@cm said: so imho thinking too much on backward-compatibility is honourable, but a handicap for groundbreaking new technologies.
just my two bits, christian
Well, as we say in LA, "You da man!" So here's the R&D gamble (e.g., where opinion + market research confront $$$$$$$).
1. IF the development and implementation DEMONSTRABLY shows that the new technology works best with a P4 x.xxGHz or faster, or it is absolutely HANDICAPPED without it, then you must go P4, with the lowest entry point a P4 2.4GHz
RISK - How many current VSL customers are on a P4? How many are on a P3? How many have P4s fast enough to run the new technology? How much effort is involved to upgrade current systems, even P4 1.8GHz if the new technology requires faster speed? How many VSL owners will do this AND order MIR in light of what they've paid so far for the library? Who would be your competitor if the majority of VSL customers couldn't afford to buy it right now?
2. IF the development and implementation shows that a good amount of use can be had from the PIIIs (Minimum specified requirements) what is the result of running on a PIII vs. a P4, and how big is the sonic difference that would cause the customer to upgrade?
3. Is MIR for the GigaStudio computer or for the sequencing computer (working with the final mix vs real time? If for the Giga, and the customer has 2 or more, how many must be purchased to achieve the desired result? One per system? If one per system, how much (projected) would a customer have to spend to upgrade 2+ systems to have MIR on every machine?
RISK - the price of upgrading isn't just about parts and money. It's about the time to make the upgrade, resetting up the system and testing/checking it out, and a point rarely discussed, changing one's work flow to accommodate the change. For the professional writing music FULL TIME, this is a problem. For the person doing for fun, self-satisfaction, semi-pro, time won't be a factor.
Of course, these are lots of questions for just one press release!
At the end of the R&D day, you still have to sell enough to make back R&D costs plus and create enough sustainable sales to show a profit (you know, steady employment....)
Will it only work with VSL? How about other libraries that were also recorded in stereo and not panned to position?
Basically, the decision path for DSP in aerospace often goes like this:
Do we use a known chip with a proven performance record or try a newer chip that doesn't have the performance track record.
For example, most communications satellites, pretty advanced technology, use PI chips.