Reading mode: Light Dark

12mp: beginning of the end of mp race?

  • Thread starter Thread starter mschf
  • Start date Start date
Go to last post
M

mschf

Guest
I find it interesting that one camera maker finally came out of the closet to make a statement about megapixels:

http://news.cnet.com/8301-13580_3-10189546-39.html?part=rss&subj=news&tag=2547-1_3-0-5

Personally I hope that other manufacturers are like-minded and have just been waiting for a competitor to step up first and declare its intent to stop pursuing the "more pixels" path. Yeah ok, I've no doubt it's also a bit of marketing backtracking by Olympus, maybe they're acknowledging in a roundabout way that 4/3's limitation is about 12mp anyways, but I'm hoping that APS-C oriented brands too are looking at 12mp-14mp being the breakpoint. If this will prove to be a industry-wide trend, I look forward to technology finally being applied to optimize the image, especially in areas such as DR.

Somehow though, I doubt it.
--
~ Martin
http://www.martinpolanic.com
 
Perhaps, with the LX3 they realized that discerning Photographers are after Image Quality no MP.

Olympus don't make sensors so they have no saying about the MP race.
Is Panasonic.

Fuji once paused for a moment but they are stupid and continued in the non sense MP race.
 
--

Olympus may not make their own sensors, but they can still have a say in the mp race. They can tell Panasonic what they want in a sensor. With Canon making their own, and Nikon in bed with Sony, Panasonic can not afford to ignore Olympus. Or Olympus can make a deal with Fuji or Sigma.
 
We should never say "never" when it comes to any technology, and there's no reason to suppose that in the future there will be 20+MP sensors (in 4/3, Nikon DX etc size) that are just as good as today's 10-12 MP sensors.

That said - it is a matter of diminishing returns, and where you put your R&D priorities.

A 12MP sensor will produce prints in the 9-14 inch range (depending on aspect ratio) at "native" 300 dpi resolution, and significantly larger without any special software or expertise by the user (and only marginal loss in IQ).

Which would satisfy the needs of - well, as Olymplus says, just about anyone and for those who don't, they'll be after full frame anyway.

So maybe, just maybe, the market lunacy will take a pause and the camera makers will turn their attention to other things for a while.

If you want me to pay more for my next camera, I'll be wanting something other than just an extra few megapixels, thanks.

I don't think there was anything wrong with taking the race up to 12MP, though.

--

 
There really is no 'megapixel race'. There has been the technology to produce DSLR's with hundreds of megapixels for years (just think, if you make a DSLR size sensor with P&S sized pixels that's what you get), however the increase in pixel counts has been quite sedate, about 25% per 18-24 month generation. This is despite the fact that in image quality terms, increasing pixel count is entirely beneficial. In particular, at any particular image size, it reduces both aliasing and AA filter detail loss. The reason that pixel count doesn't increase in the way it should is that, of course, higher pixel counts cost storage space and speed, so pixel count is essentially dictated by Moore's law in the last resort. Even Moore's law would suggest that pixel count should double each generation, so in fact the camera manufacturers are somewhat lagging on what they should be doing. I suspect the extra capacity is going into supporting lower prices and more profits by using slightly less highly specified components each generation. For example, the DIGIC IV does not have twice the capacity of the DIGICIII, though Moore would suggest that it should.
mschf said:
I find it interesting that one camera maker finally came out of the
closet to make a statement about megapixels:

http://news.cnet.com/8301-13580_3-10189546-39.html?part=rss&subj=news&tag=2547-1_3-0-5

Personally I hope that other manufacturers are like-minded and have
just been waiting for a competitor to step up first and declare its
intent to stop pursuing the "more pixels" path. Yeah ok, I've no
doubt it's also a bit of marketing backtracking by Olympus, maybe
they're acknowledging in a roundabout way that 4/3's limitation is
about 12mp anyways, but I'm hoping that APS-C oriented brands too are
looking at 12mp-14mp being the breakpoint. If this will prove to be a
industry-wide trend, I look forward to technology finally being
applied to optimize the image, especially in areas such as DR.

Somehow though, I doubt it.
--
~ Martin
http://www.martinpolanic.com
Click to expand...
--
Bob

 
The 12 mpix for a 4/3 sensor corresponds to 18 mpix for an APS-C sensor. I guess that means that those of us who use APS-C sensor cameras can expect at least some additional increase.

