I Wonder if Canon Development & Marketing Reads this Forum?

[banpreso]

What do you guys think? This site is an invaluable databank for development and market of canon gears. And I don't know if anyone outside of this site "upgrades" their bodies as some forum members here do...

[RDKirk]

There's no evidence that they read any forums.

Canon USA's director of technical media relations, Chuck Westfall, used to correspond on the old Rob Galbraith forums, but reading between his lines, it doesn't appear that Canon Inc even pays much attention to him.

Canon Inc might have people reading the Japanese language forums, but IMO the signal to noise ratio in forums is too low to spend much time on them for market research. The indications are that Canon does conventional market research with focus groups and such...and predominantly Japanese focus groups at that.

[alduin]

I still have a direct print button.

[tommykjensen]

Well they market the 7D as "You designed it Canon built it" so they must be reading something on the net otherwise how would they know what users want?

[Mark1]

This is not the gold mine companies want. If they did what we thought, they would have 1 model of camera that was way to specialized for the common person to understand. Then it wouldn't sell and then close down the company. Not a good Idea.

They have to do the same thing car companies do. No matter how "specialized" the car (read Viper, Corvette) Grandma still needs to be able to get the groceries and go to church with it. They have to cover several markets with each model. Not ideal for the 'enthusiast", but it keeps the company running.

[MikeFairbanks]

Yeah, I'd like to see them make a fully manual full-frame camera for really cheap. No bells and whistles. Just a good, solid camera like in the old days.

[RDKirk]

Yeah, I'd like to see them make a fully manual full-frame camera for really cheap. No bells and whistles. Just a good, solid camera like in the old days.

The "bells and whistles" aren't what makes a modern camera cost a lot. The only way to make anything inexpensively these days is by modern manufacturing techniques, and the "bells and whistles" aren't what make the difference in price. The difference in price between a 1Ds and a 5D is not in the length of the feature list, it's in the difference between being a consumer-level-construction camera and a "good, solid camera like in the old days."

Even in the old days, the F-1 had fewer "bells and whistles" than the FTb, but cost a lot more because it was a "good, solid" camera.

[Mark1]

Certain items take a big part of the budget. When Sony made their cheep FF camera 1/3 of the savings came from making the viewfinder a 96% vs. a 99% coverage. Not from software items.

[RDKirk]

Certain items take a big part of the budget. When Sony made their cheep FF camera 1/3 of the savings came from making the viewfinder a 96% vs. a 99% coverage. Not from software items.

That's the kind of thing I'm talking about. A high-precision viewfinder has always been a big-ticket item ('way before digital) in the top-line cameras because it involves high-precision manufacturing and tight assembly. That has always been costly.

But along with the high-precision viewfinder comes the small spot metering capability and 45-point focusing...both of which depend on a high-precision viewfinder.

[Mark1]

Not hi precision. It is % coverage of the framing the sensor sees. And actually the AF has nothing to do with the viewfinder. Manual, obviously yes. But the auto focus is actually located in the bottom of the camera not the top..

[RDKirk]

Not hi precision. It is % coverage of the framing the sensor sees. And actually the AF has nothing to do with the viewfinder. Manual, obviously yes. But the auto focus is actually located in the bottom of the camera not the top..

Yes, both have to do with the viewfinder. In any mass produced item, there are points of manufacturing that require very tight tolerances and points where tolerance can be looser. That's the basic principle that makes mass production possible--knowing where you can and can't have a degree of slop. Otherwise, you're talking about a custom production line.

The focusing screens of all SLR cameras must have tight tolerances in their vertical positions in their frames atop the mirror box, otherwise manual focusing is impossible. Their tolerance is in their horizontal position--what we see in the viewfinder as "left, right, up, and down." The Rebel-level cameras have about 2 to 3 mm of tolerance--the focusing screens can be as much as 2mm high, low, left or right of true center. Yes, they can even be canted. The screens are heavily masked to less than the actual sensor size, which masks the slop.

Because the screen may be as much as 3mm off center, the screen markings have to allow for that slop. That means the camera can't have a very small spot meter or very many tiny focus points or the position tolerance might place the screen markings completely out of the meter or focus area.

