Bass Management System Development

Last week I posted my responses to a list of requirements for a dedicated home theater bass management system. My company has some interest in developing this sort of system.

There was some interest, so it seems like the next step is to start a thread to discuss it in more depth. Here we go.

These are the requirements I listed last week:

~ Two independent channels.
Fine. Two independent inputs. How many independent outputs (in your wildest dreams)
~ A minimum of twelve parametric filters per channel (20 per channel is good too).
Fine. Only practicale limitation is processing artifacts and delays of multiple filters. Modern DSP's can handle this easily.
~ Channel coupling and series looping.
Probably not. All digital filtering can be accomplished in one pass at higher quality than two passes.
~ Three-way IN/OUT bypass.
Probably no analog bypass because we don't have analog outputs from the controller. Digital filters can certainly be cut in and out.
~ Input/output LED level indicators.
Definitely indicators on PC interface, possibly on front panel.
~ LED filter indicators for both channels.
Probably not.
~ Front panel display.
Very limited
~ Retain normal required buttons and jog dial.
Probably not, possible IR remote capability
~ 24-bit/96 kHz A/D and D/A converter powered by a 32-bit DSP.
Surely possible, although there needs to be a discussion on the recommended sampling rate for <250Hz signals.
~ Internal switch-mode power supply (100 - 240 V~ / 50-60hz).
Probably wall wart power supply with internal regulation.
~ Low power consumption.
Sure.
~ Adjustable input level (-10dBV / +4dBu) on rear.
Consumer line-level unbalanced input. Is more really required for a home theater box?
~ Hard bypass relay.
No.
~ Noise-free.
No fan.
~ 107db dynamic range.~ 0.007% THD.
On inputs and outputs.
~ High-quality components.~ High-quality construction.
Given
~ Balanced XLR inputs/outputs.
Required for home theater?
~ Soft power ON/OFF with selectable Auto-On detect via sub signal.
Probably
~ Eliminate turn-on thump and ground hum.
Sure, the digital interface will eliminate ground hum potential. The DSP can be responsible for pops.
~ Double insulated chassis with a two prong plug or a ground lift switch if needed to eliminate hum.
The external power supply will provide isolated 2 terminal power
~ Pure flat frequency response from 10hz to 20khz* (FBQ2496 is -3db @ 20hz).
Yes
~ Filter adjustment capabilities from 10hz to 20khz*.
Yes to below 20Hz (probably not to 20kHz as the system concept is for sub woofers).
~ Adjustable gain range from -24db to +16db for each filter.
Yes
~ Minimum of 4 to 6 memory presets.
Yes
~ Replace ¼ inch inputs/outputs with RCA inputs/outputs.
Yes
~ Simple shelf filter… (i.e. selectable linear boost between two selectable frequencies).
Yes
~ Selectable and/or variable 6db-48db/octave subsonic roll-off filter from 35hz to 10hz (minimum 5hz increments).
Yes
~ Time delay from 0 to 30msec or 1 foot increments w/ 0.1 foot fine up to 30 feet.
Yes, independent for each output, is 30 feet enough?
~ Variable phase correction/adjustment from 0-180.
Yes
~ Front panel USB and/or RS-232 interface for MIDI port (place under small flip cover like on computers).
1394
~ Dimmable front panel LED's / lights with OFF option.
Off option for free.
~ Soft blue and/or green LED's on front panel vs. red.
few LEDs
~ Offer unit in black or silver… if not optional, offer only in black.
black
~ Revamp chassis for home theater consumer appearance vs. pro-audio style.
~ Allow mounting brackets to be optional (include unattached).
~ Enclose/fill gaps on sides if mounting brackets are removed.
~ Add rubber feet to allow placement on top of other equipment.
~ Remove graphic design from top of unit.
home theater friendly
~ One year warranty.
certainly
~ Anti-clipping mechanism with higher dynamic peak input before clipping.
24 bit dynamic range may have to do. Clipping indication available.
~ Frequency readout in hertz instead of base + fine… (1hz increments w/ 0.1hz fine increments).
Yes, on PC application.
~ Bandwidth readout in octaves (1/3,1/6,1/24,1/1, etc.) or hertz (2hz,5hz,12hz,etc.) instead of 1/60, 10/60, 60/60, etc.
Yes, on PC application, ideally notation will agree with REW.
~ Front panel volume/gain to control output level to sub.
Probably not, possibly available over IR control.
~ Ability to slave multiple units together.
Unlikely, hopefully it wouldn't be necessary. Can you describe the need?
~ Linkwitz transform circuit.
Can't this be done with a digital filter? Looks like a shelf filter.
~ Detachable power cord.
Yes, wall wart.
~ Individual crossover filters (Butterworth, Bessel, Linkwitz-Riley) per channel w/ selectable roll-off 6db-48db/octave.
Digital crossover filters?
~ Adjustable/variable crossover range (low pass and high pass from 20hz to 200hz and w/ bypass).
Yes, if digital filters are acceptable
~ Built-in RTA mic/line input w/ phantom power and w/ GAIN control.
Yes, this would be XLR?
~ Built-in SPL meter with selectable dBA/dBC/OFF weighting.
Digital
~ Built-in test tones (sine waves - individual 1hz increments and sweep from 5hz to 400hz or 5hz to 20khz).
Yes
~ Separate RCA mic monitoring and test tone output to allow for computer program monitoring.
Don't quite understand, certainly computer program monitoring is a priority.
~ Locate RTA mic/line inputs and RCA monitoring outputs on the front face plate under a small flip cover.
possibly
~ Video output for monitoring the display would be nice, but the cost factor may prohibit this.
Display on computer connected via 1394.
~ Video display instead of LED's… (dimmable and selectable OFF).
no, but graphically rich PC interface available
~ Small and simple 6 button wireless IR remote with discrete ON/OFF buttons and Preset 1,2,3,4 buttons.
IR control could replace many front panel buttons, and interface with home theater control systems.
~ Rear IR jack for remote control… to use with repeaters.
Yes
~ Add volume/gain control button to remote if feature is made available.
More likely on remote than on front panel
~ Add other control buttons to remote if cost feasible.
Given IR control, more buttons is not a cost issue. We probably wouldn't include a remote, more likely to offer a configurations for Harmony and other programmable remotes. I know my theater doesn't need another remote!


