[Blue]

I wonder then, if we will see a system similar to the old auto110. a compact separate system with a small set of dedicated lenses and accessories. this would be wise in my opinion, as just releasing a body and one or two lenses, will not separate it enough from the current crop of micro 4/3 . . .

I wouldn't mind a digital body that could take the 110 lenses.

[philbaum]

One thing that seems clear to me is that based on the success of the K-7 and K-m, all Pentax DSLRs are going to be small and getting smaller from now on, regardless of format or function. This is my fearless prediction

Bingo, thats what i'm thinking. Everytime Pentax has gotten smaller, e.g. Km, k200, K7, now Kx, they've been rewarded by a lot of buzz, excitement, and sales.

Considering this trend, it seems reasonable to assume that HOYA is or has consideried the next SMALL step.

When i bought my first DSLR in 2007, i was impressed by the small size of the Oly 410. The thing that put me off on the 4/3's was the lack of 3rd party lenses. Also i remember the VF of the Pentax K10 as being better.

[*isteve]

Are you sure?

The 14-45 is about as good as it gets, optically, for a kit lens.

The 45-200 is kind of weak at the longer end of its focal length but surprisingly good at the short end.

The 14-140 I admit to not having read much about.

The 7-14 is impressing pretty much everyone who's used it so far.

Preliminary samples and impressions of the 20mm are promising (far moreso than the 17mm Oly, whose flaws were apparent pretty early), and the new 45mm macro... well, Leica saw fit to put their name on it, so I expect it to be pretty much as good as it can be.

The Oly m4/3 14-42 is pretty decent, too, considering that it's a kit lens.

Some people are too easily impressed.

Panasonic Lumix G Vario 14-45mm f/3.5-5.6 ASPH OIS - Review / Test Report

[er1kksen]

That's because it's designed to be digitally corrected, to allow for more compact design and less compromise on the sharpness. It's actually a 12mm at full-wide angle, to allow for those corrections in RAW processing.

Some people consider this "cheating." Hasselblad doesn't seem to think so. I personally consider it a perfectly viable and forward-thinking strategy.

So yes, color me impressed, though I'm not so sure that for them it was especially "easy." To put all those elements and innovations together must have taken at least a little work.

[thibs]

That's because it's designed to be digitally corrected, to allow for more compact design and less compromise on the sharpness. It's actually a 12mm at full-wide angle, to allow for those corrections in RAW processing.

That sucks if you ask me. Correcting things in camera (or in RAW developper) is OK but using that principle, let's manufacturer crap, the software will correct it anyway except that one soon realize that those very manufacturer will lock you in there own software. Conspiracy theory? maybe, maybe not.

[falconeye]

That sucks if you ask me.

Let me offer a different opinion.

I have awaited this development with much anticipation. I welcome the arrival of the first lens of a new generation, namely the above mentioned Panasonic zoom.

To me, camera + lens is a system which must be optimized at the system level.

If processor power has improved to the level that distortion and vignetting can be corrected in-camera in real time (or alternatively, in RAW converters upon import) then it is wise to make use of this possibility and drop optical correction of these parameters in favour of better correction of other aberrations, i.e., for better sharpness, bokeh, contrast etc. Or simply to make a lens wider.

It isn't a vendor lock-in because the correction is licensed to RAW converter manufacturers. In the case of Panasonic (upon pressure of Leica), the RAW format is licensed to converter manufacturers on the condition to include the correction. BTW, a possible vendor lock in came with the RAW format, not the correction. Adobe LR, ACR and (free) DNG converter includes the Panasonic correction module and you can't disable it.

Adobe is currently working at an extension of the DNG format to include the necessary correction meta data.

[*isteve]

Let me offer a different opinion.

I have awaited this development with much anticipation. I welcome the arrival of the first lens of a new generation, namely the above mentioned Panasonic zoom.

To me, camera + lens is a system which must be optimized at the system level.

If processor power has improved to the level that distortion and vignetting can be corrected in-camera in real time (or alternatively, in RAW converters upon import) then it is wise to make use of this possibility and drop optical correction of these parameters in favour of better correction of other aberrations, i.e., for better sharpness, bokeh, contrast etc. Or simply to make a lens wider.

It isn't a vendor lock-in because the correction is licensed to RAW converter manufacturers. In the case of Panasonic (upon pressure of Leica), the RAW format is licensed to converter manufacturers on the condition to include the correction. BTW, a possible vendor lock in came with the RAW format, not the correction. Adobe LR, ACR and (free) DNG converter includes the Panasonic correction module and you can't disable it.

Adobe is currently working at an extension of the DNG format to include the necessary correction meta data.

This ignores the obvious fact that there are pentalties involved in boosting shady edges and correcting distortion. As anyone would know from messing about in photoshop or a RAW converter, you cannot make a silk purse from a sows ear.

