New Canon digital SLR - the 20D is official

[jim_larson1]

I have *no doubt* that this camera will produce better images than a 10D. I hear the "more noise due to smaller pixel argument", but I *refuse to believe* that this sensor is not more advanced than the current 6mp sensors and was specifically designed to maintain, if not improve, image noise.

 

A 20D being a yawner is almost as inevitable as the specs. Most people on this forum were agreed months ago that a new camera would look alot like this one.

 

Consider: In February 2003 the *cheapest* dSLR was the D60, and that camera had been *off the shelves* since November 2002. In February 2003 there was a significant pent up demand for products. If D60's were re-released in February at $1999 (the going price for D60's), they would have flown off the shelves. So out comes the 10D at 75% of the cost ($1499) and an AF system that actually worked. . people (like me!) snapped them up.

 

Now look at the market today: The 10D's are in plentiful supply. If you want to go cheap, you can get a 300D for 60% of the 10D price. So naturally, even if this camera does have a step change sensor improvement, you won't have droves of people lining up because the demand for dSLR's is not there like it was 18 months ago.

 

But it would be silly to buy a 10D today with this camera available in a month.

 

****

 

On another note: I think the pricing of this camera will probably match D70 pricing. Rumour of a $1300 street price (not clear if this includes a lens). I don't see the 10D staying in production with a 20D floating around.

 

One can speculate that there is a huge gap between the 20D and the 1DII. There is still definately room in the product line for a 3D sporting a 8mp 1.3 crop sensor, 45pt AF, but with only a small buffer. . .That would be something 10D owners would drool over (and perhaps look deep into their wallets). . .wonder if that would be a September surprise?

[r_n]

Yup -- the photos were really there.

 

The joystick has little ridges on the NE/SE/SW/NW corners. That is a pretty good indication that it needs to point specifically in those directions, hence it looks safe to assume that the 9pt AF will indeed cover the rule of thirds hot corners. That is a very good thing.

 

I agree that the IS lens is disappointingly slow, but if the image quality is there it could be useful. I think I would stay with the 17-40 f4L however, with the 10-22 3.5-4.5 ef-s being the only ef-s lens I'd really need (77mm filter size). That lens looked big to me; the profile and size are akin to the 17-40 f4L! But from the picture it looked reasonably well-constructed.

 

Another note: the A95 was clearly pictured with 5 Megapixels. There is also an S70 which appeared to be the same as the S60 except done in charcoal gray and black (preferrable to the silver imho.)

 

The G6 still needs the adapter. I always found the need to attach an adapter to use a filter or hood unacceptable. Still do.

 

ONE MORE NOTE: The G6 was pictured with a flash -- a 550 EX flash. Since the 550EX was pictured, not a 550EX replacement, it looks like a 550 EX replacement is deferred. No new 550EX this time?

 

The fabled "3D" with the 1.3 crop chip from the 1DmkII will probably come out next year. Since the 1.3 + appears to be the pro / prosumer line for Canon now, I'd expect a 550EX pro flash update to be released either with the 1DsII or the "3D".

[bobatkins]

I believe that this information is real and that Canon simply made a mistake. After pulling the page they fogot to disable directory browsing of the image directory for press releases, but they've finally got around to doing that too.

<p>

I'm going to let this thread run since I think 20D speculation is now much more than an unfounded rumor. We now have a "semi-official" source (Canon's own webpage). Though it would be a great joke, I doubt that Canon are into "spoofing" people, so I have every confidence that the data that leaked out was accurate.

<p>

As soon as we have an OFFICIAL press release by Canon, I'll have an article on the new products up on photo.net. Keep an eye on the "what's new" column on the front page.

<p>

Interestingly enough, there was nothing on any of the Canon pages that suggested any sign of a 300D upgrade, an EOS-3D announcment or the widely rumored upgrade to the 1Ds. Make what you will of that.

<p>

I have a <a href="http://www.bobatkins.com/photography/digital/rumors.html">Canon rumors</a> page on my website if anyone wants to see the pics.

[phule]

<< I believe that this information is real and that Canon simply made a mistake. >>

 

I don't believe for a second it was a mistake. Personally I think Canon did that "oops" on purpose. Of course, I have no real proof of this, but it certainly wouldn't be the first time this has happened.

