R. L. Drake also makes equipment that can do this for less than Scientific Atlanta charges. I priced some of it last year. I only remember the rough order of magnitude of the component pricing (too high for my application) which was to receive and demodulate broadcast 8VSB on one UHF channel and convert it to another for alternate channel reinsertion into the broadcast 8VSB band where I was concerned about otherwise having an adjacent channel interference problem. I think that a QAM demodulator, an 8VSM modulator, the supporting rack cage and power supply would run you somewhere around $1,000 to $1,500.
R. L. Drake and Blonder-Tongue both make devices that take the 30 MZ transponder-width QPSK DirecTV and DISH network transmissions and somehow convert them into 6Mz wide QAM (64QAM, maybe?). I think this was done in a single device, and if it is, such a device would seem to internally incorporate functionality comparable to, but not identical to, what sdossick wants to do, so the price of that device might be a guideline as far as the likely market price of an integrated QAM to 8VSB converter if anyone decides to produce such an animal using the prevailing commercial signal processing architecture.
But for member sdossick's purposes, the output of the R. L. Drake, Scientific Atlanta, or Blonder Tongue demodulator alone would probably be more useful than a reconstituted 8VSB stream, since it would, at that point, be in suitable form for viewing or recording. But that is what he can already get out of a consumer-grade QAM tuner, which costs less than any of the commercial product alternatives.
Analog broadcast NTSC signals can be economically converted to other broadcast and cable channels through a process called heterodyne conversion in which the broadcast signal is mixed with a fixed oscillator output while the NTSC analog content of the signal remains the same. The cost of such circuitry, which is used in all agile NTSC tuners to bring broadcast and cable TV channels down to so-called "intermediate frequency", is no more than a few dollars per mass-produced unit. But converting from QAM to 8VSB requires reprocessing the signal twice, first in the manner of its ordinary processing for viewing, and then in the manner ordinarily used for making this intermediate product suitable for transmission. There will never be significant demand for such a product so as to warrant its mass production, and its cost would always be greater than the cost of a simple demodulation device.