Comparing lens resolutions

[ajohnw]

I'm the original poster of the thread that pjbw was advised to take a look at and, since he came and woke us all up from our post-new-year snooze in the dpreview-Sony Cybershot forum , I thought I'd follow him back here and see what you folks have been saying.

As the thread there makes plain, the samples of output from my brother's example of the marque did not do the R1 justice as I tried to compare it with the RX10. They did, however, illustrate the overruling weakness of the R1 [I've owned 2 of them in recent years and I bought the one my brother still has, so I know the camera very well] the viewing screen and the EVF are rubbish: the screen is the size of a large postage stamp and its resolution is far too low and the EVF is equally poor. What's worse, the focus point indicators on the screen often tell lies - not deliberately, they just cannot show you everything that they embrace because the screen resolution isn't good enough. These completely inadequate ways of framing pictures lead to a high incidence of incorrectly focussed images with a wisp of shrubbery or other unseen interloping object pin sharp against a blurred version of the intended subject.

Before anyone leaps in and alleges poor technique: probably! But manual focus with screen magnification, and moveable spot focusing, are far to clumsy to capture even slow-moving children and most of us granddads are blessed with the standard-issue, quick-moving variety anyway. To show that I can use an R1, and that I know how good they are, look at this image...

http://www.theformsonline.com/pics/16.jpg

To illustrate how clumsy the R1 is at capturing scenes like that, with objects in more planes than a wall has, I have to tell you that I kept only 4 of the 12 images I made as my granddaughter cleaned her grandma's stone circle on the first sunny day of spring 2013. In all of the rest, the wisp of berberis visible in the left front of the image linked above was crystal sharp against a blurred little girl - granddad's have no use for blurred images of their nine-year-olds. The image was shot with fill flash and is everything a good photograph should be [no artistic comments please ] but it's tricky to get shots like that without a substantial admixture of discards.

To get to the point of all this preamble: look at this image...

http://www.imaging-resource.com/came.../YDSC00821.HTM

As I said, I know the R1 inside out and I loved both of the examples I owned, but an R1 is simply incapable of the IQ of that shot.

My advice to pjbw is twofold. If you were only looking at the RX10 because it might be the update of the R1 that you'd perhaps like to have but you are satisfied with your results, why change?

If, on the other hand, you'd like a modern screen and EVF and a distinct performance hike, get yourself down to your local camera store and try the RX10. It is superior to the R1 in every measurable way and is a joy to use.

It does landscapes...

http://www.theformsonline.com/pics/01.jpg

...and livestock comes and poses for it...

http://www.theformsonline.com/pics/02.jpg

Ed Form

Your shots on dpreview got off to a bad start as any sort of comparison isn't really possible. Most of the R1 shot is out of focus. Why - because both cameras are set around F5 but the r10 has a smaller sensor so the depth of field is greater. Find a portion of the R1 image that is in focus and I would be inclined to say it has better gradation.

The ladies face with pores is roughly twice life size and shot with a focal length of 161mm. Doesn't really signify anything unless a similar twice life size image shot is taken with the R1. As it doesn't have the same zoom range this may mean putting the camera closer to the subject.

The link already posted does a better job of comparing the cameras but goes a little bit silly on the telephoto shots as both cameras are set to a focal length of about 70mms. It also mentions that raw noise performance at the R1's highest setting is better than the R10's. The R1 suffers at 3200 because of it's internal jpg software. Neither building shot is of much use as a comparison but if the R10's shot is reduced to the same size as the R1's there is little or no difference.

The landscape shot you posted isn't good enough to tell what the R10 can do in that respect. It's hard to tell where the camera is focused and if there was sufficient depth of field. On the other hand this could just be the usual small sensor problem - detail is only just there so results are mushy full sized and fine reduced for web galleries etc.

The livestock shot is odd. Highlights are blown in places but the detail in the 2 heads isn't reflected in the bodies. I regard that as a small sensor jpg processing software effect but in this case could also be down to off axis performance of the optics as well. I've noticed that some how or the other a lot of bridge cameras tend to capture subject detail fairly well but backgrounds not nearly so well.

Landscapes can have problems. Take this one. Small things aren't clear as there is insufficient resolution. More pixels might help but the lens may not be up to it. It's gone some way towards the one you have posted. These are straight camera jpg's with all enhancements set off. These 2 are a way of judging a cameras performance on landscapes.

http://www.23hq.com/ajohnw/photo/8155737/original

Change focal length and small detail becomes clear.

http://www.23hq.com/ajohnw/photo/8155738/original

Nothing to do with R1's and R10's as both were shot with a cheaper camera but they do illustrate why I feel that the R10's landscape shots provided so fare don't really prove anything. Actually at 28mm full frame equivalent I would be surprised if any camera / lens combination would do a significantly better job of the 1st shot but notice that nothing is noticeably fuzzy at full size.

