Tube Valve 5.1 Home Theater [Early Research]

[dannyf]

Has anyone characterized tube-amplifier distortion and simulated it in a DSP?

Two (potential) sources for "tube sound":

1) clipping behaviors: tubes clip gentally vs. hardly by solid state amps. This allows the similarly rated tube amps to handle more peak power without sounding as harsh as ss amps. This part is likely difficult to be characterized by a dsp.

2) part of the tube distortion comes from transformers (particularly output transformers). It is fairly doable with a dsp.

Multitudes of experiments have shown that when not pushed outside of their performance envelops (not clipping), tube amps are indistinguishable from ss amps.

But to remain within its performance envelope, ss amps will need to be rated much higher than ss amps, when content of high crest factor is played.

[ampdoctor]

As Nick said, they've been trying to model tubes for decades with very little success until fairly recently. It's close, but it still not quite right. Having said that, once the audio gets past the mixing desk and over to the mastering house, the whole argument becomes moot. At that point the signal has been pushed through so much digital signal processing software that nobody is going to hear the difference. As such, I stand by my original comment that tubes are great for sound production at the front of the chain, but utterly pointless for reproduction.

[VK5RC]

The Australian popular electronics magazine "Silicon Chip" in its August 2014 out today, publishes a Valve sound simulator, it is described as having some high level clipping with mostly even order harmonics and a "soft' sounding spectrum with a bit of bass resonance circuits built in as well. Uses FET 2N5485.
I don't like that "Valve" sound myself but accept others do.

[ampdoctor]

There's so much more to the "valve sound" than just second order harmonics. It's even AND odd order harmonics as well as their ratios and whether or not they're higher or lower order. Depending on the program material you'll get compression and/or hard limiting on the positive excursions depending on the magnitude and the frequency of the signals while the negative going signal remains relatively unadulterated. But that's not always the case as cathode bypass capacitance and local feedback can play havoc with this as well. Add in some sag and ringing from a shit power supply, and poorly designed output transformers, and a truckload of negative feedback on the power amp. Attempts to emulate the effects of all this at the same time is really like trying to hit a moving target and the result is a close approximation at best.

[SirNick]

That's probably true, but if you modeled each of those elements individually, the net result would probably be as close to the behavior of any given real tube amp as any other given tube amp is.  I.e., it doesn't need to be a perfectly accurate reproduction of exactly one -- just a realistic emulation of any one.

I think Peavey's amp sim product tries to do this (simulate the components), but I'm not sure how well they pulled it off.  I got a demo a couple years back, but it didn't work very well for some technical reason that I can't remember now.  Many of the others (Waves, etc.) seem to follow the "complete signal path" modeling approach.  There are, indeed, probably too many variables to make the complete model behave identically to the original, even if a particular demo clip sounds similar, and that might be OK if it sounds good in its own right.

[bob808]

You could also go for a hybrid setup, with tube preamp and SS output. That should give a bit of "warmth" or "tube sound" to the whole setup.

[Lightages]

I have yet to see a budget attached to this. The pricing could go from $200 per channel to almost unlimited.

You can eliminate the output transformers as there are such things as output transformerless amps, or you can buy off the shelf transformers, or even get them custom made.

Do you want to power any random speaker? Are you willing to buy or make speakers specifically made with high efficiency just for this purpose? Is the LFE channel going to be tube also?

Do you want to go silly audiophool or practical? How many watts? How many tubes per amp are you willing to manage?

The only I see for using tubes is because you like them, rather than they are a better way of reproducing sound. If you agree, then using tubes for such a practical and specific use of a 5.1 theatre sound system is for bragging rights just because you want to. To do a 5.1 system correctly means having some power, and low requirements for maintenance and setup. Tubes rarely offer this at low cost.

So what is your cost limit? Why do you really want tubes?