My crazy Processor idea: Below is an idea I had that may not be 100% practical but its fun to dream sometimes.
My idea is to build a pure digital preamp/processor that has no preamp stage. It would have 1-7 HDMI inputs to connect sources plus some coax/optical digital inputs as well. These feed into a set of the top of the line 64bit DSP’s for processing. On the output we have a Monitor out HDMI to go to the screen. Also there are about 4 additional HDMI outputs which will only pass HDMI digital LPCM 6-8 channel audio which will feed the sound out to the amplifiers. Also there may be a set of 2-4 RS-232 serial ports which will be used to talk to the amplifiers to control them. Also has a network port which can be used to control multiple amplifiers that have a telnet over Ethernet control option. There may be some coax/optical digital outputs for legacy amplifiers that lack HDMI inputs. Finally there is a calibration microphone for speaker setup and room correction.
We have done away with all the analog parts of a normal pre/pro so how is it going to work? It takes in a source over HDMI like a DTS-MA track for example. It processes it through any digital Filters, EQ, DTS:X etc to produce lets say 10 speaker output signals depending on the current speaker setup. In a normal design these 10 outputs would then go through a DAC to the balanced/unbalanced pre-outs. Instead each speaker channel is assigned to one of the 6-8 channels in one of the 4 HDMI outputs and sent to the amplifiers internal DAC and amplification stage which then outputs to the correct speaker. Taking away all the analog requirements make power supply and other electrical design problems much simpler.
What Amps would it support? Most HDMI capable AVR’s are theoretically usable but there are a few requirements: They need good DAC’s, Good Amplification stages, Good ability to bypass all processing and just pass the digital input to the DAC’s/AMP’s without interference, digital volume control instead of analog POT and ideally the ability to be controlled over RS-232 or network. Most mid-high end AVR’s made in the last 5 years have these boxes ticked. Users may want to support existing traditional 2-7 channel power amps and for this an options external device could be created that has and HDMI input and 7.1 balanced/unbalanced analog outputs.
How the AVR’s controlled? Normally a pre-amp applies master output volume to all analog pre-outs before the amp stage but in this setup the pre-amp is inside the AVR so we would use the volume control of the AVR to set the final output volume. Any individual channel dB gain/loss set in calibration would be applied by the processor to the PCM digital signal before it is sent but any output volume adjustments are done by sending the required control codes over serial/network to the AVR’s. Also the commands to set the correct input and output mode for each unit is also sent at the beginning. Any AVR’s that can’t be controlled like this may be usable if the volume is just fixed manually and then the output level is adjusted by reducing the amplitude of the PCM digital signal being sent instead.
How would you get the levels/timing right on such a setup? During the microphone based setup and calibration stage the processor would have to measure the connected AVR’s volume and delay per speaker. For volume it would have to make several measurements at a range of volume settings so it can calculate the required mapping of processor master volume level to AVR volume level. Speaker distance and 3D position would also be nice to get for DTS:X/Atmos processing.
What about analog stereo and other sources? Lack of analog inputs on the processor may seem limiting but you can get external Analog to digital converter devices to solve this. Another cool feature of this setup is using the AVR control already setup above you can take one of the AVR’s that has a digital zone 2 output option and connect the source to this device instead. Get a digital cable linked back from zone 2 out to digital in on the processor and when the source is selected zone 2 activation control commands are sent and you get the selected audio. May add just one stereo RCA with a ADC to use with AVR’s zone 2 analog pre-outs would be helpful. Could possibly use this same external zone control with loopback trick with some AVR’s with HDMI Zone out option to add more HDMI inputs allowing less to be required for the Processor.
For basic setups you would only need this processor and one AVR to get 7.1 channel setups. Two AVR’s would do nearly any channel setup you wanted. Adding any more would only be done for extreme power requirements and special cases.
The secret to this solution is the fact that almost everything is now controlled in software and support for different AVR amps is just a firmware update or script update away (creating a file format to describe the control codes used for an unsupported AVR brand/model would allow end users to do this work). Support for new sound formats is also just firmware updates as well and even if the AVR’s being used don’t support DTS:X etc instead of replacing it we can still use it and we are not limited by the number of speaker channels of our original Amp. Bi-amping or Tri-amping any channel we want is a simple config change away as well. Want to separate left from right channel just assign one AVR for Left channels and one for Right. Zone 2/3/4 support is just a software setting as well and you can run a long HDMI cable to a second room and control an AVR in that second room and still get the latest room correction and processing from this older AVR.
DSP performance limitations may limit the processor to 10-12 active speaker channels. Would be cool if you could get two processors and chain them together with an HDMI and Network connection so that the second (or third?) slave processor handles another 10-12 channels allowing larger DTS:X setups for large rooms. This works by passing the currently select source before it has been modified to the second unit via HDMI and then over the network letting it know which speakers it should process form this source.
Another extra feature that could be added that no one else offers is to allow end users to remove the analog crossovers and do it all digitally with one amp channel per speaker cone type. There are kits and processes right now for doing this conversion to turn a passive multiway speaker into an active monitor speaker as an all in one type unit. But instead for this solution we simply bypass the crossover and connect the internal speaker elements to the outside. As an example a 3 way speaker might have 2 woofers wired together and to the outside and the midrange and tweeter wired outside as well. 3 Amp channels are assigned to this one speaker and then the microphone based speaker calibration routine is run but each element is tested separately and then the integration of three frequency bands is calculated and crossover points selected. This is sort of the same process speaker designers use to design their own crossovers but instead we are replacing it with an automated process. Would be good to give options to customize the crossover points by hand so end user can tune it to their own tastes and knowledge of the speakers limitations. The advantage of this systems are many. It optimizes not just the speaker on its own but also for your room and listening environment. Each speaker element can have different level and time delay if needed. Digital crossovers can be far more customizable and efficient than analog ones which often use LCR losses to heat to remove unwanted frequencies/levels. Instead we just don’t output the un needed frequencies to the amp and the speaker receives power for just the right range and there is very little lost energy. It is not a perfect system as it can’t easily overcome the supplied speaker drivers physical limitations. You can EQ out frequency problems but the speed of the speaker and other factors are much harder to solve.
Problems with this Idea: First is who would bother to build it? All the companies that are skilled in this kind of thing already have their existing analog centric pre-amps they are trying to sell or AVR’s. They will not want you re-using older generation AVR’s as Amps as they want to sell you their new stuff. A smaller company that doesn’t make AVR’s or amps could build it but then they wouldn’t have the skills to do all the DSP sound processing etc.
Second possible problem is HDCP protection. Not sure if they would let you terminate the HDCP connection in the processor and then pass the raw PCM version of the decoded audio out via digital audio. May be able to re HDCP encrypt the data going out to the AVR’s to cover this. It may also not be an issue being un-encrypted as it is not the raw digital source we are outputting but a modified volume adjusted and room corrected pre-amp digital signal and not the original source that you could easily copy.
Final problem is how to charge for something like this. The build cost per unit is low because it has little hardware in it compared to existing equivalents but still costs money to design and the main cost is in the software development involved.
Anyway just my idea that will never be built
