Can Olympus compete in the pixel wars?

[DarkPenguin]

I believe Oly needs to create a new class of camera--take existing E-3 sensor and processing engine and repackage in a small (510 size would good) electronic viewfinder (WYSIWYG) based body that uses the 4/3rds interchangable lens format--this could become the DSLR killer. 
[a href=\"index.php?act=findpost&pid=175221\"][{POST_SNAPBACK}][/a]

If used in RAW mode how would this be different for the end user from the E-510?  Other than having an a viewfinder that lags the action?

[Atlasman]

If used in RAW mode how would this be different for the end user from the E-510?  Other than having an a viewfinder that lags the action?
[a href=\"index.php?act=findpost&pid=175271\"][{POST_SNAPBACK}][/a]

Lag can be fixed--and the keyword in my statement is "WYSIWYG". Also: no mirror-slap = SILENCE.

[DarkPenguin]

Ah.  A quiet shutter is the requested feature.  That would be nice.  The E-3's AF would be nice in a weenier bodied camera, too.

[Er1kksen]

Not to mention that they could make the evf as large as they wish, and high-resolution, too, which could actually be an improvement over the tiny ovfs of their consumer models (never really bothered me, but a common complaint).

However, I think they would need a mirror assembly for the af setup, unless they want to do it p+s style? However... since there's no need to send light to the prism, more light could be diverted to the AF sensors, resulting in greater low-light sensitivity and precision.

[panoak]

I can't imagine why Oly should care about the pixel wars, when the 1D Mk.III has exactly the same res.  As previously mentioned, to those of us who know what megapixels are, 10 is plenty.  That will stand for exactly the same reason why 35mm film stood for so long: it fits the standard magazine format to the end of human perception.  There is a huge matter at play here that doesn't get enough coverage.  The higher res. sensors are going beyond the ability of lens technology to carry that higher res.  On this very site, M. Reichmann falls out of love on the 1Ds Mk.II with the 100~400L, because aberrations become apparent @ 16 Mp.
     Where Oly does need to concentrate is on improving things like DR, buffer size, cont. frame rate and so forth, but they are fine right now, at the price point.  If I was a new E-3 owner,  (not my brand) I'd be satisfied and confident that my camera will be competent for a long while to come.

[Er1kksen]

It'd be nice if everyone realized that 10mp is quite adequate as you do, but the reason Olympus has to care to some degree is marketing. Olympus is already very far behind in marketing, and if all their competitors are marketing 14mp cameras and they only have 10mp, they will run into some problems.

[Ray]

No, no, no! These incremental jumps from 6 to 8, from 8 to 10, from 10 to 12mp are trivial. Added together, however, they mean something. The difference between 6mp and 12mp is definite.

The next FF DSLRs are going to be 24mp. That's a significant jump from say 12mp and it's an area where Olympus will fear to tread.

[Er1kksen]

Well, personally, I use Olympus, the 7.5 mp E-330, and I don't want to upgrade untill I can get a nice 4000x3000 12mp image... otherwise I see little reason to upgrade.

[BJL]

The difference between 6mp and 12mp is definite.

The next FF DSLRs are going to be 24mp.
[a href=\"index.php?act=findpost&pid=176992\"][{POST_SNAPBACK}][/a]

But each successive doubling of pixel count is less important, or important to a smaller range of of photographic and print-making situations.
- From 1MP to 2MP made even snapshot sized prints noticeably better, and so was of value to anyone who printed at all, or had a high enough resolution monitor.
- From 5MP to 10MP shows up only with far larger prints and not on any current monitors. Starting at about 8"x10" or above, and then only with fairly close viewing and some types of subject matter.
- From 10MP to 20MP might only be relevant with the rather extreme combinations of:
-- prints viewed from unusually close; far closer than the typical viewing distance of roughly the maximum dimension of the print, and
-- sharp, high quality lenses: only fairly good primes and the best of zooms.
Because with most lenses, the extra pixels will mostly add greater awareness of lens aberrations and not much more image detail (like 14 bit A/D convertors mostly revealing finer details about the sensor noise, unless the sensor has very good DR.)