If you remove heavy distortion, you have to interpolate data = reduced resolution, in some cases in the corners by quite a large amount.
If you correct vignetting, you are brightening shadows = increased noise and again in the corners by over 2 ev, which is quite a lot of noise.

If I subsequently apply any PP to these images, I will simply magnify the problem. So, I have not removed any problem at all, I have just moved it from one place to another, and saved a few quid in the design of the lens, and made it impossible to manipulate the resulting image without degrading it to an unacceptable degree.

You cannot correct for a shit-poor lens performance using software, you can only make it less visible, but mucking around with raw data to mask lens inadequacies means you are simply starting your post processing with a
compromised image.

IMO, the lens should be as good as possible to start with and the correction should not need to be extreme. Since the K7 applies correction to DA lenses, perhaps we should now ignore their inadequacies? Not in my book.

I want to know how the lens performs independent of the system so I can pick the best one for my needs, and then I want my RAW to be RAW.

Panasonic are on the thin end of a nasty thick wedge, denying photographers the freedom to fully exploit RAW. I dont like RAW NR and I like RAW lens correction even less. If you want to do it in JPEG, then fine. Thats different.

I wouldn't touch a Panasonic micro camera with a long pole for precisely that reason. Sorry.

[falconeye]

This ignores the obvious fact that there are pentalties involved in boosting shady edges and correcting distortion. As anyone would know from messing about in photoshop or a RAW converter, you cannot make a silk purse from a sows ear.

If you remove heavy distortion, you have to interpolate data = reduced resolution, in some cases in the corners by quite a large amount.
If you correct vignetting, you are brightening shadows = increased noise and again in the corners by over 2 ev, which is quite a lot of noise.

Steve, I can understand your sentiments and value your opinion.

Two details need clarification:

Distortion correction interpolates data but this doesn't reduce resolution if the lens doesn't outresolve the sensor in corners.

Vignetting correction reduces dynamic range and freedom of noise in corners if you compare at a constant aperture. However, electronically corrected lenses can be made in larger apertures leading to better center dynamic range and freedom of noise. Or, they can be made using fewer lens elements improving resolution, contrast and luminance. Hell, even a f/1.0 lens could be made for the PK mount if electronically corrected ...

You need to keep constant the correct parameter. Like price or weight.

Overall, your thinking is old thinking (which isn't necessarily a bad thing). But the engineering task is to provide a best possible image, not a best possible pinhole lens emulation. To enforce an arbitrary constraint to stop optimization at some arbitrary boundary (the raw data in your case) is just poor engineering. Believe me, the professionals (lens designers) know it better and thank Panasonic to pave the way (and Leica actually).



As for the "build crap" first hand and "fix it in software" afterwards theory:
The recent discovery of extra-solar planets have only been possible by the advances of exactly this type of system-level optimization in (astro-)photography ...

[er1kksen]

Steve, I think the problem is in your thinking that Panasonic is offering "shit-poor" lenses with the intent of making it all look better in software. It's not a shit-poor lens, it's actually quite good optically in some respects. It just has a lot of distortion and a bit of vignetting, which could really detract from its performance, except that it's actually a 12mm fully wide-open, and processing through a RAW developer like RAWTherapee which doesn't do corrections will show you that. That gives leeway for optimizing corrections without real degradation of the image.

The end goal of any camera+lens pairing is to give you good images. All indications are that the Panasonic 14-45, used as intended, will give you images at least on par with your average kit lens. Given that it is your average kit lens, that extra bit of sharpness it has over most is all the better.

There are a lot of users with Panasonic m4/3 lenses who do heavy PP, and I haven't heard any complaints of the nature you mentioned about "magnifying the existing problems" so far.

Again, Hasselblad is one manufacturer I can think of right off the bat who is open about their commitment to pair lenses with correctional software for optimal results. Nikon also corrects a lot of things in software, though I don't know if distortions are included in that.

All the major RAW developing software companies will have access to the correctional data, so no big conspiracy there. I wouldn't be particularly surprised to see it show up in some freeware RAW developers, too.

You may dislike the approach, and certainly no one is forcing you to use it, but profaning it just seems a bit ridiculous. It's a perfectly viable product (heck, some might even call it "good") in a perfectly viable system, and while it may not fit everyone's tastes and ideals, the sooner we all admit that it's perfectly fine rather than grabbing the torches and pitchforks, the better.

If you want a shit-poor lens, go grab an A 35-70 f4-5.6. And I don't know if even that lens is worthy of profanity.

[*isteve]

Steve, I can understand your sentiments and value your opinion.

Two details need clarification:

Distortion correction interpolates data but this doesn't reduce resolution if the lens doesn't outresolve the sensor in corners.

Vignetting correction reduces dynamic range and freedom of noise in corners if you compare at a constant aperture. However, electronically corrected lenses can be made in larger apertures leading to better center dynamic range and freedom of noise. Or, they can be made using fewer lens elements improving resolution, contrast and luminance. Hell, even a f/1.0 lens could be made for the PK mount if electronically corrected ...