Hopefully, we will soon see improvements in sensor design. I am looking forward to the 30-50 mpix cameras. It is nice to be able to do some serious cropping when needed and to also be able to print large sizes. Hopefully the improved sensor design will also give us corresponding improvements in IQ at high ISO settings.

I hate it when someone tells us what we need. We already "need" 20 mpix. Even Costco prints at the 330 dpi level - that corresponds to 20 mpix for an 11x14 print without an cropping.
 
Leon Wittwer said:
Olympus had said it when they were not in deep financial troubles.
Now, it seems like an attempt to justify the decision to go to the
smaller sensor.
Click to expand...
From the article:
Leon Wittwer said:
"We don't think 20 megapixels is necessary for everybody. If a customer wants more than 12 megapixels, he should go to the full-frame models," Watanabe said.
Click to expand...
I'd stay this is a statement of target audience, as in "We cater to these specific customers. People who have other needs have other options." I see a lot of self-confidence here. Whether that is genuine or just marketing I can't tell.

Concentrating on a specific product area and doing it well is a valid business strategy. It worked great for some companies and destroyed others. Which will be the case for Olympus, only time will tell.

Cheers,
AdSR

--
'I don't need a better camera, I need better imagination.'
 
Arrowman said:
We should never say "never" when it comes to any technology, and
there's no reason to suppose that in the future there will be 20+MP
sensors (in 4/3, Nikon DX etc size) that are just as good as today's
10-12 MP sensors.
Click to expand...
In business, "never" is never "never". I'd say what they meant was "in the foreseeable future". How long is that in modern technology business? 2-3 years maybe?

I pretty much agree with the rest of your post.

Cheers,
AdSR

--
'I don't need a better camera, I need better imagination.'
 
Am I right in thinking this is not just about number of pixel but quality also? I am reminded of when I buy my first hi-fi in 1970's power of amplifiers was ever being increased, but distortion was itself not necessarily being improved at all so extra power was in a sense useless.

I am not familiar with science of sensors but is there a similar situation?

--
Veejay Chatterjee
 
I thought about that too, as far as electronic hardware goes, it makes sense that camera manufacturers would want to maintain a pixel-to-performance ratio. But what about optics? Do you think new designs can be engineered in the not too distant future that will be able to cope? I know there's been a lot of research on this over the past few years but nothing's come of it to date (granted, this takes time). Perhaps the barrier has more to do with optics but I wonder how much the manufacturers are holding back on pixel quantity because of this.
--
~ Martin
http://www.martinpolanic.com
 
veejay chatterjee said:
Am I right in thinking this is not just about number of pixel but
quality also? I am reminded of when I buy my first hi-fi in 1970's
power of amplifiers was ever being increased, but distortion was
itself not necessarily being improved at all so extra power was in a
sense useless.
Click to expand...
Well, it would be if there was any evidence that smaller pixels are 'lower quality' than big ones. The science (taken overall) says there should not be a difference until you get to very small scales. The practice seems to suggest that smaller pixels perform, if anything, a bit better than bigger ones, but that's probably because smaller pixels also tend to be newer designs. In the case of hi-fi, distortion was being improved, or at least total harmonic distortion was. It turned out that some types of distortion were more important than others, and also that some 'audiophiles' thought that things sounded better with lashings of second harmonic distortion. The very most expensive (and therefor highly rated) amplifiers are often not the best performers, on any objective basis. If digital photography goes the same way, we'll end up being sold $10000 digital Holgas.
veejay chatterjee said:
I am not familiar with science of sensors but is there a similar
situation?
Click to expand...
No

--
Bob

 
That's almost as silly an idea as the one that was attributed to an IBM executive in the 70s: "no one will ever need more than 640kB of memory in a computer".

Pixel counts will keep expanding as long as they can figure out ways to do it. The chips will get faster. The future is video. All cameras will capture video, then you'll sort through the images to find the one perfect still you were trying to get. The camera will be buffered- it will capture images for 5 seconds before you push the shutter button, so you'll never miss that critical moment in the action.

Cameras will have DOF image stacking, HDR processing, and 3D model construction built in and it will be fast. You'll be able to alk around an object with the camera and transfer the image data to a 3D printer and print out a scale model of the object in minutes.