I can't commit to this product yet, but feedback would definatly help define a possible product and validate the potential market, both useful in our business case development.

Barry

I was unaware of using professional amplifiers for DIY subwoofers. That makes sense. I would expect to provide a different output module with XLR connections if there was a demand. Keeping the unbalanced output module inexpensive is a priority.

Is there a similar situation for source signals on XLR? I assumed most dolby decoder type boxes would be unbalanced.

Barry

In the interest of full disclosure I work with Quantum Parametrics.
To date we have exclusivly produced 1394 test equipment. As I mentioned previously, we are introducing a new 1394 protocol standard and are looking for demonstration product applications.

Initially the bass management system seems a good fit both in technology, complexity, and market size.

It's much too soon to say how or if or who might produce this product. I'm just saying I'm interested in discussing it.

[Sonnie]

Hopefully we can get some worthy discussion for you... I know we all have the desire to see a unit produced that will fit our needs better than what is being offered at this time.

[brucek]

I would expect to provide a different output module with XLR connections if there was a demand. Keeping the unbalanced output module inexpensive is a priority.

Is there a similar situation for source signals on XLR? I assumed most dolby decoder type boxes would be unbalanced.

I don't know what you mean by dolby decoder type boxes.

The unit has to be able to accept both unbalanced and balanced inputs to the two channels. The input to the device will be fed line level signals from home theater processors, receivers, preamps that have low output impedance (usually less than 100 ohms). The device has to have high input impedance that accepts a switchable range of levels since there are two quasi standards in the industry (-10dBV and +4dBu).

The balanced connections can be TRS or XLR and the unbalanced RCA type.

The output of the device (low output impedance) will be feeding subwoofers with their own amplifiers, or DIY subwoofers with external amplifiers (high input impedance devices). Some of these devices accept balanced and some unbalanced. Again the levels usually conform to one of the two standards.

Here's an FBQ2496 from Behringer.
You can download its PDF and look at the specs etc.

Here's the type of specs we would be looking for. (except most people would want the unbalanced connection as an RCA rather than an auto detect TS jack..



brucek

Thanks for the input. I'm warming to your position, but I'm not quite there yet .

