The 5D replacement rumor mill...

[jonoslack]

Hi Peter

1) Removable sensor - this is nothing else than wishful thinking. Of course, theoretically it could be done, but actually no manufacturer of FF DSLRs would really like to offer. And mainly because such a design would be very expensive (see Leica's DMR solution) and not allow the camera manufacturers to make the desired profit on future models.

It's already been done AND on full frame as well - when Kodak brought out the SLR/n, they offered a sensor upgrade from the previous model (14n). It was actually about 1/3 original cost of the camera.

Everybody said that if the sensor could be improved . . . they should give it away for free:ROTFL::ROTFL:

I would think that anyone else looking at the possibilities would take one look at that experience and run in the other direction.

 

[ptomsu]

Hi Peter



It's already been done AND on full frame as well - when Kodak brought out the SLR/n, they offered a sensor upgrade from the previous model (14n). It was actually about 1/3 original cost of the camera.

Everybody said that if the sensor could be improved . . . they should give it away for free:ROTFL::ROTFL:

I would think that anyone else looking at the possibilities would take one look at that experience and run in the other direction.

Well, of course it can be done, issues are:

1) it cannot be for free - otherwise manufacturers would make no more revenue but rather loss on these models

2) it s pretty obvious that a sensor change will imply more data to be handled (more pixels, more bit depth, etc - at least for the next years of evolution), so al the electronics for image processing would also need to be changed. And this means even more expensive. Can actually only be done in the way as it is done in MF digi backs, where you change sensor and image processing hardware together in the same back. Broken down to a rather small DSLR compared to MF would result in heavy and bulky DSLRs.

 

[jonoslack]

Well, of course it can be done, issues are:

1) it cannot be for free - otherwise manufacturers would make no more revenue but rather loss on these models

2) it s pretty obvious that a sensor change will imply more data to be handled (more pixels, more bit depth, etc - at least for the next years of evolution), so al the electronics for image processing would also need to be changed. And this means even more expensive. Can actually only be done in the way as it is done in MF digi backs, where you change sensor and image processing hardware together in the same back. Broken down to a rather small DSLR compared to MF would result in heavy and bulky DSLRs.

It WAS done with the Kodak - and yes, they had a module which included the sensor and much of the electronics, it certainly didn't result in it being heavy and bulky. It wasn't popular though, and I doubt if anyone would care to repeat the experiment.

 

[woodyspedden]

Jono

I fundamentally agree with all you have said except................The performance of the original SLR/N was frought with problems. Color was pretty terrible, noise was unacceptable etc. However it was the first FF sensor so there was a pretty loyal following.

Kodak, I believe priced the upgrade so as not to lose money and to pacify their customer base. Not only did they include a new sensor but super low ISO (6 if I recall) and a number of other features and fixes that made it a pretty darned good camera, at least for me.

I would not think this represents any kind of model for a normal upgrade. Since an upgrade cannibalizes the sale of a new camera a substantial profit would have to be made on the upgrade.

Woody

 

[jonoslack]

Hi Woody

Jono

I fundamentally agree with all you have said except................The performance of the original SLR/N was frought with problems. Color was pretty terrible, noise was unacceptable etc. However it was the first FF sensor so there was a pretty loyal following.

They were interesting times - I did a lot of Beta testing for Kodak of the firmware for both cameras - and of the RAW converter (was it called photodesk?)- hours and hours doing spreadsheets of magenta and green values across the screen (very similar to the corner problems with the M8).

The colour on the 14n was completely fixed before the upgrade - and many felt that the upgraded camera wasn't quite as sharp as the original.

Kodak, I believe priced the upgrade so as not to lose money and to pacify their customer base. Not only did they include a new sensor but super low ISO (6 if I recall) and a number of other features and fixes that made it a pretty darned good camera, at least for me.

The ISO 6 option was available on the original camera, (with a firmware update only).
The upgrade certainly didn't pacify anyone :ROTFL: it was met by complete fury.

Still, it was always intended that the camera would be upgradeable - and the internal design made it reasonably easy.