 

Anyone who has a NDA with them should be upset.

[keith_van_hulle1]

My only point is that the 17-85 IS lens looks so fake. Why is it that the 17-85 text appears as a completely different shade of white (bright) on the front ring compared to the much more grey appearing other text? Sticks out like a sore thumb an looks bogus.

[phule]

Keith Van Hulle,

 

The images were taken directly from the Canon website. There is no reason to believe that any of them were faked.

[jbq]

Jim Larson: "I have *no doubt* that this camera will produce better images than a 10D. I hear the "more noise due to smaller pixel argument", but I *refuse to believe* that this sensor is not more advanced than the current 6mp sensors and was specifically designed to maintain, if not improve, image noise."

 

The question is whether the same improvements packed into a new 6MP sensor would have resulted in a better image quality than jumping to 8MP (at a same process, the read logic takes about a constant surface independently of pixel size, i.e. larger pixels have a higher fill factor, which results in lower photon noise even after resampling). Also take into account the fact that the signal measured in large pixels is much stronger, i.e. less sensitive to pre-amplification noise, which makes a big difference at high sensitivities.

[markci]

Framkly, I think think Canon has a little more experience developing imaging chips than you do, Jean.

[tan]

Hey, for the price of a 10D with all these small but signficant improvements,

how can anyone possibly complain? It IS better than the 10D, is it not? The

faster 20D is very significant to me, as the 10D is painfully and embarrassingly

slow compared to the D70. The 8 megapixels with a Digic II processor is also

huge for those (like me) who regularly print to 13 x 19". And the ability to use

the new Canon EF-S lenses is great. All these and much more ... without any

price increase in the MRSP. I mean, common, get real ... you want a full-frame

sensor for $1,600? Yeah, it will happen ... eventually ... but not at this time.

[iskandar_azaman___kuala_lu]

I think this is not the camera for the 10D owners looking for an upgrade but is more of a camera for those looking to buy a new digital camera of around the same level as 10D but slightly improved.

 

Either way everyone wins. 10D goes down in price and you can buy it cheaper. Anyone looking for a new camera can buy the 20D.

 

However the lenses don't really look that interesting. Can someone explain to me whether it would be easier for canon to just make full frame sensors instead of designing new lenses? What are the cost trade offs?

[gary_g1]

Bobby Tan: I take it that 6.3MP doesn't cut it (quality/sharpness-wise) on 13"x19"? I just bought a Canon I9900 printer, waiting on the Canon 20d before buying a digital camera.

 

(I have done 8x10 prints from my wifes Canon A80, they're pretty awesome).

[michael focus97.com lee]

On the pixel count/dimension/noise front, it is not accurate to say that the pixels have to be smaller for the 8 MP sensor to still have the 1.6 CF. If you 'do the math', you'll get dimensions for the pixel size, based on 6 or 8 or whatever million pixels packed within a given area. Doing so for a 1.6 CF, and 6 MP, will give you the area of a pixel, assuming that the pixels are all touching each other. As far as I'm aware, there is spacing between pixels. Following this, doing the above calculation will not give you the pixel size because you don't know how big the pixel acutally is (i.e., how much space the pixel *actually* occupies within the space calculated.)

 

Depending on the spacing between pixels, Canon could pack 2 million more pixels into the same 1.6 CF by just reducing the spacing between adjacent pixels. And... as far as I'm aware, more photons will fall onto pixels, instead of falling in between them, and this will require less amplification (and so, less noise) to produce the image.

[eos 10 fan]

<a href="http://www.hunapu.dk/images/canons.jpg">Fronts</a> <br>

<a href="http://home.comcast.net/~craigsingle/d60_vs_10D_vs20D-back.jpg">Backs</a><p>--

[jbq]

Mark: understanding how something works doesn't require fist-hand experience doing so - we've invented languages to be able to convey knowledge. Going further, in some cases we can predict how something will work before anybody has any experience doing so (the predictions aren't always accurate). That's just basic science.

 

I'll gladly listen to any argument that explains to me why light can't be modeled as a stream of independent photons for the purposes of phtoography, or why smaller sensors can have a greater well depth or a greater fill factor than larger ones.

