You really think so? Actually, the reason for the price is just
that we're paying for the R&D, and the profit Canon wants to make.
They can easily change the price tag from $6,000 -> $3,000, from
$3,000 -> $1,499, they can do $799, too.
Well, I never like betting against human ingenuity, but I really do
think that big sensors are expensive to manufacture and may not
drop in cost very quickly. If you think otherwise, I'd love to see
an example where anybody has manufactured a CMOS chip the size of
the 10D sensor for less than several hundred dollars.
The only reason they won't sale 1Ds for $1,499 or 10D for $799 is
that there are no competiter yet. I recall last year when D60,
D100, S2Pro, and SD9 just announced their list price was set to
$3,000, and people celebrate for those "afforadable DSLR"...
I'd like to see some facts to back up these claims. How can Canon
make CMOS chips at lower cost than anybody has ever been able to do
despite 30 years of trying? Perhaps you have an argument that
sensors are much less sensitive to defects than other kinds of
chips. This is a reasonable starting point, but I'd like to see
somebody carry the argument through.
You're certainly right that competition has forced manufacturers to
accept lower margins on their products, which has probably slowed
down R&D a bit and led to more incremental products like the D60 ->
10D upgrade.
Just in several year. When the R&D is paid off. When there're
enough market for DSLR, under $1,000, or even under $500 DSLR would
not be a dream.
If it's only about amortizing R&D costs and has nothing to do with
variable costs then Canon could do it now. I have little doubt
that the demand curve is such that they'd quickly recover their R&D
costs in added volume. There was probably also some squeezing of
their distribution channel involved too.
The fact that they haven't dropped the price already this is
actually pretty good evidence of high variable cost per chip.
Let's crunch the numbers: A big manufacturer like Canon might able
to tolerate a 1.5X markup to their channel and then push their
channel retrailers to accept just a 1.5X markup when selling to
consumers. (It's usually 2X for both of these for smaller
producers. I'm giving you the benefit of the doubt here.)
If we believe that, apart from amortizing some R&D costs, Canon
could sell the camera for $1000 (let's say $990 to make the math
simple) today, then this means that it can't cost them more than
$440 to make today. So, at a $1000 price, this would give Canon a
$330 margin per camera (990 - (1.5 * 440)).
Now, let's suppose that Canon is currently selling to distributors
at a price that yields a $1500 street price with a 1.5X markup.
This means that Canon is accepting $1000 for the camera and
clearing $560 per camera.
So, if your hypothesis about Canon's cost structure is correct, it
would take only a 1.7X jump in demand (560/330) at the sub $1000
price point for Canon to take the plunge. In fact, it would
require even less than this because Canon would make money on
accessories for each additional camera they sell and they'd place
value on stealing market share from their competitors.
To make a convincing case that Canon can pull of the sub $1000
digital SLR in the short term, you're going to need to establish at
least one of the following:
- Evidence that Canon has figured out how to make CMOS sensors for
less than it costs eveybody else on the planet to make similar
sized CMOS chips for other functions.
- A compelling argument that the current high prices are simply the
amortization of R&D costs by arguing that demand for such cameras
would not increase significantly if Canon reduced the cost to
$1000, allowing Canon to amortize the costs even faster.
I really do hope you're right and that Canon can do this. I just
want to damp what I think are some unwarranted generalizations from
other high tech areas. Making big chips is really, really
difficult and expensive. Traditionally, manufacturers have avoided
this problem by dramatically increasing the amount of stuff they
cram on small chips, while only very slowly cutting the cost per
sq. mm of silicon.
--
Ron Parr
FAQ:
http://www.cs.duke.edu/~parr/photography/faq.html
Gallery:
http://www.pbase.com/parr/
My take on everybody shouting about only using Canon lenses, and not using 'crappy' Sigma & Tamron (until I see first hand, at least) is that they are just a little too brand loyal.