In order to produce a viewfinder with a 95% or better coverage, the focusing screen must be very precisely positioned in its frame--that takes extremely tight machining of the focusing screen and its frame atop the mirror box, in addition to accurate positioning of the frame with the sensor mount. When the tolerance is that tight, there is no room for slop. Precise machining in both dimensions is considerably more expensive than precise machining in one direction. It means the dies and tooling must be more expensive, the rate of parts rejection is much higher, and the assembly process is more complex and expensive.

But once you have a high precision viewfinder, you can have metering systems and focusing systems that make use of it--small spot markings and closely placed focus points.

[tkbslc]

Yeah, I'd like to see them make a fully manual full-frame camera for really cheap. No bells and whistles. Just a good, solid camera like in the old days.

Back then the sensors were not included, though. I am sure they could still do it that way for nice and cheap.

[Mark1]

Light goes through the mirror and down to the auto focus. As far as the AF is concerned the focus screen is only a user interface. It can operate absolutely fine with the focus screen totally removed.

The same with the metering. only a user interface. The metering will be fine with it off center. However the user will not be metering where they think they are. Some metering is done through the screen not on the screen. When you change the screen yo have to tell it what one you are useing because they transmit a different % of light. Some block more than others.

If your 3mm tolerance is an exaggeration for example that is fine. But I cant imagine anything on a camera with a 3mm tolerance. I would suspect a body would get rejected with a strap eyelet that is 3mm off. And the manufacturing tolerances are not as far off as you may thing from the top od the line to the bottom. It is the physical size of the pentaprism that is the determining factor. Much like a diamond. The bigger it is the more it costs exponentially.

[RDKirk]

If your 3mm tolerance is an exaggeration for example that is fine. But I cant imagine anything on a camera with a 3mm tolerance. I would suspect a body would get rejected with a strap eyelet that is 3mm off. And the manufacturing tolerances are not as far off as you may thing from the top od the line to the bottom. It is the physical size of the pentaprism that is the determining factor. Much like a diamond. The bigger it is the more it costs exponentially.

No, it's not an exaggeration. Popular Photography used to publish centering error in their tests, and they did find centering errors that great among Canon Rebels. Canted screens as well.

Yes, pentaprism size is a major cost factor--I certainly was not attempting to claim that viewfinder tolerances were the only cost factor, and I would not claim that pentaprism size is the "determining" factor, either.

You seem to have completely missed my point about the necessary correlation of focus screen markings with focus and metering sensitive areas.

[number six]

3 mm? Geez! That's more like a carpentry tolerance than a machining tolerance. For my own machining I consider .01 mm a pretty tight tolerance...

-js

[RDKirk]

3 mm? Geez! That's more like a carpentry tolerance than a machining tolerance. For my own machining I consider .01 mm a pretty tight tolerance...

-js

That was for film Rebels, with an itty-bitty pentamirror over a big film frame. They have certainly improved them since then--Canon gets better more cheaply every year, which is why we see a 100% viewfinder in the 7D.

[Mark1]

No, it's not an exaggeration. Popular Photography used to publish centering error in their tests, and they did find centering errors that great among Canon Rebels. Canted screens as well. .

WOW! Its no wonder people claim to have problems then. 3 mm is huge on something so small as a camera. No less as small as a screen.

Yes, pentaprism size is a major cost factor--I certainly was not attempting to claim that viewfinder tolerances were the only cost factor, and I would not claim that pentaprism size is the "determining" factor, either..

Your argument here is with Sony then not me. As that was what they pointed out in the press release, as being 1/3 of the drop in price.

You seem to have completely missed my point about the necessary correlation of focus screen markings with focus and metering sensitive areas.

Yes, I have. I will admit it! I understand it is the biggest key part in manual focusing. But it is not even included in the AF process. I have had reason to remove the screen for a few jobs. Without the screen, the viewfinder becomes omnifocus. And if you know the focus points you can still auto focus properly.

[RDKirk]

Yes, I have. I will admit it! I understand it is the biggest key part in manual focusing. But it is not even included in the AF process. I have had reason to remove the screen for a few jobs. Without the screen, the viewfinder becomes omnifocus. And if you know the focus points you can still auto focus properly

However, most customers will not accept a focusing screen with the marked focus points and the marked spot meter area totally outside the actual sensitive areas.

[Mark1]

I get you now. We have been debating to different points. You mean 'precision' as in the ability to locate focus points in the framing and them being actually the points the sensor sees. I have been meaning 'precision' as to the accuracy of the focus itself.