Fun with Bayer interpolation

[Ron Parr]

No, he (and everyone else) carefully reports all three times (FPS
until full, FPS when the buffer is full, and buffer flush time. If
you think the buffer flush time is sooo unimportant, go over to the
Oly forum and read the flame wars about the E10/E20. It has the
very X3-like characteristic that the fill time may be fast but the
flush time is slow. (It also buffers the same number of frames
whether in RAW or JPEG.) See how popular this design choice is.
Some people see it as a fatal flaw, others say they can live with
it as they don't need this kind of performance.

One difference here is that I'm assuming it's relatively easy to increase buffer size by using cheap and plentiful memory solutions.

I do agree that slow buffer flushes can by annoying. If you had pushed this point from the beginning, I would have agreed that it can be viewed as a drawback by some.

Perhaps the hidden message in what you're saying is in the slow
response times issues I raised about.

You still keep reading FAR more into my claims than I've ever said.
All I've ever said was either SLOWER or MORE EXPENSIVE I/O. Not
unusable or unacceptable (although perhaps dissappointing to some.)
The inevitable consequence of producing X3 more data. I've also
said that based the cost/performance curves we have from current
cameras, this means that any sensor cost advantage may be negated.

Well, I don't think anybody claimed that the sensor would be less expensive than, say, Canon's CMOS sensors.

Any savings from elminating other parts in the camera would likely be in reduced design time, which could be a big deal for a firm like Sigma, rather than reduced parts cost for a $3000 camera since the part's we're talking about (either added or removed) would not be a huge part of the cost for the camera.

Having a larger buffer won't change the FPS rating of the camera.
All it will change is the length of time at which it can sustain
that rate.

Finally, we agree on something. (Is the world ending?) But all
cameras have 2 FPS rates: one burst and one sustained. Every review
measures both (unless the camera is too cheap for them to be any
different.) You keep ignoring or dismissing the second one as
unimportant. Fine. It does not matter to you. It obviously matters
to a large number of other photographers.

It's actually pretty difficult to find the sustained FPS numbers for many cameras. I agree that they matter to some types of users for good reasons, but it's not one of the more prominent statistics used to describe the cameras. For example, it typically does not appear in Phil's summary page for cameras.

Of course. That's likely one of the reasons the D30 is slower than
the 1D. Stopping and restarting has a price on both ends. BTW, I
consider the D30 "state of the art" for it's price class. If you
are saying you need something at least as good in the SD9, then you
are in the unhappy position of agreeing with me again. (Actually, I
suspect we agree 95% of the time. It's just when we disagree, we
let it get out of hand.)

Well, the D30 is two year old design (announced PMA 200) that has been discontinued. I don't view it as state of the art.

What I'm saying is that the easy solution, which requires no special dual ported memory, is currently adopted by some cameras that Phil has tested (e.g. D30). So, I don't agree with your claim that dual ported memory is needed. I also argued that if one demanded that writes to CF continue during shooting, one could achieve very close to dual ported speeds by introducing a small and inexpensive additional buffer (or by exploiting fast memory speed).

--Ron ParrFAQ: http://www.cs.duke.edu/~parr/photography/faq.htmlGallery: http://www.pbase.com/parr/

 

[Erik Magnuson]

Funny, the initial subject name of this thead. I watched this
instead of TV today. Loved it.

Glad to see there's some benefit in it. (Thank you, Thank you. We will be appearing next week in rec.photo.digital. Enjoy the veal.)--ErikFree Windows JPEG comment editor http://home.cfl.rr.com/maderik/edjpgcom

 

[David SL]

The discussion was about frames per second for the camera. FPS
ratings for digital cameras are always measured while the buffer is
filling, not in terms of throughput to CF. Check the FPS rating
for any camera reviewed by Phil. All of these are rated at a
particular speed until the buffer fills.

No, he (and everyone else) carefully reports all three times (FPS
until full, FPS when the buffer is full, and buffer flush time. If
you think the buffer flush time is sooo unimportant, go over to the
Oly forum and read the flame wars about the E10/E20. It has the
very X3-like characteristic that the fill time may be fast but the
flush time is slow. (It also buffers the same number of frames
whether in RAW or JPEG.) See how popular this design choice is.
Some people see it as a fatal flaw, others say they can live with
it as they don't need this kind of performance.