You need to keep constant the correct parameter. Like price or weight.

Overall, your thinking is old thinking (which isn't necessarily a bad thing). But the engineering task is to provide a best possible image, not a best possible pinhole lens emulation. To enforce an arbitrary constraint to stop optimization at some arbitrary boundary (the raw data in your case) is just poor engineering. Believe me, the professionals (lens designers) know it better and thank Panasonic to pave the way (and Leica actually).



As for the "build crap" first hand and "fix it in software" afterwards theory:
The recent discovery of extra-solar planets have only been possible by the advances of exactly this type of system-level optimization in (astro-)photography ...

Look I do understand the theory, but I think you are missing the point.
Unless you have a single objective in mind, you cannot "optimise" a system. Of course you can optimise the signal processing to resolve more stars if you build a lens to optimise resolution at the expense of aberation, but to remove such aberation accurately in software you need limited controlled conditions.

I can actually tolerate a certain amount of distortion in many shots, but light falloff is far less acceptable. Pulling up corners by 2ev i bad enough at ISO100, but an ISO800 shot is going to look pretty bad.

Nor am I trying to discover stars. Resolution is not the be-all and end-all of photographic achievement. For a general purpose system it is better to optimise each component of the system and perform the minimum software manipulation. It is also better to perform such manipulation manually because its extent (and relevance) will depend to a large degree on the circumstances. In some cases I may not want/need to remove it at all.

Besides camera companies are profit motivated, astro-photographers are quality motivated. Panasonics motivation is to build smaller, cheaper lighter lenses. If you get noisy corners and distortion corrected WHETHER YOU LIKE IT OR NOT then most people dont care. I do.

Photography is an artistic medium, not a scientific one. Panasonic may think they know what the "best" image is, but they dont. Thats why I shoot RAW. I can then apply corrections in the correct order to maximise the quality of the eventual output which will be different for each shot. Everyone knows its best to perform PP steps in a particular order, not have half of them done for you beforehand.

For general photography it is best to maximise the RAW input quality (lens and sensor) and then let the photographer decide what to do with it. Once RAW is no longer RAW, the lattitude allowed the photograher is reduced.

I didnt support NR in RAW and I sure as hell dont support image correction in RAW either.

[MikePerham]

I am with *isteve on this one!

[Angevinn]

I am with *isteve on this one!

Me too!

The lenses need to be designed to offer the best optical performance possible. Having to rely on the manufacturers specific software to process the images is for the birds.

thibs makes a good point also.

Personally speaking I use a Panasonic MP3 player when I run (it fits comfortably in the palm of my hand and has great sound). The only way I can put music on it is through the Panasonic software. I like Panasonic as a company but I don't want to get locked into software like this ever again. It's a pain!

[*isteve]

Me too!

The lenses need to be designed to offer the best optical performance possible. Having to rely on the manufacturers specific software to process the images is for the birds.

Well it seems that you can rely on both ACR or Silkypix to interpret the RAW corrections, but this is likely to lead to some inconsistency, not to mention restrict ones choice of RAW converter.

But in principle I agree with you. I dont want to match up a lens and a software package.

My real issue is this. The Olympus 14-40 4/3 lens on their SLRs is a stellar lens
Olympus Zuiko Digital 14-42mm f/3.5-5.6 ED

It shows NONE of the Panasonic lenses issues, yet is equally as sharp. I can make minor corrections without introducing any issues.

The problem on the micro 4/3 camera, is that the registration distance has shortened and therefore blown away the advantage of light hitting the sensor at near perpendicular angles. This was the raison d'etre of the 4/3 system, now lost.

So Panasonic are not doing everyone a favour by optimising lenses, they are working around a serious engineering challenge using a software fix.

Leica incidentally had to use offset microlenses on the M9 for the same reason.

AFAIK this is and will remain a major issue with micro cameras compared to their SLR stablemates. Its a shame because I really think this format has real promise, but I would rather they found a better solution to the problem.

Like I said, they can make any corrections they like in JPEG, but leave them out of RAW please.

iSteve

The statement of the day. In fact, I think it holds more than a real promise. They will figure it out soon enough. They most always do.

[NickC]

Me too!

The lenses need to be designed to offer the best optical performance possible. Having to rely on the manufacturers specific software to process the images is for the birds.

It's all a compromise otherwise you wouldn't have more than one lens covering the same focal length. If all lenses had to be the best optically then the kit lens would be the 16-50, with the price and weight penalties involved. If the 18-55 can be made to perform the same by using software correction I'm all for it.

The limiteds are also compromises - they don't offer the maximum possible optical performance or the fastest apertures, but are smaller and more portable because of it. If some of the IQ can be fixed in software then that's good. The alternative would be much bigfer heavier lenses, which I'd see as against their philosophy.