I would not think this represents any kind of model for a normal upgrade. Since an upgrade cannibalizes the sale of a new camera a substantial profit would have to be made on the upgrade.

Woody

Whether it's a good model or not, I think it was intentional, but it unquestionably backfired, and on that basis it certainly wasn't a good model!

I agree - by the end it was a great and underrated camera - I often wish I hadn't sold mine.

 

[woodyspedden]

Hi Woody


They were interesting times - I did a lot of Beta testing for Kodak of the firmware for both cameras - and of the RAW converter (was it called photodesk?)- hours and hours doing spreadsheets of magenta and green values across the screen (very similar to the corner problems with the M8).

The colour on the 14n was completely fixed before the upgrade - and many felt that the upgraded camera wasn't quite as sharp as the original.




The ISO 6 option was available on the original camera, (with a firmware update only).
The upgrade certainly didn't pacify anyone :ROTFL: it was met by complete fury.

Still, it was always intended that the camera would be upgradeable - and the internal design made it reasonably easy.



Whether it's a good model or not, I think it was intentional, but it unquestionably backfired, and on that basis it certainly wasn't a good model!

I agree - by the end it was a great and underrated camera - I often wish I hadn't sold mine.

Sometimes wish I had not sold mine as well and yes, the software was called Photodesk

Best

Woody

 

[asabet]

It's already been done AND on full frame as well - when Kodak brought out the SLR/n, they offered a sensor upgrade from the previous model (14n). It was actually about 1/3 original cost of the camera.

That reminds me; didn't Kodak have a camera with a user-replaceable AA filter at some point?

 

[Jack]

Well time to place your bets...

The 40D replacement, the 50D, is now official so the 5d replacement cannot be far behind. Now here is what is interesting: the 50D has 15MP (4.7u pixels), new Digic IV, 6 frames per second and expanded ISO 6400 and 12,500...

My money says that if Canon can do all that on the crop sensor, it bodes very well indeed for the specs on the 5D replacement

 

[asabet]

There is a Japanese poster in the DPReview forums named dHitman who predicted the 18-200 IS back in April (though he said it would be bundled with the 450D), and judging from his posts may also have had a bit of advanced word on the 50D. He says the 5D replacement is a 21MP camera. Not outside the realm of possibility IMO.

 

[Guy Mancuso]

Yes but they need to add some more benefits over there own 50d so this may turn out to be a interesting camera the 5d replacement. We can count on the new Digic IV for sure and most likely at least 5 FPS and my bet a 18mpx or 21mpx FF sensor. But i would like to see a very weak AA filter or i won't even look at it.

 

[Ben Rubinstein]

There is a Japanese poster in the DPReview forums named dHitman who predicted the 18-200 IS back in April (though he said it would be bundled with the 450D), and judging from his posts may also have had a bit of advanced word on the 50D. He says the 5D replacement is a 21MP camera. Not outside the realm of possibility IMO.

Not at all actually. Would make the lack of pro AF almost worth it!

 

[Terry]

Well time to place your bets...

The 40D replacement, the 50D, is now official so the 5d replacement cannot be far behind. Now here is what is interesting: the 50D has 15MP (4.7u pixels), new Digic IV, 6 frames per second and expanded ISO 6400 and 12,500...

My money says that if Canon can do all that on the crop sensor, it bodes very well indeed for the specs on the 5D replacement

The first night I had my D700 and was shooting at a party, there was someone there from Canon Imaging that insisted that the new 5D would blow away the D700. It will be interesting to see what they come up with. I'm happy with my choice but camera improvement keep moving the bar higher and that is a good thing.

 

[robmac]

A 21 MP 5DII, I can see it now - forums filled with cries of 'soft' images from folks slapping their xxxx L zoom on their new baby and wondering where the hell the resolution went.

 

[asabet]

21MP in a 5D replacement would make a lot of sense. No pro AF or weather sealing to compete with the 1Ds III. Not enough frames per second to compete with the 1D III or its replacement (which will likely be a lower MP full frame model). It would also compete well with the D700 without the need for pro AF or weather sealing.

 

[LJL]

A 21 MP 5DII, I can see it now - forums filled with cries of 'soft' images from folks slapping their xxxx L zoom on their new baby and wondering where the hell the resolution went.