Perhaps the hidden message in what you're saying is in the slow
response times issues I raised about.

You still keep reading FAR more into my claims than I've ever said.
All I've ever said was either SLOWER or MORE EXPENSIVE I/O. Not
unusable or unacceptable (although perhaps dissappointing to some.)
The inevitable consequence of producing X3 more data. I've also
said that based the cost/performance curves we have from current
cameras, this means that any sensor cost advantage may be negated.

Having a larger buffer won't change the FPS rating of the camera.
All it will change is the length of time at which it can sustain
that rate.

Finally, we agree on something. (Is the world ending?) But all
cameras have 2 FPS rates: one burst and one sustained. Every review
measures both (unless the camera is too cheap for them to be any
different.) You keep ignoring or dismissing the second one as
unimportant. Fine. It does not matter to you. It obviously matters
to a large number of other photographers.

You don't need dual ported memory to implement a buffer of the type
we're discussing. The first and obvious thing you can do is just
pause writing CF when you're feeding in a new image from the
sensor. This is what the D30 does.

Of course. That's likely one of the reasons the D30 is slower than
the 1D. Stopping and restarting has a price on both ends. BTW, I
consider the D30 "state of the art" for it's price class. If you
are saying you need something at least as good in the SD9, then you
are in the unhappy position of agreeing with me again. (Actually, I
suspect we agree 95% of the time. It's just when we disagree, we
let it get out of hand.)

Perhaps there have been misunderstandings between us in our
dialogue, but I don't think that any of this justifies the tone
that things have taken.

Well, I did get a chuckle over the idea that this forum somehow
needed to be protected from me. I spent a lunchtime plotting on
how I could use my evil influence to take over the world.
Fortunately, I forgot how I was going to do it a few hours later,
so you are all safe now.

I don't take it personally and harbor no hard feelings. We'll know
a heck of a lot more once cameras actually get into reviewer's and
user's hands. Until then, it's all just hot air.
--
Erik
Free Windows JPEG comment editor
http://home.cfl.rr.com/maderik/edjpgcom

--
The significant problems we face can not be solved at the same
level of thinking we were at when we created them.
(A.E.)

--DSL

 

[Erik Magnuson]

I do agree that slow buffer flushes can by annoying. If you had
pushed this point from the beginning, I would have agreed that it
can be viewed as a drawback by some.

It's been my point all along. It's what I mean by I/O. The "O" has to go somewhere useful.

Well, I don't think anybody claimed that the sensor would be less
expensive than, say, Canon's CMOS sensors.

Are you reading the same Forum I am? When the Foveon zealots claim that it will drive mosaic sensors from the market (because of price among other things), I've not seen any "unless Canon pefects their CMOS sensor" caveats. And since the Canon technology in question has yet to appear on an under $1000 camera, it's hard to assume that it's inexpensive.

Any savings from elminating other parts in the camera would likely
be in reduced design time, which could be a big deal for a firm
like Sigma, rather than reduced parts cost for a $3000 camera since
the part's we're talking about (either added or removed) would not
be a huge part of the cost for the camera.

Like I said, it's all speculation at this point. Personally, I think that this design choice will limit the acceptance of the SD9 unless an SD9a with this fixed comes out quickly. When a camera lacks something that's common (and commonly used) in every other camera in its class, customers tend to be unhappy.

It's actually pretty difficult to find the sustained FPS numbers
for many cameras.

The only reason it's hard to find is because for most cameras it's the same as the single shot write time. It's in the timing section of every camera that supports it under 'Timing and Sizing: Continous Drive Mode'.

Well, the D30 is two year old design (announced PMA 200) that has
been discontinued. I don't view it as state of the art.

Well, until someone introduces something better, it's still "state-of-the-art". Is there a better camera in the $3k price class?

What I'm saying is that the easy solution, which requires no
special dual ported memory, is currently adopted by some cameras
that Phil has tested (e.g. D30).

Well, let's revisit this when the SD9 is reviewed/shipping. If I'm wrong and the SD9 has better sustained continuous drive mode performance than the D30 or D60, I will buy you the beverage of your choice. No obligation.--ErikFree Windows JPEG comment editor http://home.cfl.rr.com/maderik/edjpgcom

 

[Ron Parr]

I do agree that slow buffer flushes can by annoying. If you had
pushed this point from the beginning, I would have agreed that it
can be viewed as a drawback by some.