John
-

[edform]

I have to say that every one of the images posted looks terrible showing a combination of or individually - focus errors, camera shake due to incorrect shutter speed selection, massively blown highlights, over processing (common on small sensor cameras) poor WB and I think even a de-centred lens.

That's odd since there are three different cameras involved.

Perhaps you can be more specific and give us an individual critique of each of the shots in turn so that we can all learn by my mistakes and those of the pro who took the portrait of the lady.

The portrait, incidentally is an out-of-camera jpeg with no processing at all and if it's as bad as you say, post something of yours that's better by way of illustration.

Perhaps you can bear in mind when appraising the shots that the only other shot entered in this thread, the landscape posted by the originator of the thread, was delivered at a tiny file size and has been processed to give a painterly effect. My landscape was loaded into DxO as a RAW after which I lifted the shadows a click or two because the area under the trees on the opposite bank of the lake was actually in deep shadow. It also had the horizon corrected [about 2 degrees]. That's the full extent of the processing. The shot was taken at 1/640th and f4 so if it shows camera shake I must have bashed the camera with a hammer to make it happen. There are no blown highlights and the sky really was that free of clouds - in fact the wisps that can be seen low down in the sky were not visible to the naked eye because the light was harsh enough to shut one's pupils down to pinpricks.

Then if you read my post again, you might notice that I went into detail about the big problem with the R1, which is that it is really, really hard to see what you've focused on in a complex scene because the screen and EVF are hopeless. The intended subject of 16.jpg is soft but reasonably detailed at 100% and one click of the mouse wheel beyond that the pixels become visible. I hadn't noticed before, but the camera actually focussed on the little green bush in the bottom right corner with the intended subject in the DOF. Who cares? It's an example of beautiful lighting on a beloved subject and every time I look at it, it fills me with happiness. Everything a good photograph should be and do. It was shot at 1/500th and f8, so again, no camera shake.

As to your diagnosis that the lens is decentred: nonsense! This is the same camera a few weeks later...

http://www.theformsonline.com/pics/03.jpg

This was shot at a low shutter speed, with very low, late-evening sun coming from front left of the view - so all of the harsh, razor-sharp shadows behind each of the tiny little stones scattered on the cast iron manhole cover must be evidence of camera shake and lens decentring.

Finding any of the faults you list, in the picture of the two bantam cocks is astonishingly difficult for an image you describe as terrible. It's derived from an RX10 RAW file dropped into DxO with as-installed settings - in other words, very light processing with no hand work at all. So it is not over-processed, I can't see any blown highlights, detail in the feathers on the back and wings is fine as is the fluffy down between the neck feathers, and this 100% crop of one of the birds' heads shows little evidence of camera shake even though it was 1/160th at f4...

http://www.theformsonline.com/pics/04.jpg

It's a little soft this big, but the picture is somewhere around A3 at this magnification. I'm at a loss as to where you're coming from.

Top of my head, I'd guess your a M43 man, although most of that fraternity post constructive comments rather than unsupported, and largely unsupportable, throwaway bile.

Ed Form

[edform]

Again I definitely agree.However, in spite of its vastly superior paper specification and huge array of bells and whistles I did not think it was worth dipping into a savings account for. However I remain to be persuaded...

If you're happy with your shots and have evolved a system of working that bypasses the small-screen focusing difficulties, why change? The R1 is quite a camera.

The real appeal of the RX10 for me, the thing that made it worth the £899 I paid, was the longer maximum reach and the fact that with the VCL-DEH17R 1.7x extender lens that reach goes out to 340mm at f2.8 and IQ remains top notch. I'm currently drawing an adapter plate to take the bayonet from the R1 adaptor but fitted to the shape and dimensions of an RX10 - should be in the CNC man's workshop next week. I'll publish the details when I'm sure it's right.

Ed Form

[edform]

The ladies face with pores is roughly twice life size and shot with a focal length of 161mm. Doesn't really signify anything unless a similar twice life size image shot is taken with the R1. As it doesn't have the same zoom range this may mean putting the camera closer to the subject.

The perspective would change with a closer approach, but relative detail handling would still be down to the lens/sensor performance alone, so a comparison shot is perfectly feasible. Forget framing to look the same, just shoot to make a head the same size; I don't believe an R1 can record an image with that kind of detail.