And with the near inevitable move of all DSLR systems to offering 14MP and above over the next year or two, the visible gap between 14MP and and higher pixel will become even more rarely seen in prints.  Near inevitable because of the empirical

Law of the Pixel Count Race: Every new high of pixel count in digicams is soon matched or exceeded by every DSLR system.

(Even if this is at some point driven mostly by marketing considerations.)

And Sony just announced a digicam with 14MP, 1/1.7" format sensor.

[Ray]

But each successive doubling of pixel count is less important, or important to a smaller range of of photographic and print-making situations.
[a href=\"index.php?act=findpost&pid=177335\"][{POST_SNAPBACK}][/a]

Put simply, there's a law of diminishing returns that applies when sensor resolution is improved without simultaneously improving the lenses used. That's true and it is a concern. Every time Canon releases an upgrade with a higher pixel count, there are those who worry if the potential improvement in resolution will be realised.

However, having compared my 20D with my 5D using the Canon 100-400/5.6 IS zoom at 400mm from the same position, I can see greater detail and resolution in the 20D shot, to an extent which would cause me to choose a more expensive lens in preference to another cheaper lens if I saw such a difference during a comparison in a store whilst comparing lenses before buying.

Since the 20D has the same pixel density as the 1Ds3, and since the Canon 100-400 zoom is a medium quality lens in the general scheme of things, I have no doubt that a 24mp FF 35mm sensor would produce a worthwhile jump in resolution and detail compared with my 5D, with most good lenses, even if they weren't the best.

This is of course a dilemma for me. A 16mp upgrade to the 5D is likely to be too marginal for me, considering only the resolution factor. A subsequent upgrade again from 16mp to 24mp would also be too marginal. But an upgrade from 12mp to 24mp is likely to be irresistable. I have a wide format printer, you know!  

[BJL]

... I can see greater detail and resolution in the 20D shot ...

Since the 20D has the same pixel density as the 1Ds3 ... I have no doubt that a 24mp FF 35mm sensor would produce a worthwhile jump in resolution and detail compared with my 5D, with most good lenses, even if they weren't the best.[a href=\"index.php?act=findpost&pid=177412\"][{POST_SNAPBACK}][/a]

Agreed with respect to the central 40% of the 24x36mm frame that you are seeing in files from the 20D. I am not so sure about the remaining 60% around the edges when scrutinized in 24MP 24x36mm format files.

Still, I am sure that you are one of those who is willing and able to upgrade lenses as needed to get beyond the resolution limits that you are apparently suffering with the mere 12.7MP of the 5D!

The market for smaller formats like 4/3, EF-S and DX is not the market for "highest feasable resolution", but a different and historically always far larger SLR mainstream. In that sense, the short answer to the question that started this thread is "no, and none of 4/3, EF-S or DX puts a priority on trying match the pixel counts that larger formats like 24x36mm offer". (Though with the D200 etc. and then the D300 etc., Nikon and Sony were happy to come close to the slightly low-ball choice of 12.7MP in the 5D, probably taking many sales that would otherwise have been split between the 5D and 20D/30D.)

[jani]

- From 5MP to 10MP shows up only with far larger prints and not on any current monitors.

Well, that's not entirely accurate.

The monitor in the link has a resolution of 2048x2560, and it's a monochrome display. Given the resolution loss from monochrome conversion of Bayer pattern sensors, it can display more resolution than what you can get from a 5 MP or 6 MP camera.

If you look around, you can get a used 9 MP (3840x2400, WQUXGA) colour display, under the IBM, Iiyama, ViewSonic, ADTX or IDTech brands. None of these displays are currently manufactured, though.

Toshiba will, apparently, begin production of a WQUXGA display this spring, so the format ain't dead yet.

And even if you can't view an image in full size on 16:10 or 4:3 displays, there is always that infernal cropping of images to suit the subject. Darnit.

Nitpicking aside, I don't think people will view images up close and personal on a computer screen today, but I don't think we should pretend that they won't in the future.

IF the subject matter survives having a high resolution image, I don't see any particular technical reason why one shouldn't desire to take advantage of that. Who knows if your pictures might not sell in 2020 as well as in 1995.