Now you know.
 
that is a wet dream we ought not to have or else we will be forever disappointed.
geoson said:
Or Olympus can make a deal with
Fuji.
Click to expand...
...but in order to further torment myself, wouldn't it be nice if there was a fuji superccd in one of those m4/3 retro bodies of Oly's... oh dear, the anguish...

--
http://www.flickr.com/photos/gearporn/
-equipment in profile-
 
the end of photography as we know it. how sad...
mark_rehorst said:
That's almost as silly an idea as the one that was attributed to an
IBM executive in the 70s: "no one will ever need more than 640kB of
memory in a computer".

Pixel counts will keep expanding as long as they can figure out ways
to do it. The chips will get faster. The future is video. All
cameras will capture video, then you'll sort through the images to
find the one perfect still you were trying to get. The camera will
be buffered- it will capture images for 5 seconds before you push the
shutter button, so you'll never miss that critical moment in the
action.

Cameras will have DOF image stacking, HDR processing, and 3D model
construction built in and it will be fast. You'll be able to alk
around an object with the camera and transfer the image data to a 3D
printer and print out a scale model of the object in minutes.

Now you know.
Click to expand...
wish i didn't though...

--
http://www.flickr.com/photos/gearporn/
-equipment in profile-
 
bobn2 said:
There really is no 'megapixel race'.
Click to expand...
I'd say there was a race of sorts but more like racewalking than running, where everyone agrees to go slower than their maximum pace. If there is one thing business has learned it is to make incremental improvements. If you produce the best widget you can immediately you will get increased sales until the others catch up. But after that you are stuck as further improvements are much more difficult for everyone. Although I'm no MBA, I'd guess that incremental improvements increase profits for everyone...

The one area that may have slowed pixel counts is yield. If the yield is poor then a tiny sensor is still profitable since you are only thowing away a small piece of silicon. For a larger sensor the yield has to be greater for it to be as profitable...

I would like to see a 48MP Bayer APS-C sensor which actually only produces 12MP output. They could then eliminate the AA filter since four pixels are combined into one saved RGB pixel. There is a speed issue with so many pixels, though, which might make the camera more expensive on the hardware side (A/D conversion) but the software burden is reduced as complicated Bayer processing algorithms are not needed...

Cheers, Keith
http://www.kotay.net/keith/photo/photo.shtml
 
You know, for FF cameras that pixel density would mean 48Mp.
There is still a lot of room left...
mschf said:
I find it interesting that one camera maker finally came out of the
closet to make a statement about megapixels:

http://news.cnet.com/8301-13580_3-10189546-39.html?part=rss&subj=news&tag=2547-1_3-0-5

Personally I hope that other manufacturers are like-minded and have
just been waiting for a competitor to step up first and declare its
intent to stop pursuing the "more pixels" path. Yeah ok, I've no
doubt it's also a bit of marketing backtracking by Olympus, maybe
they're acknowledging in a roundabout way that 4/3's limitation is
about 12mp anyways, but I'm hoping that APS-C oriented brands too are
looking at 12mp-14mp being the breakpoint. If this will prove to be a
industry-wide trend, I look forward to technology finally being
applied to optimize the image, especially in areas such as DR.

Somehow though, I doubt it.
--
~ Martin
http://www.martinpolanic.com
Click to expand...
--
http://www.pbase.com/akimage
 
In the early days of stereo high fidelity, the primary goal was adequate frequency response, especially at the lower frequencies (where the human ear has a problem). That, by the way was the reason the "loudness control" knob was added to the preamplifiers so that the low frequency would allow a better response by the ear at lower volume levels.

There lies the rub. It takes a significant increase in power (amp wattage) to start pushing low frequency out at a desireable level. It's noteworthy that when observing a low frequence bass note displayed on an oscilloscope, the "rise time" of the note has a high frequency component. A hard lick on a base drum initially shows a need of high frequency reaction time.

How this applies to sensor design, I have no earthly idea. A little humor there, albeit, very little.

derFritz
 
You must log in or register to reply here.

Keyboard shortcuts

f
Forum
n
Next
w
Next unread
m
My threads
Mobile site
About
Editorial content
Cameras & Lenses
Community
All content, design, and layout are Copyright © 1998–2025 Digital Photography Review All Rights Reserved.
Reproduction in whole or part in any form or medium without specific written permission is prohibited.
When you use DPReview links to buy products, the site may earn a commission.
©GPS Media - Guides, Products, Services.
Back
Top