The landscape shot you posted isn't good enough to tell what the R10 can do in that respect. It's hard to tell where the camera is focused and if there was sufficient depth of field. On the other hand this could just be the usual small sensor problem - detail is only just there so results are mushy full sized and fine reduced for web galleries etc.

You went on to post a pair of comparison images to illustrate that zooming in brings more distant objects into recognisable clarity, but didn't seem to allow for the fact that the image I posted has a total back to front distance of almost half a mile. Your examples show a building about 100-150 yards from the camera. I didn't attempt to record the detail on the opposite bank of Diss Mere; the shot was taken at 33mm equivalent focal length and 1/640th at f4. Focus was general centre-weighted out-of-the-box with the camera horizontal and at eye height. I zoomed in from 24 to 33mm to place the ducks just where I wanted them in the scene. The shot was meant to portray the big, tree-bounded, watery, haven of peace that lies just off the main street of a bustling market town. If I'd wanted to let you see the detail of the buildings on the far bank I had a 6x zoom range available to pull them towards you. What's more, I didn't offer the shot to show that the RX10 is better at landscapes than the R1 - read what I said again: 'It does landscapes' - in other words, don't worry, you can get just as nice landscapes with an RX10 as you currently get with an R1.

And surely the flippant wording '...and livestock comes and poses for it...' was sufficient clue that getting 6 feet away from a pair of birds that live wild at the side of the mere when they ought to be in someone's protective custody is a rare thing and a good case for taking a frame or four. As I believe I said in one of the posts in the dpreview forum, I'm unskilled in Photoshop and don't really trust myself to do a good enough job of processing images. Nevertheless, to show that the virtually untouched distillation to jpeg from Raw via DxO's default settings isn't all that the images contain, here is a new version from the same RAW file so that you can see that the feathers are packed with detail and blown highlights are not a significant issue.

http://www.theformsonline.com/pics/05.jpg

If you're interested enough to try processing some of the RAWs themselves, to see how good they are in better hands than mine, I'll happily post a bunch and let you have the links either here or in PM.

Ed Form

[ajohnw]

If you're happy with your shots and have evolved a system of working that bypasses the small-screen focusing difficulties, why change? The R1 is quite a camera.

Ed Form

Can you elaborate on that. Robin is a Nikon man by the way. I am primarily into M 4/3 which is why I feel you should elaborate. Like the Sony these cameras are mirror less and manual focusing needs a magnified view of sufficient magnification to achieve a near pixel level match between the sensor and the electronic viewfinder/screen. For instance my E-M5 has a 1.4mp viewfinder and a 16mp sensor. For really critical manual focusing that needs a 10x view at least. 7x can be used but there is a good chance 100% res views will not be sharp. This means that AF is the best option in many situations and various styles of indicator in the electronic view show where the camera has decided to focus. The decided aspect is bad news on most cameras as given full reign it may focus on any old point it decides to what ever style of camera it is but there is usually a way of guiding it.

So in what way does the RX10 help?

John
-

[Black Pearl]

I'm a Nikon and a Pentax man to be exact though what that has to do with anything I've no idea and I have even less of an idea as to how me posting an image taken with my gear would help with a discussion about a compact with a small sensor.....unless you're hoping to pick faults with my camera equipment/technique in a hope it will strengthen your argument.

As to the images taken on the equipment being discussed:

I have to say that every one of the images posted looks terrible showing a combination of or individually - focus errors, camera shake due to incorrect shutter speed selection, massively blown highlights, over processing (common on small sensor cameras) poor WB and I think even a de-centred lens.

Focus Errors. The camera has entirely failed to get a correct focus on the girl, who I'm assuming is the main subject of the shot. I feel that blaming a poor EVF/Rear LCD is not a suitable reason when the photographer could have simply make sure the camera was using a fixed (generally centralised) focus point to ensure this didn't happen.

WB. I personally feel this shot is too warm, probably down to the camera seeing the ambient colour temp but not correcting, or being manually corrected for the flash firing.



De-Centering. I'll concede that this may or may not be the case here but the top left corner has a very unusual look to it that could be down to a poorly centred optical path.



Over Processing. Common on most if not all compacts with a small sensor is an overly processed look to the fine detail particularly as the distances increases. While I'm the first to argue that images shouldn't be viewed at high magnifications if you do in this case the detail is very poor with a mushy lack of any fine detail. I would also go as far as to suggest either the camera has failed to focus correctly or the photographer has failed to use the correct focus mode to ensure suitable focus for the scene.



