The Riot and iPod are made in large quantities - I suspect the manufacturing runs are at least 50,000 pieces for the iPod, for example. Volume makes things cheap - you can get agressive with the silicon suppliers, you'll have multiple factories dying to get your business to build the thing, etc.

Also, the iPod/Riot are much simpler internally than the empeg. They play music. No audio in chip, no ethernet chip, no serial port, no multichannel out with low impedance drives, no complex power supply, no temperature sensors, no supervisory PICs, a display that almost costs an order of magnitude less, etc etc etc. Yes, they have a battery and a charge management chip, but that's about the end of it.

The software on these devices is also a lot simpler; you notice it with things like the way the iPod locks the ui completely during HDD spinup, the lack of advanced playlist reordering when the player is running, etc. Still, they amortised the software development costs over a lot more units, which means that they can make less margin and still be profitable on the product.

The auto aftermarket, IMHO, would have been able to support the empeg as a low-volume product. We were working on figures for small distribution deals with specialist dealers (people like Cambridge Car Audio, who can be trusted not to screw up the installation, for example). As with any growing company though, cashflow was an issue - you can't build bigger batches from the money coming in from a smaller batch, you have to grow organically. We were hoping that Rio would make a bigger splash marketing-wise (ok, a splash at all) and would fund the (expensive) development of a more consumerised next-gen player (the mk3) which would have had wider appeal, moving it from just specialist aftermarket to high-end general aftermarket. This didn't happen for a number of reasons, the tech bust being one of them - automotive wasn't a core market, so it got cut when savings were being looked for.

Hugo