The cost goal is going to be difficult to meet. XLR connectors alone can cost $5 each. My estimate is 95% of the customers for this device wouldn't use any XLR inputs (feel free to make a better estimate). That means the system would cost at least an extra $25 for something that wouldn't add any value to 95% of the customers. On the other hand, an XLR for a calibration mic could possibly add value to 95% of the customers.

On the bright side is the flexibility of the Firewire architecture of this proposed system. I consider it the true professional audio interface. If a high end customer saw value in implementing a balanced input to the system he could attach another device like a M-Audio Solo or Miglia HA02 ($120 or less). These devices could digitize the balanced signals in a very high quality manner and pass them to the bass management device. I think the Miglia even does the soft clipping you were interested in. If this really turned out to be a popular option for the system, we could develop simple lower cost devices to perform this same function. The Miglia and M-Audio devices include many other functions.

A customer could presumably gang as many of these devices as necessary to satisfy all his input requirements. The memory in the main system would select the appropriate input for each setup. In my future dream world, all the source devices are providing digital streams over Firewire and no converters are necessary until the amplifier.

All this flexibility comes nearly free with this system architecture. The main unit stays as low cost as possible with 2 unbalanced inputs that add value for 99% of the customers. Even most of the XLR users will likely have an unbalanced device they want to hook up.

I'm still open to the probability I'm missing something. Please feel free to correct any of my assumptions. I appreciate you sharing your experience and I'm willing to adjust to compelling arguments.

Is this really the only issue you find on the list of requirements? I can't imagine it's so close to ideal on the first pass! Would you like to discuss DSP options?

Barry

[Sonnie]

Personally I would not be offended it it did not have XLR inputs/outputs. There is quite a lot of fine equipment out there that doesn't have XLR and generally they will all have RCA... plus we are talking sub frequencies, so I don't see it being a make or break issue myself, but I could be missing something.

Thus far let's consider the features on the list you object or somewhat object to:

~ Channel coupling and series looping.
Probably not. All digital filtering can be accomplished in one pass at higher quality than two passes.

~ Three-way IN/OUT bypass.
Probably no analog bypass because we don't have analog outputs from the controller. Digital filters can certainly be cut in and out.

~ LED filter indicators for both channels.
Probably not.

~ Retain normal required buttons and jog dial.
Probably not, possible IR remote capability

~ Front panel display.
Very limited

~ Hard bypass relay.
No.

~ Front panel volume/gain to control output level to sub.
Probably not, possibly available over IR control.

~ Linkwitz transform circuit.
Can't this be done with a digital filter? Looks like a shelf filter.

~ Separate RCA mic monitoring and test tone output to allow for computer program monitoring.
Don't quite understand, certainly computer program monitoring is a priority.


These are really all we need to discuss.

[Sonnie]

~ Channel coupling and series looping.
Probably not. All digital filtering can be accomplished in one pass at higher quality than two passes.

This probably isn't a big deal. I don't really understand the digital limitations of looping one channel to the next for double the filters, but with a dozen filters per channel, I think most everyone would be satisfied.

If I'm remembering correctly, coupling simply allows both sets of channel filters to be input simultaneously. With REW automatically loading the filters into either channel, I don't see this as a major loss.


~ Three-way IN/OUT bypass.
Probably no analog bypass because we don't have analog outputs from the controller. Digital filters can certainly be cut in and out.

~ Hard bypass relay.
No.

I think these go together. I can't remember what the issue was with needing a hard bypass relay.

How would we set the input levels without bypass... again, I'm sure I down understand the process completely, but for the average guy like me, I'm seeing this is something we need to do (set input levels).

So when the unit is OUT, does the signal still get passed on to the sub?



~ LED filter indicators for both channels.
Probably not.

~ Front panel display.
Very limited

These tie in together. I assume by some of your other remarks that there will be some sort of monitor output and/or computer connection for monitoring/viewing the settings. So having a computer nearby is now a necessity and the manual method of doing things is out the door.




~ Retain normal required buttons and jog dial.
Probably not, possible IR remote capability

~ Front panel volume/gain to control output level to sub.
Probably not, possibly available over IR control.