[jim_larson1]

BOB: Thanks for letting this thread run. This is afterall a *gear forum* -> and what could be more fun in a *gear forum* than talking about new gear?

 

Especially if we are all well behaved and keep it to only one or two threads. . .(If it was 20 threads. . .that would be too much!)

 

JEAN: I hear you regarding an *improved* 6mp sensor instead of an 8mp that merely matches current 6mp technology. From a marketing perspective, however, Canon *had* to bump up to 8mp => consider all the Sony 8mp digicam models floating around (including the new Canon Pro-1). If Canon did not bump up the new dSLR specs, then Canon would probably lose market share. Also, 8mp gives a big edge over the competing D70.

 

Not to mention, mere "noise reduction" in a 6mp camera would not encourage existing owners to upgrade, or perspective 300D buyers to pony up an extra few hunderd dollars.

 

I also agree with Bobby and Iskandar. This camera is better than a 10D. The difference is bigger than the difference between a D60 and a 10D. I suspect this camera would make alot of D60 owners upgrade. 10D owners (like myself) who bought more recently will likely skip this generation (or keep waiting for the holy grail of dSLR's: The 3D)

 

(who am I kidding: Where would I get $2K to buy a fancy new dSLR? I got too much in this hobby already!)

[dhiren]

It seems that EF-S lenses are here to stay for a while and so the 1.6 crop factor.

Canon recently upgraded their extension tubes to make it compatible with EF-S lenses, here

 

http://www.adorama.com/CAETEF252.html

 

so now it makes more sense prosumer camera + EF-S lens & professional camera + L lenses

 

i was hoping to see something in the range of 200-400, 400 f4 L IS ?

[jbq]

Jim - totally agree with you, there are plenty of very good marketing reasons why the upgrade to 8MP is totally justified. Improvements in sensor technology are likely to improve the noise quality over the previous generation, especially at low and high sensitivities where the noise is limited by the sensor itself. And in a competitive market it won't matter much if a camera has the lowest possible noise, as long as it has less noise than the competition.

[r_n]

For more oops! spec release see:

 

http://www.sdcard.be/Nederlands/store/viewitem.asp?idproduct=8979348

[suhaskulkarni]

Here is extract from specifications

 

QUOTE

 

 

File Size 1) Large/Fine: approx. 2.4MB (3072x2048 pixels)

2) Large/Normal: approx. 1.3MB (3072x2048 pixels)

3) Middle/Fine: approx. 1.4MB (2048x1360 pixels)

4) Middle/Normal: approx. 0.7MB (2048x1360 pixels)

5) Small/Fine: approx. 0.9MB (1536x1024 pixels)

6) Small/Normal: approx. 0.5MB (1536x1024 pixels)

7) RAW: approx. 8.2MB (3072x2048 pixels)

 

UNQUOTE

 

 

So.... where is 8 MP image ???

[bobatkins]

There's a bunch of errors in that "ad". Some places they call it the 10D, they claim 25fps (!) and say it has an EF mount (not EF-S).

 

It's clearly something they've made up and rushed togther without a lot of care - and they've not done a very good job at it.

 

My guess is that they're just trying to attract traffic, and as soon as the real press release comes out they'll post the real specs.

[jim_larson1]

I'm with Bob. This ad is not internally consistent, and therefore should be viewed with suspicion.

[r_n]

Yeah, upon further review it looks like they took the "known" specs to the 20D (more or less as posted at the head of this thread) and cut and pasted them into the previous 10D product description. Nothing new and some obvious mistakes (25 fps?!!! -- obviously meant to say 25 frame buffer -- but did not answer the big question -- what is the buffer size for RAW?)

 

My apologies: Sorry folks, false alarm.

[bob_katz1]

I can hear the pitter-patter of thousands of little lemming feet already...

[keith_van_hulle1]

And every one of them thinking "Finally, a piece of equipment to match my skills so I'll be able to take better pictures".

[peter_phan]

All the internal specs aside, I like what I see just on the outside. No more up/down button to navigate the LCD is a good thing. The addition of the joystick is a good thing. That taller pop-up flash is a good thing. The elimation of the seperate and space-consuming on/of switch for the Quick Control Dial is a good thing. But it looks like they removed the Assist button, too, which I liked having.