Camera Shake. I'm not convinced this shot is as sharp as it could be due to camera movement and that is before I get onto excessing noise - typical of small sensors - or the blown highlights.

If I were joining a forum to defend my choice of camera or to point out how superior I felt is was over the one in discussion I would make sure the shots I posted were of the very highest standard. I personally don't feel these are particularly good examples and they certainly wouldn't make me change from an R1 with its larger sensor.

[edform]

Can you elaborate on that. Robin is a Nikon man by the way. I am primarily into M 4/3 which is why I feel you should elaborate. Like the Sony these cameras are mirror less and manual focusing needs a magnified view of sufficient magnification to achieve a near pixel level match between the sensor and the electronic viewfinder/screen. For instance my E-M5 has a 1.4mp viewfinder and a 16mp sensor. For really critical manual focusing that needs a 10x view at least. 7x can be used but there is a good chance 100% res views will not be sharp. This means that AF is the best option in many situations and various styles of indicator in the electronic view show where the camera has decided to focus. The decided aspect is bad news on most cameras as given full reign it may focus on any old point it decides to what ever style of camera it is but there is usually a way of guiding it.

So in what way does the RX10 help?

My remarks were made in the context of R1 versus RX10 and highlighted the real advantage the RX10 has over its older ancestor - it's EVF and its rear screen are both more than good enough to let you see which object or plane in the frame has been chosen as the autofocus target. I haven't discussed critical focus or achievement of full value for every single pixel at all; I've only pointed out that in autofocus mode on an R1 it will persistently and infuriatingly choose the wrong target and you will not see the mistake because the postage stamp, low resolution screen and the tiny, low resolution EVF are simply incapable of showing you what has happened. As you indicate that you are a critical focus man while I am a point and shoot believer, you would probably be at home with the R1, using its display magnification and manual focus to maximize the value of every pixel. I'd loose the spontaneity and the 1000 nuances of my favourite subject - my granddaughter - because a soft capture of something, an expression or action, is to me, infinitely superior to any slowly taken and perfect portrait of her.

The RX10's viewing apparatus is so good that I see false focus choices and correct them, usually by simply moving the camera a few degrees till it snaps onto what I want and then half-pressing the shutter release to lock it and moving back to the framing I chose- a split second's work. You can do this with an R1 as well, but you can't see that you need to. You'd see it with your critical focus approach; I haven't got time or my shot is gone.

As to the M43 throwaway line: that was insulting to M43 users, to whom I apologise. I've only come to photography forums since I got my RX10 - a short time - and it's been interesting to see that M43 users are the stoutest and most numerous defenders of any single system's place in the photo-sphere. Inevitably this means that people who defend by throwing acid are more frequent in the M43 brigade also - it's simply a numbers thing. On the other hand, the most helpful and best informed posters are also most commonly from the M43 infantry.

I don't have any illusions that my R1s were, or that my RX10 is, better than the best that M43 can do now, although the R1 certainly was when it first appeared and for some years afterwards. I chose the RX10 because I won't have a camera that doesn't have a lens or lens-set of top-grade Zeiss equivalent quality and in M43 I simply cannot afford to buy gear that good.

Although I'm a long-term R1 user I didn't buy it for it's fixed lens either; I bought it for its peerless lens. I'd have happily lugged an SLR and a kit of lenses about but I only had the original list price of the R1 to spend and no outfit of SLR plus lenses could even approach the performance of the RI for less than 3 times that money. So I bought an R1 on eBay for £172. I then added both of the big, really heavy extender lenses with their bayonet adapter, and a whole bunch of other kit to my big LowePro case and lugged the whole lot about everywhere.

When Sony announced the RX10 I realised that it was the R1 updated with solutions to my main gripe and probably offering great performance, but I also looked long and hard at an EM1 with the two F2.8 lenses that straddle the range of the RX10. I would have bought it in preference to the RX10 had I been able to afford the extra £1500, but I couldn't, so no M43 for me. The decision to buy the R1's descendant was based on the same argument as made me buy the R1s - performance for money that nothing even comes close to.

What does intrigue me is the closer ties being forged between Sony and Olympus. Although the EM1 gets top grading, the sensor in it is a blind alley as far as development is concerned - absolutely the best that can possibly be squeezed from that sensor family but having no on-going development path without moving to a new sensor generation that Olympus don't have. I'd guess you guys must be intrigued by the possibilities for good and bad that such an alliance could offer.

Ed Form

[Polar01]

This thread has some interest to it, however I am confused as the working title of the thread is "Comparing Lens Resolution" instead it is more on one camera model compared to a newer model, with no or less talk about lens.