We will either have to have these on the front panel or on the remote. Somehow we have to be able to make the adjustments, input filters, etc. OR, are you suggesting this will all be done via computer program that you will furnish?

I'm confused about these.



~ Linkwitz transform circuit.
Can't this be done with a digital filter? Looks like a shelf filter.

Not a major loss IMO. This was suggested because it is supposedly a relatively inexpensive add-on, but I don't understand how it all works... others can chime in on it.




~ Separate RCA mic monitoring and test tone output to allow for computer program monitoring.
Don't quite understand, certainly computer program monitoring is a priority.


If we have a mic input and SPL meter built in to the unit... that information has to get to a source that can read and display the information it measures. I guess maybe it's a given that this would be included.

Thanks for breaking it down.

I'm still no fan of series looping. If channel coupling means running the same filters on both channels, that seems very easy to accomplish.

The two bypass issues do go together. A wired bypass breaks the architecture.

Input level setup is a good point. I would expect this to be the first step of calibration. I envision a process step where a signal is applied to the system and the system reports any signs of clipping. Then the input must be lowered or raised until maximum dynamic range is achieved without clipping. I'll bet someone can find holes in this strategy. Hopefully you can also recommend a fix!

I don't see why one of the presets couldn't be OUT. In this mode no signal would be sent to the sub. Does this serve the OUT request? I thought OUT meant no change to the signal (even though it's being A/D'd and D/A'd.) Certainly another preset could implement this "no processing" option.

Let me expand on the computer interface. My concept is that all setup would be controlled from a computer connected via Firewire. It seems to me that calibration with REW already requires a computer. Additionally requiring a Firewire interface wouldn't be a show stopper if it adds the value I expect it to. In the simplest case the computer application could just be knobs and buttons like a physical interface (the manual method), but I imagine there is great opportunity to automate tasks and integrate with other applications like REW. The goal is to reduce user interface cost on the hardware, and enable user interface development on PC's where it is easier, less expensive, and more flexible. I'm sure attracted to the Java multi-platform approach of REW.

So yes, setup would be done on a computer running a provided application.

After setup, the computer would not be required for operation. Hopefully there will be a small list of required actions during operation. Currently I can only think of one "change preset". I'll bet we find at least a few others that will be needed during operation. These actions must be supported by front panel buttons or IR control. I also need a list of feedback (LEDs) required during operation. That list needs to be as short a possible to keep costs down and focused on operation, not setup.

SPL information from the microphone would also be sent to the computer for display and use in the setup application.


Does all that make the system more attractive or less?

[brucek]

My concept is that all setup would be controlled from a computer connected via Firewire.

A lot of people don't have Firewire interfaces on their PC's that they use for REW. I would much rather see USB for the PC connection.

I don't see why one of the presets couldn't be OUT

You're a bit confused on the meaning.
The BFD has a button on the front panel that is labelled IN/OUT.
The LED on this switch indicates its present functionality.
LED on = Filters engaged, VU meters showing output signal levels.
LED off = Filters disengaged, VU meters showing output signal levels.
LED flashing = Filters disengaged, VU meters showing input signal levels..
We use the switch to test filters in or out and also to set the input level.
We do this by playing a bass heavy movie at the loudest volume we would ever have the sound system, and then we adjust the receiver/processors subwoofer trim up or down to have the signal level just showing a yellow LED and the RED LED flashing. We hope for the best S/N and dynamic range once that's set.
Yes, it's a crude way to set the input level, but you have to understand that this isn't a fixed line level as would come from a CD player. This level is at the mercy of the receivers volume control. We need to have monitoring on input and output for clipping.

My estimate is 95% of the customers for this device wouldn't use any XLR inputs

Anyone with a DIY IB or subwoofer would likely have a pro amp. These use XLR connectors, balanced inputs. Some allow the use of unbalanced inputs by using a TS plug, but adding the likelihood of hum. Subwoofer are very prone to hum since they amplify that frequency range so well.

In addition there are those that have processors with XLR balanced outputs. There's a certain pride and admitted insanity about audiophiles that you may not be aware of. Most wouldn't be interested in using a "converter" to switch to unbalanced..... Just my opinion though.

brucek