Official Rythmik Audio Subwoofer thread - Page 154

50hz highpass? Not a problem. Our preamp board is very easy to fiddle with. We can discuss that off-line aboutwhere to solder what to get that. This is strictly for experimenting purpose. In production, I don't want to have 10 different configurations. That will wear me out


We can help you out to trade them in for D1510 in the end. We are flexible.

In the real world, the flat frequency response may not yield the best sound, in particular for front speakers. If a flat FR is all we need, the speaker designers will be replaced by computer software. A lot of times, the flat FR is the first shot solution (or so-called initial solution), and then the designer comes to fine-tune it using his ears (a process called iterative improving). Maybe the tweeter is too bright, so we can attenuated more on tweeter. Or the midrange is too forward or too layback,... This is just one aspect. Eventually he may find the Q value (or impulse response), phase response, voice coil weight, rubber surround material (creep effect), cone material, and other factors not listed here all affect sound. This is just the complexity of the problem. Then the components in crossover design. Some components sound brighter than others.. That is the second layer of fine tuning. If one has unlimited budget, component cost is not a problem. However, on commercial speakers, finding good sounding components with reasonable cost is trickier.

If you think about it, this industry is a bit disconnect from various related fields. The transducer designers do not understand how electrical circuits and components affect sound and think transducers are the only one matters. The electrical engineers think flat frequency response is the only one matters. Not too mention we have software guys have their opinion.

Even for transducer designers, we have one with Ph.D background (you know who) think his waveguide design is best and for bass, all you have to do is add a bunch of subs and mix them. Sure, subwoofers are just commodity. Why doesn't this "many subs and commodity idea" apply to his waveguide speakers? All engineers look for unified theories so that they can apply the idea to a wider scope. If I believe servo subs sound better, I will naturally think servo midrange drivers should also sound better. Bass is just other frequency band: common sense applies.

In short, to answer why flat F/R bass may not be the best? I don't think we fully understand how we perceive sound quality. Just like someone has an illness with many different symptoms, including headache. If the doctor cannot find the root cause or the relation between various symptoms, all he can do is prescribe pain reliever and treat the most obvious. Well flat FR is our easy target here. So if flat FR gives us good sound, that is great. But if it doesn't, I won't be surprised such cases do exist. But that is not to say having flat FR has no value. It is a good initial solution. A tweek here and there maybe needed.

For boominess, it can be the echo time and Q value of the extension. I would calibrate XT32 with 14hz low damping and afterwards, change the extension filter setting to high damping 14hz or 20hz. If you'd like the sound to "firm up", you can also play with the PEQ setting on our plate amp. Setting it to +3db at 80hz with mid to max bandwidth to bring to midbass forward to you. That can change the perception of muddy sound. Have you tried set your front speakers to large and play in the same room? It will give you the idea of the bass tone quality from the front speakers.

[EDIT] BTW, what does XT32 say the distance of the subwoofer is? That information is important. If the distance information is wrong, then you need to change the crossover to another frequency that XT32 is correct on subwoofer distance.

The best sound for music is 14hz/hi damping. High damping has less ringing and it has a much more gradual roll-off.

The setting for auto calibration is just temporary. Let us assume the software is capable of achieving same consistent calibrated response if the pre-calibration curves are well behaved. BTW, switch damping control does not introduce new peak or null, so it is reasonable to assume Audyssey will come back with very similar calibrated curves. If one set the extension to 28hz, the software may even boost the band between 20hz and 30hz. So the idea to present the low damping curve to calibration program. And then check the sound. If you think it is too boomy, then you can switch the damping back to high. If one had used high damping for calibration, then there is no "ultra" high damping setting to fall back to. If the sound is good, there is no need to make further damping adjustment.

I encourage everyone to adapt the idea. Alternatively, one can also use 14hz/med damping for sealed subs and afterwards, one can use low damping to make the bass even more "full body" (or boomy)by switching back low damping, or more articulate by switching back to high damping. Of course, there is nothing wrong using 14hz/high damping for calibration if the result is right on what you expect. My rationale here is the software should also give you very similar calibrated curve if the damping were set to med damping or low damping. For vented subs, my recommendation is to use 14hz/low damping as the vented subs already sound more full body than sealed subs.


MJ.com,

When I say fine tuning, it is not meant to do huge adjustment. You may do a +/-3db to adjust because bass is less sensitive. In the example case I mentioned for front speaker design, a +/-1db adjustment already sufficient to make audible difference.

The reason I mentioned distance is important is one can use a common sense of physical distance plus and minus 1/2 wavelength as a sanity check window. If the software comes back with difference distance, it means the software was not able to calculate the distance correctly because the input data have huge anomalies that is either from strong reflection or other source. That is the first indication there is some trouble in the setup. We should do something to "help" the software. One thing we should know is sound travel faster then air. If one place the mic too close to the wall, hypothetically it can cause problem. Note that 1/2 wavelength means 180 degrees in phase difference. It already has sufficient margin to account for all errors. In reality, the actual value should be well within that range. For instance, wavelength of 80hz (the crossover point) is 12ft. Half wavelength is 6ft. So your sub distance should be within physical distance +/- 6ft. When something like this happen, try a different set of inputs and placement. If you got to have that placement location, you may need to do manual calibration instead.