Just my opinion for what's it worth.

[Black Pearl]

I don't have any illusions that my R1s were, or that my RX10 is, better than the best that M43 can do now, although the R1 certainly was when it first appeared and for some years afterwards. I chose the RX10 because I won't have a camera that doesn't have a lens or lens-set of top-grade Zeiss equivalent quality and in M43 I simply cannot afford to buy gear that good.

Before we go any further would you please clarify that statement.

Are you saying your RX10 is better than most M4/3 offerings because it has a Zeiss lens? You do realise that you don't actually own a Zeiss lens - you have a lens made in Japan by Sony (albeit with some Zeiss input) on a production line that has been signed off by Zeiss. It is not in any way, shape or form a genuine Zeiss lens.
Zeiss make their glass in Germany and their lenses are hand assembled in their German factory.

I'm not knocking it as a decent bit of glass but to say it alone offers more than a camera system that is vastly superior in ever way is a bit lop sided. Are you saying we should all rush out and buy a Nokia phone to replace our m4/3 systems or even our APS-C systems because it has a (faux) Zeiss lens on the front?
As much as I admire Sony for jamming 20mp into a 1" sensor and getting anything nearing a decent output out of it I would never choose a tiny sensor with limited dynamic range and a tiny usable ISO range over a sensor with larger photo sites and a choice on almost unlimited glass - including real Zeiss optics.

[edform]

I'm a Nikon and a Pentax man to be exact though what that has to do with anything I've no idea and I have even less of an idea as to how me posting an image taken with my gear would help with a discussion about a compact with a small sensor.....unless you're hoping to pick faults with my camera equipment/technique in a hope it will strengthen your argument.

At the primary level the discussion is actually whether a 2013 generation sensor of 116square millimetres can keep up with a 2.67 times larger sensor dating from 2005. The portrait image posted by the Pro from imaging-resource.com shows that it can. My 'post an image to compare' line was merely a repetition of a request that R1 users do that, a perfectly reasonable suggestion. I have criticised no one's images here, or in the dpreview forum where the hook to this topic found me, and I certainly don't intend to start by criticising yours. It would just be good to have your swingeing criticisms backed up with an example of a similar image showing what really good is compared to 'terrible'.

Just to set the matter straight, I looked and found a suitable candidate image in the 2005 imaging-resource review of the R1. here it is...

http://www.imaging-resource.com/PROD...1OUTFACAP2.HTM

As you can see, the 2 images we can now compare are framed the same, although the one I've just found is better lit while the RX10 example was shot against strong sunlight and is somewhat moody by comparison. Nevertheless, the lower resolution and narrower gamut of the R1 is absolutely unmistakeable. The RX10 has twice the pixels and a greater dynamic range - read the reviews - in situations where the light is good it will do a better job than the R1 can. There's no rocket science here, you just have to open your eyes.

With the small sample of reactions to you that I've been able to see so far, I have a strong idea that you are actually pretty good, so an example of a hard lighting portrait from one of your obviously superior camera/lens combinations would probably be an object lesson for many of us. Had you been civil in your first post and rather than grotesquely unpleasant, this might have turned into a pleasant learning experience for me and loads of folks like me.

Anyway, onwards...

As to the images taken on the equipment being discussed:





Focus Errors. The camera has entirely failed to get a correct focus on the girl, who I'm assuming is the main subject of the shot. I feel that blaming a poor EVF/Rear LCD is not a suitable reason when the photographer could have simply make sure the camera was using a fixed (generally centralised) focus point to ensure this didn't happen.

I don't recall blaming the poor display system of the R1 for my errors. I criticised the poor display system because it makes the errors really easy to miss. In this specific incidence, the image is actually the fourth of a sequence of 12 and the first three were all scrap because the only sharp object in each of them was one or other of the wisps of berberis [the red shrub] visible in the front left of the scene. The focus was set to use centre-weighted standard auto mode. For this particular image I moved round so that the berberis was of to the side of the view and assumed the clear line of sight to my granddaughter meant she would be the focus target - the focus frame did actually halt on her face. When I pressed the button, however, the focus system grabbed the little green box plant at the extreme bottom right and the combination of the main subject being almost in the depth of field and the EVF being incapable of showing even substantial softness produced a false focus image. The focus target movement with no corresponding movement of the green target box on screen was almost certainly a result of the slow, early-generation processor of the R1 but the RX10 would have shown me the target dance. I still value the shot for it's strong colours and the memory of the conversation Aimee and I were having at the time - there was a strong element of cheekiness involved; something about how decrepit and old I am.