It's been my point all along. It's what I mean by I/O. The "O" has
to go somewhere useful.

We'll, I don't think this was exactly what was being argued, but I'm pleased that we've converged on somehting.

Well, I don't think anybody claimed that the sensor would be less
expensive than, say, Canon's CMOS sensors.

Are you reading the same Forum I am? When the Foveon zealots claim
that it will drive mosaic sensors from the market (because of price
among other things), I've not seen any "unless Canon pefects their
CMOS sensor" caveats. And since the Canon technology in question
has yet to appear on an under $1000 camera, it's hard to assume
that it's inexpensive.

Any claims that it will drive Mosaic sensors from the market would be motivated by the knowledge that it's no more expensive than CMOS + Bayer interpolation.

It's generally agreed that CMOS is a less expensive technology than CCD in the long run. There are good reasons why it has currently been used in larger sensors and why it will migrate to smaller sensors shortly.

The Foveon approach has the potential to offer superior performance with with fewer spatial pixels on the surface of the chip, which implies larger pixels. Thus, they are pushing the curve forward faster than what would happen normally with Bayer interpolation anyway.

Like I said, it's all speculation at this point. Personally, I
think that this design choice will limit the acceptance of the SD9
unless an SD9a with this fixed comes out quickly. When a camera
lacks something that's common (and commonly used) in every other
camera in its class, customers tend to be unhappy.

This could well be correct. The product seems to smack of "first generation." I've never disagreed with you or anybody on this point.

It's actually pretty difficult to find the sustained FPS numbers
for many cameras.

The only reason it's hard to find is because for most cameras it's
the same as the single shot write time. It's in the timing section
of every camera that supports it under 'Timing and Sizing:
Continous Drive Mode'.

Hmmm...

Well, there are some differences. They're not that big for the D1X, but they're pretty big for the D30, presumably because it pauses writing to CF when it's taking new shots.

In any case, the single shot write times aren't listed in Phil's summary pages for cameras either.

I don't disagree that some people care about it, but it doesn't seem to be so important that it's listed alongside with the buffered FPS rate.

Well, the D30 is two year old design (announced PMA 200) that has
been discontinued. I don't view it as state of the art.

Well, until someone introduces something better, it's still
"state-of-the-art". Is there a better camera in the $3k price class?

I don't want to split hairs about this, but when you said, "state of the art or better" I assumed you meant using top shelf current technology. The D30 is two years old, however if you want to hold that out as state of the art, then it will be easier for Sigma to achieve "state of the art" performance.

What I'm saying is that the easy solution, which requires no
special dual ported memory, is currently adopted by some cameras
that Phil has tested (e.g. D30).

Well, let's revisit this when the SD9 is reviewed/shipping. If I'm
wrong and the SD9 has better sustained continuous drive mode
performance than the D30 or D60, I will buy you the beverage of
your choice. No obligation.

I never claimed that it would have better sustained rates than the D30 or D60. I would expect that it would have less than either of them.

--Ron ParrFAQ: http://www.cs.duke.edu/~parr/photography/faq.htmlGallery: http://www.pbase.com/parr/

 

[Erik Magnuson]

We'll, I don't think this was exactly what was being argued, but
I'm pleased that we've converged on somehting.

Well, it's what I've been attempting to argue.

This could well be correct. The product seems to smack of "first
generation." I've never disagreed with you or anybody on this
point.

The only question is whether these "first generation" problems are there because they are hard to solve or just inexperience with the technology. No data yet.

I don't want to split hairs about this, but when you said, "state
of the art or better" I assumed you meant using top shelf current
technology.

To me, it's not "current" until you ship a product with it. Too much vapor out there. It reminds me of the saying:

"In theory, there is no difference between practice and theory. In practice, there is."

I never claimed that it would have better sustained rates than the
D30 or D60. I would expect that it would have less than either of
them.

OK. The offer still stands if you care to take me up on it. (Notice how little we disagreed this time? Just a little at the very top. Maybe there's hope for peace in the Middle East yet.)--ErikFree Windows JPEG comment editor http://home.cfl.rr.com/maderik/edjpgcom