wrote:
>
> So, just how long will flash be
> around before someone invents some new sexy "standard"? It could be
> tomorrow. I've got a perfectly good digital camera tied to the older
> PCMCIA form.
>
> Compact Flash has a roadmap of between five and 10 years, but our machines
> aren't necessarily tied to CF, because we made the decision to have the PC
> Card form factor which allows all sorts of other media ( Microdrives, ATA
> Flash, Sony Memory Stick, DataPak etc. ). It's the bus rather than the
> card that we are working with. In our major markets such as broadcast, we
> are looking to have product ranges with long lifetimes, because our users
> demand it. At the minute solid state is still in an interim phase, so we
> are keeping the options open as to developments here.
I'm certainly hoping the PC Card form and slots stay around, I've got
quite a few thousands of dollars tied up in a digital camera based on
this form. It's one of the real annoying things about computer stuff
that they often seem to change just for change's sake. Or the
alternative view, to force us to buy a new set of equipment. It's been
quite a while since any of these changes really improved the
functionality for what I do, and I do pretty heavy graphics and audio. I
think a lot more than just the pros are looking for long lifetimes.
Computers should be able to last as long as TV sets, and I'm using a 20
year old TV set.
It's the life expectancy that worries me the most in these new forms of
recorders. In my lifetime I've watched a lot of things come and go. With
each change in technology the life expectancy seems to have gone down.
If it was compensated by cheaper prices it might be ok. But a solid
state recorder certainly does not fit that so far. We keep getting
promises about the flash memory being cheap, but it never seems to
happen. Or maybe their definition of cheap and ours is different.
> Most of the things using the compact flash assume it went ok and don't
> really test or even provide a testing method. They really are not doing
> much to the memory, so it probably will do most of the time. I think you
> can forget having a recorder that can both play the recording as you
> make it from the chip and record at the same time to the chip. The
> programmers seem to assume perfect reliability.
>
> This is a good point, and one that I have discussed extensively with our
> engineers in Japan. In some ways playing from the flash whilst recording
> is not really an issue, depending on where the monitoring is taken from in
> the circuit, since it's extremely unlikely that you would have a failure
> in the actually write process. If there is a failure, then it is much more
> likely to occur in the logic or software, so if you are listening at some
> point in this chain ( rather then you would hear any failure. However, I
> would very much like to have an effective way of proper confi monitering.
> It is possible in software to read and write from the flash
> simultaneously, but the main issues are memory overheads in doing this. Be
> assured though that we are looking at the possibility of doing this.
Digital tends to be all or nothing, and I'm comfortable monitoring off
the chain. The idea of the third head monitoring for tape was a good
idea. With tape you could just have a head get dirty or the transport
slip and mess up, so monitoring the tape itself was important. Now it's
probably more like habit than necessity.
To my mind the better use of memory buffers is as a record ahead buffer.
It would be nice on things like MD if it would have a buffer long enough
to allow the drive to not be spinning in record pause. That would make
it so that pause would not use battery as fast, or at least I think it
should. Same would be true of any of the mini HD forms. Of course solid
state is in effect, it's own buffer.
Walt
________________________________________________________________________
________________________________________________________________________
|