WB. I personally feel this shot is too warm, probably down to the camera seeing the ambient colour temp but not correcting, or being manually corrected for the flash firing.

Look at the pink and white stripes on the T-shirt: the white is snow white and the pink is pale and very well controlled. Look at the white chalk marks that Aimee is making in the cement lines between the York-stone slabs they are grey white without a colour cast, and Aimee's hair is a rich golden brown lighting up in sunlight just as you see.

We're back with the portrait from imaging-resource.com again and your analysis of it couldn't get any further away from the facts. There was no flash used in this shot as a close look at the eyes ought to have told you. There are clear sharp reflections of something behind and to the left of the camera in both of her pupils which fill flash would have drowned with flare-reflections. The strong light coming from the right over the subject's shoulder and blowing out the highlight on the tip of her nose is Nashville sunshine. The girl is a fairly well known country singer on the Nashville circuit and she is wearing stage make up with a tanning foundation and heavy applications of rouge on her cheeks. If you look at her septum, however, [the groove under her nose running down to the top lip], you will see that the tanning foundation has sweated off and her skin is pale, even sallow, with a scatter of blackheads - I said in another post that she can't have been very keen on having this portrait published in all its gory detail. Now look at her eyes again and you will see that she has very deep green pupils and that the colour of her outer eye is cold and grey with no trace of warmth at all. Now move on to the area at the top right, which you said was strange and suggested lens decentring. First of all, its *very* sharply in focus, odd for a lens with optical alignment problems. Second, the pointing is cement made with coarse sand and grit and both colour reproduction and detail retention is outstanding. There isn't a trace of lens trouble in the photograph. It is sharp on a 20 inch monitor when expanded to 100% with the width of her cheeks almost exactly filling the screen [Professional quality monitor with a video card that cost twice as much as the camera - second hand] and it remains printable up to half as large again. Some of the detail is too a very high standard indeed - check the eyes: the white blaze at the left hand side of both pupils is a reflection of a white painted circular metal staircase and the buildings that form the perimeter of the yard she's standing in are also clearly visible.

Your criticisms of this photograph have no foundation in reality and plainly reflect an unreasoning disdain for 'small sensors' that can't possibly perform. Letting your contempt cloud your judgement and fog your eyes and parading that misguided attitude in a public forum has exposed a side of you that I don't really want to see any more of. I may well have posted imperfect images, but this particular image is stunning by almost any measure. You just couldn't bring yourself to look.

I've already added comments on the other images I posted in another post, so I won't bother going further in this particular exchange.

To other members of this forum - I really don't like being goaded into an exhibition of bad temper by nasty, unreasonable and unreasoning trolls - it gives a poor impression of me as well. My apologies! I'll unsubscribe now so that you don't have to put up with any more of it.

Ed Form

[ajohnw]

Well I do hope I haven't upset Ed. My main points made were intended to help. The original OP mainly. I do feel there is a need to clear a couple of things up.

Firstly all cameras with electronic views. The viewfinder usually has the most pixels but given the number of them in it and on the sensor there is little chance of reliably achieving even postage card sized sharp pc screen images just using the viewfinder at 1x magnification. Like many people who bought Pen's I bought a number of manual lenses and adapters for it. Fun trying say a 500mm lens even with only a 5x magnified view but still rather hit and miss. 5x always is. That's why the R1 reviewer states that they feel it should be higher. Actually this argument has cropped up before else where so some one was asked to do it with a moderate telephoto. They found they couldn't even at postcard size. Others have tried as well. Image stability while the shot is being taken can be surprisingly good and EM-5's allow IS while using manual lenses but long lenses are still difficult to handle.

It seems that the EM-1 uses Sony's latest style sensor. Phase contrast on the sensor itself. Canon have developed something similar. According to a Jessop's sales person Nikon have switched to Sony sensors as well for guess what - phase detection on the sensor itself. If it can be done the change makes a lot of sense. The EM-5 appears to use a Sony sensor. Olympus sold off all Panasonic sensored Pen's very cheaply before it and the new Pen range came out. Not sure what Panasonic are using. The link up between Sony and Olympus appears to be a bit one way. Sony gets the sensor IS and Olympus's jpg software and access to Olympus's optical design facilities.. The 2 companies will also use the same distribution network and share camera components. I'd guess Olympus also get a secure sensor source and some input into what they want. I have a feeling they are currently field testing Sony's latest offerings as well. Sony once let another company down but at least they supplied free replacement sensors for a very long time after the camera was defunct - Dimage. Sony wanting Olympus optical design. Sounds strange but they make some of the best optics on the planet. At a bit of a price at times when they really go to town.

