Yep. The licensing on the silicon must be pretty expensive. That's the only reason I can see the cost being so high given the parts. Technology marches on...

Meanwhile, the closest I can find to my ideal Class-D 7-channel amp is the Rotel RMB-1077. It's price has reportedly just dropped to under $2000.

At 100 watts per channel, it runs extremely cool, and only weighs 17 pounds. But it only offers analog, unbalanced inputs.

Wouldn't it be great to have an amplifier like this that had a HDMI for its audio input?

Dave K.

Probably not, for a couple of reasons.

Connectivity: What are you going to connect it's HDMI input to? I don't want to integrate a full surround processor into my amp, so I'd need my processor to have a dedicated HDMI audio output that provides conveniently decoded, processed, EQ'd, and level-balanced multichannel PCM separate from the HDMI video output that goes to my display. There isn't such an animal available now.

DAC's are also an issue. The method by which they amplify an analog signal is conveniently called "digital," but these "digital" amps are still acting on an analog signal. If you input a digital signal, you need to add a multichannel DAC into the amplifier. That DAC would convert the multichannel PCM into analog. But since your pre-amp/processor will still need DAC's and analog outputs for use with the subwoofer and with other amplifiers, that means you just bought two multichannel DAC's.

Yes Gonk, you are correct. I'm hoping that Outlaw's 997 will have at least two HDMI outputs.

The goal ought to be having NO analog connections between digital components- including Class-D amplifiers (if they ever become widely available).

Class D amps are not digital components - they are still analog, and you still need analog connections to them.

Not sure where you're getting your info, Gonk, but Class-D amplifiers are switching mode amplifiers that operate on a PWM signal. A digital input (like SPDIF or HDMI audio) can be directly converted to PWM in a DSP chip without ever going into the analog domain. In fact, HDMI supports sending multi-channel PWM (a la DSD, as is used with SACDs) directly.

A pre-pro internal digital signal that is converted to an analog output and fed to an analog input on a Class D amplifier must still be converted to PWM inside the amp. This adds an unnecessary D/A conversion at the output of the preamp/processor, and also requires the signal to travel in the analog domain through conventional analog cables between the two boxes- things that could be avoided if the signal was left in the digital domain all the way to the DSP in the amp.

It has been a while since I read up on class D, so I had to go back and do some digging. I have read in several places that the common term "digital amps" is a misnomer, with the "D" simply being the next letter in line at the time that the technology first appeared. (See this Wikipedia entry, which mentions both the class D identifier and touches on PWM as pertains to analog signals.) Another article that touches on digital vs. analog input to class D amps is this Audioholics article. It does point out that it is technically possible to build a class D amp that accepts a digital signal (which I had either not heard before or had forgotten), but it goes on to suggest that this approach may actually be inferior to a design that operates with an analog signal.

Glad to see we're on the same page again, gonk.

I am familiar with that Audioholics article and other writings by Bruno Putzeys. When he wrote that in 2004, he was indeed describing what was the current state of the art of analog vs. DSP at that time. But DSP chips have gotten a LOT faster and more powerful since then- which has closed the performance gap considerably- perhaps to the point of inaudability.

In any case, I shall continue to lust over the possibility of someday eliminating the giant wad of cables connecting all of my components together without [audible] sacrifices in sound quality.

If the technology can be made to work, the idea of a digital output to the amps is definitely intriguing, but even if we see those amps enter production I worry about trying to do it with today's HDMI. We start with the nightmare that is splitting HDMI - most processors with two HDMI outputs don't have both active at the same time because of the HDCP problems that it generates. They we get to the issue of controlling the audio output via HDMI: is it PWM or PCM? While the HDMI spec may theoretically work with PWM, has anybody actually done it? And when will processors incorporate the software to provide that signal? Then we have the issue of what you do if you don't have all of your amp channels in a single amp chassis. We have theaters with five channels, seven channels, and now (with the arrival of PLIIz) possibly even nine channels. How do we handle that if not by creating a single audio connection for each speaker? That connection gets messy in a hurry unless you require people to buy an amp with all of the channels in a single chassis.

I suspect the realities of electrical engineering plus economies of scale are conspiring. It's cheap to build really good amps using proven techniques, and so we keep doing it.

Digital amps, despite what some more vehement supporters asserted almost a decade ago, are hard to make right for an arbitrary load. *I* had hoped we'd see more self powered speakers and within those, class D amps that were precisely matched to the driver they powered. No dice though.

I do find the self powered Martin Logans intriguing though.