Little do they know is that is just a way to invite noise, either from the relay or signal switch for the source AVR, or from the noise contracted by the interconnect cable. There is a lot of reporting of solving noise issue by switching from one cable to another. Solving that problem is just a "relative" measure because the problem may still be there, just with less degree.

-
Edited by Rythmik - 1/26/13 at 8:51am

This type of discussion of music vs HT is not brand new. It happened in audio industry numerous times in the past. Early Japanese marketing gurus studied the sound preference of US customers. They come to the conclusion there were east coast sound and west coast sound. Elsewhere we have US sound vs British sound. We also have speakers that is good for classic and bad for rocks or vice versa. I think it is human nature to think that way. But the truth.... I dont care (or we should not get hung up on it) We should not let this type of arguement sway us one way or another because everyone is different.

That being said, a lot of times, the recommendation of which sub works best for customer is based on room size, existing subwoofer (ported vs sealed), and how they like the sound from current sub and that helps us to direct the customers to the correct model.You will be surprised how many times I have heard "well it is ok for movies, but for music it is muddy or boomy". So indeed there are differences these customers can hear. Those are the type of customers we want to get. If the customers tell me they all sound the same, just not power enough, I have a different strategy.

-
Edited by Rythmik - 1/26/13 at 8:50am

There are indeed some ambiguity in terms of how LFE is handled in non-native 5.1 setup (or x.x). LFE is not a complete independent channel. Think of a gun shot. It has low frequency content that naturally goes to LFE channel, but it also have high frequency components going to other channel. They should be really combined together in signal domain (where speaker phase response and distance are completely irrevalent) before they are processed further for bass management. But there are docuements showing it is not done this way, which completely beats me. How what do we have .1 channel? It is encoding efficiency. Instead of using standard 44k sampling rate, .1 is sampled with less than 1k sampling rate in order to save the storage space.

The model of the amp is HX800. In addition to HX800, we will have HX300, HX600 (dual 300W x2). They are supposed to be released last year. I have a customer pre-order a pair of them and my priority is air ship units to fulfill his order first. The rest should arrive late March.

The current A370 plate amp can driver up to 4 16FR drivers on paper (because the DCR will be 3ohms just like our sealed or ported subs). But there are reports the amp gets a bit warm (because the impedance peak from OB driver is not as strong as in sealed or ported subs) and hence Danny recommended 2 or 3. HX800 class-D (it is no secret that they are based on Hypex modules) and hence more energy efficient. I do feel that each 400W should comfortably drive even 4 16FR and that gives you 8 16FR total. I have proposed more than once that we should re-parameterized 16FR to SW12-8FR and have an amp specifically designed to use two OB drivers per channel. The customer should expect to save $120. But he is not sure if that is a good idea or not. You can help me to persuade him. Here is the thing. T/S parameters are scalable with respect to DCR. But all current amps have a sweet spot at 4ohms (or 3ohms for the good ones). I really think we should standardize the OB drivers as 8ohms load (or 6ohm DCR) so that each 250W to 300W module can drive two. If we need more drivers, switch amp from HX300 to HX600.

The servo feedback circuits are truly isolated. But we use the trick of pseudo complementary (or BTL) so that there are always equal (or similar) amount of current drawn from the positive and negative power supply caps. If you mix drivers, that condition of pseudo BTL is no longer true. All of advantage is lost. In short, we don’t recommend mixing different drivers on our dual mono amps.

-
Edited by Rythmik - 1/26/13 at 9:49am