Maybe, or maybe not. If some of the new tech, like the redesigned microlenses and Digic IV processing works out, Canon may be able to reduce the AA filter and not lose so much detail at the higher pixel density of smaller size. There are laws of physics about some things, but I am not convinced we have hit the wall on all of them yet. What if they were able to unleash a 21MP camera with FF sensor that did deliver detail without "mushiness" at high ISOs? That seems doable, but we will have to wait and see. The images it and others produce will tell the story, not the marketing hype

LJ

 

[robmac]

If they skinny-out the AA filter, etc, etc it might work. What I am assuming is that we will see the current tech as per the 1Ds3. Given that C and N seem emotionally attached to AA filters, my expectations aren't high.

So far the 50D doesn't look like they've reacted (or had time to react) as many would have hoped (e.g. radically) to the 'new' Nikon.

I do think they need to be a tad careful releasing a vastly new & improved 21MP sensor at a pro-sumer price to soon after the 1Ds3. I don't think a lot of natives would take that well ;> This is of course assuming it's a 21MP sensor in the first place...

That said, ya never know. And, as you say, it will take months before we know if the various packages work as advertised - and within what limits.

 

[Jack]

Canon makes the 1 series for pros, and most working pros already own the latest 1's. While the market is not saturated, it is likely 90% of the folks who were ever going to buy a new one already have. Thus, the delayed release of the new 5 makes sense; fill the pipeline with the runner-up...

However, Canon arrogance finally bit them in the butt --- the mistake they made was waiting too long, as they have clearly lost a significant share of the market to Nikon, a share they will have a tough time recovering unless the new 5D is something spectacular in terms of ISO, AA filter and improved IQ...

Personally I waited so am not affected, and will go with the new 5D regardless simply because the lenses both mount and manual-focus in the correct direction. In 25 years of shooting Nikon, I *never* really overcame those two issues... My dyslexic right-brain I guess.

PS: My bet is on 18MP or thereabouts, being less than 20 for marketing reasons against the 1DS3. The practical difference in 18 and 22 on full-frame DSLR is going to be insignificant since few lenses are good enough anyway...

,

 

[jonoslack]

PS: My bet is on 18MP or thereabouts, being less than 20 for marketing reasons against the 1DS3. The practical difference in 18 and 22 on full-frame DSLR is going to be insignificant since few lenses are good enough anyway...

,

I guess you're right.
I find it fascinating that Canon set so much store of making sure that a camera doesn't compete with it's bigger brother (no weather sealing on the 50D for instance).

Nikon on the other hand really do seem to bring out the best camera they can for the money. I think that's probably where they have done so well in the last year.

The focusing issue is interesting - I hear what you say, but maybe us brits are so used to having to swap everything around everytime we drive in another country - it seems to me like very much the same sort of transition.

 

[Ben Rubinstein]

Of course with that much resolution you run into serious diffraction issues. Might be worth keeping at least one of my original 5D's just for when I have to stop down past f11!

 

[LJL]

Again, this may or may not be the case. With the past and existing sensors and how they were configured with different microlenses, we could see diffraction limits kicking in, especially with overaggressive AA filters. If Canon, and this is a big if, went to a newer sensor design (like placing the electronics beneath the photosites rather than inside as most CMOS sensors are built), and if the microlenses are reconfigured to better match the photosites on the sensors, it may be very possible to reduce some of those diffraction issues the kick in hard on some smaller pixel sensors now.

Until we start to see actual results, it may be premature to dismiss what this newer camera may or may not be able to do. I am not being a fanboy or anything like that here, as that is not my nature by a long shot. I just think that folks may want to hold judgment and decisions a bit longer until they actually see what some of the newer tech really is able to deliver. It may not amount to anything different, or it could be a very significant improvement that starts to rewrite some of our "rules" about how capable some things may be. My first inclination is to agree that diffraction could become a player, but I am not convinced it will be until we start to see actual shots. Even on the MF side of things where they have a lot more sensor real estate to work with, pixels are getting smaller, and IQ is not getting worse.

LJ