Pixel counts on m 4/3? Silicon has a low but real co efficient of thermal expansion. This can tend to break things over time as the parts on it get smaller and smaller. Bigger the sensor the bigger the expansion and contraction. They get round this problem in some semi conductor areas by gradually strengthening things up as they get nearer the edge. I wouldn't want to even guess what the limits are at any sensor size. The biggest factor really is cost. Lots of parts are made on a wafer in one go. More parts = lower costs.

In my view Nikon have done the sensible thing with 1in sensors as used in the RX10. The Nikon 1 range with it's relatively low zoom range rather high res lenses. Trouble is that the V1 doesn't offer the controls many would want. The V2 may but the costs goes up. Olympus have done that with Pens too. In my view the E-PL1 was also crippled in firmware.

Personally I feel that the original poster should wait as there are clearly some changes coming but from what I have seen he may well be unhappy in one way or the other with all 1in sensor offerings except perhaps the V2 and needs would best be met by something with interchangeable lenses and a bigger sensor. On the other hand as he is currently using a wide zoom range lens he might be happy with a single zoom lens rather than 2. When camera's mention Zeiss Planar etc Germany the name really does mean something - usually. Costs are a problem and there are no free lunches associated with optics. The name Zeiss used in any other context is nothing more than reassuring. The main aim in design will have been cost so it isn't very sensible to compare with famous Zeiss prime lenses. Yeh I buy compacts with Zeiss lenses as well. Came as a surprise when they had to improve the optics to cope with HD video. It most certainly looks that way to me as I have owned 3, 2 pre HD. Great cloud exposures most of the time but oh the dark end. Not much is recoverable either.

John
-

[xpatUSA]

Well I do hope I haven't upset Ed. My main points made were intended to help. The original OP mainly. I do feel there is a need to clear a couple of things up.

Pixel counts on m 4/3? Silicon has a low but real co efficient of thermal expansion.

2.6x10^-6 per deg C change. Therefore, a 5um pixel changes it's width by 0.013nm per deg C compared to say green light at ~550nm. 0.022nm for a D700 with it's 8.45um pixels.

This can tend to break things over time as the parts on it get smaller and smaller. Bigger the sensor the bigger the expansion and contraction. They get round this problem in some semi conductor areas by gradually strengthening things up as they get nearer the edge. I wouldn't want to even guess what the limits are at any sensor size.

Certainly an interesting solution, John - is there a reference for that?

[ajohnw]

Sorry Ted no. I only know due to wondering why automotive parts sometimes take some time to get through qualification so I asked what they do if parts can't take temperature ranges of -40+85C in this case. That's ambient temperature. Any self generated heat has to be added on top as far as the part is concerned. The bits that do the work get hotter than the cases as well.

0.013nm doesn't sound much does it. This may be of interest

http://en.wikipedia.org/wiki/10_nanometres

http://en.wikipedia.org/wiki/32_nanometer

This chip uses 1.1um pixels and is probably a complete camera on a chip. Pixels are one thing and the wiring and logic are another

https://www.aptina.com/products/image_sensors/ar1230/

They also do some "broadcast" quality video stuff. "Only 10mp in a 1in sensor" Being me I can't help wondering why.

http://www.ukiva.org/news/news-item/191

Last time I looked at Canon DSLR hacking they still made use of Motorola micro controllers. I was dealing with them when the reps were over the moon because a Japanese company had chosen them. I believe the rest is a sensor and asic but maybe some have that on the sensor as well. Motorola semiconductors is now 2 separate companies On Semiconductors and Freescale. I often wonder if these are the people who actually developed Canon's sensors.

John
-

[pjbw]

Personally I feel that the original poster should wait as there are clearly some changes coming but from what I have seen he may well be unhappy in one way or the other with all 1in sensor offerings except perhaps the V2 and needs would best be met by something with interchangeable lenses and a bigger sensor. On the other hand as he is currently using a wide zoom range lens he might be happy with a single zoom lens rather than 2.
John -

I am fascinated to see how my original little post has generated so much interest. It all goes to show the completely fluid state of the digital camera 'market' - pity the poor marketing people in the manufacturing companies!
I am now firmly of the (humble) opinion that I will hold on to my R1 until I can see reasonably definitely which way the bridge camera market is going.
Could I ask for some help?
A photo taken with my R1 has received a lot of praise. I am sure people would be interested in what can be achieved with an 8 year old camera. Unfortunately the JPEG does not shrink and compress well. It needs to be seen at 3600 x 2400 pixels and 3.1+ MP. Can anyone suggest a hosting website which I could upload it to and which could give me a URL to it which I could paste into a post here? (Like Ed Form's).
Peter

[ajohnw]

I use this one. They allow a very reasonable number of uploads per month without any charge. Adds are not a problem and it's easy to link to a shot like this one. Just click on the image after it's been uploaded to your page and it will open reduced size to fit the browser. Clicked again and it goes full size. They stipulate that the site mustn't be used to sell things. Signing up is easy and having done so it hasn't cause me any problems at all.

http://backup.cambridgeincolour.com/...98480/original

The Dpreview on the R1 is interesting. The reviewer gained the impression that the camera was well over priced and then considered what it would cost to buy a DSLR with a similar quality zoom on it.

Just a note. I posted that one because some one in another site thought that the just had to get rid of the "kit lens" on an E-PL1. PP is a bit OTT but the lens doesn't need it really. Actually I am getting more and more convinced the lens is better than the movie optimised 12-50mm I bought with the OMD-EM5. Ideally I need to replace it with the new 12-40mm

John
-

[xpatUSA]

Could I ask for some help?
A photo taken with my R1 has received a lot of praise. I am sure people would be interested in what can be achieved with an 8 year old camera. Unfortunately the JPEG does not shrink and compress well.

Why is that, may I ask?

It needs to be seen at 3600 x 2400 pixels and 3.1+ MP. Can anyone suggest a hosting website which I could upload it to and which could give me a URL to it which I could paste into a post here? (Like Ed Form's).

I have a satellite internet connection which can sink to as low as 3Kb/sec. Therefore, even though your undoubtedly excellent image "needs" to been seen at max size and minimum compression, I for one get quite tired of waiting minutes for a masterpiece to download.

Anyway, my rant notwithstanding, Google has about 15Gb waiting for you for free, in the form of what they call 'Drive' if you don't mind being suckered into their social networking thing.

[ajohnw]

Not being sucked in to anything is why I use 23hq Ted. It's Europe based as well. They don't intend to "suck" anybody in.

John
-

[pjbw]

I use this one. They allow a very reasonable number of uploads per month without any charge. Adds are not a problem and it's easy to link to a shot like this one. Just click on the image after it's been uploaded to your page and it will open reduced size to fit the browser. Clicked again and it goes full size. They stipulate that the site mustn't be used to sell things. Signing up is easy and having done so it hasn't cause me any problems at all.

The Dpreview on the R1 is interesting. The reviewer gained the impression that the camera was well over priced and then considered what it would cost to buy a DSLR with a similar quality zoom on it.
John
-

John,
This is one of my efforts with the R1:

http://www.23hq.com/pjbw/photo/15099119/original

titled 'Winter Evening' (swan pasted in where one did fly in as I was setting up the R1). Unfortunately right-clicking on 23's picture only shows a very truncated Exif.
It was taken with Adobe RGB (1998) JPEG, much better than sRGB or SR2 RAW in the R1. It is post-processed in Corel's Paint Shop Pro 9.02 (JASC's 9.0 had a nasty bug). Luckily the camera, PSP, my NEC monitors and Photobox all match so I needed a minimum of post-processing; save as TIFF, crop, paste, sharpen, resize and then optimize and save as JPEG for printing.
Back in the naughties when PSP 9 had the bug I joined the herd and bought Photoshop CS2. This was certainly the worst bit if Windows software I had ever used (XP). Very luckily I ordered 2 CF cards from Lexar and each came with a mini CD with PSP 9.02. I have used this ever since. 9.02 does not do very much but what it does it does do it does very well. I was told that PS CS5 worked properly in Windows and Adobe offered a reasonably priced up grade for CS2. I use PS occasionally for such as Content Aware (PSP's takes me for ever!), HDR.
Peter

[Polar01]

Is this the photo that you talked about in your post #54, that "has received a lot of praise", that you linked to, or is it "one of my efforts", in other words not that photo.

Cheers: Allan

[ajohnw]

It looks a little flat to me Peter but that's probably down to Adobe RGB. I only work in sRGB. It should print well even at a size over the one you intend. Only slight problem I can see is the pale line on the horizon. Probably down to sharpening but might be a line left over from selection.

As it looked a little flat on my sRGB monitor I made a very slight adjustment - just set the black point up a little and left the rest of the tone curve straight. Hope you don't mind. People sometimes do that on here. It's good practice for them and also helps posters.



John
-