davephoto Posted October 24, 2006 Share #21 Posted October 24, 2006 Advertisement (gone after registration) 4/3 sensor and an optical viewfinder i like that, salvatore. or partner up with fuji to use -- or just buy from them for that matter -- the unbelievably low-noise 1/1.7" sensor fuji is sticking into their F series compacts. Link to post Share on other sites More sharing options...
Advertisement Posted October 24, 2006 Posted October 24, 2006 Hi davephoto, Take a look here D-Lux 4 Market Research. I'm sure you'll find what you were looking for!
OttoG Posted October 24, 2006 Share #22 Posted October 24, 2006 Regarding the lens cap, I was skeptical about it at first, but then I realized that it is very useful for improvising a camera support. Put the camera on a solid surface, and if it needs to be pointed a bit more upwards, just put the lens cap underneath the extended lens. So, it is actually quite useful, and it always accompanies the camera. (My experiences refer to the DMC-LX1/D-Lux 2.) Regarding an optical viewfinder, please no, at the very certain expense of a larger and heavier camera. A hot shoe for external viewfinder would be no problem, especially if it doubles as an external flash connector. The real issues are sensor sensitivity and flash. Putting in a larger sensor or faster lens is probably not realistical without significantly increasing size and weight, but if Fuji can make a small sensor with high sensitivity, why can't Leica/Panasonic? A sharp but somewhat noisy image in RAW mode at ISO 400 would be a great improvement. The internal flash, which is almost useless at ISO 80, except for close-up and wide-angle pictures, would of course automatically be improved by better sensitivity, but what we really need is a connection for an external flash with the ability to bounce off a ceiling with full automation. Preferably with the choice of using it together with the camera or separately with a cable or wireless connection. (By the way, has anyone tried the Metz 28 CS-2 digital with a D-Lux 2/3? Or any slave flash. Does the camera always fire at least two flashes, or does it with some settings only fire one?) Otto Giesenfeld Link to post Share on other sites More sharing options...
hawkeye_a Posted October 24, 2006 Share #23 Posted October 24, 2006 i like that, salvatore. or partner up with fuji to use -- or just buy from them for that matter -- the unbelievably low-noise 1/1.7" sensor fuji is sticking into their F series compacts. Hey Dave, That gallery your linked to is awesome. Any chance of getting higher resolution versions of a couple of em ? Cheers Link to post Share on other sites More sharing options...
Guest stnami Posted October 24, 2006 Share #24 Posted October 24, 2006 Bugger this pussyfooting around lets go the D-Lux 5CL Link to post Share on other sites More sharing options...
Riley Posted October 24, 2006 Share #25 Posted October 24, 2006 provide a 'turn off venus 3 switch' Riley Link to post Share on other sites More sharing options...
humanized_form Posted October 24, 2006 Share #26 Posted October 24, 2006 i would suggest that Panasonic/Leica do NOT make a D-Lux 4 until they actually have something innovative to offer in a new model. i advise that they stop with the annual(semi-annual?) release of the very slightly upgraded "new" models that simply add features and more pixels. i think you will eventually lose the interest of your market if you continue down the path of marginal upgrades, cameras released just to have something "new" on the shelf. Link to post Share on other sites More sharing options...
peterb Posted October 24, 2006 Share #27 Posted October 24, 2006 Advertisement (gone after registration) Since we're all dreaming here I tell you what I dreamt last night. The DLux 4 would have the Foveon chip designed for the Sigma cut to 16:9. Pixel count isn't so important as noise level is. Plus because the Foveon concept is 3 layers of pixels analogous to color film, the pixels would be quite large and therefore would have less interaction between. The total count would be the sum of the layers, thus THREE layers of 4MP each would translate to a chip with a grand total of 12 million pixels. Not too shabby eh? Plus a nice accessory shoe for optional optical finders. And with that, the option to turn off the LCD. A built in finder would be a compromise and the add ons I've found are quite accurate. While the buffer for RAW is good if there's a chance to make it bigger I'm all for it. Plus a vario Summicron. Oh wait...then no one would buy the M8.... Regards, P P.S. Dave....AWESOME SHOTS. Very enviable collection there. Link to post Share on other sites More sharing options...
farnz Posted October 25, 2006 Share #28 Posted October 25, 2006 The total count would be the sum of the layers, thus THREE layers of 4MP each would translate to a chip with a grand total of 12 million pixels. Now this is the bit that confuses me about Foveon sensors: as I understand it each photo detector captures red, green and blue wavelengths at different depths and combines them to produce a composite pixel (picture element). In non-Foveon sensors each photo detector site produces a composite pixel using a Bayer filter so 4 million detectors would yield about 4 million pixels. Similarly then in a Foveon sensor there can therefore only be a total of 4 million photo detectors or pixels rather than 12 million as the Foveon's marketers would have us believe? Or have I misunderstood? Link to post Share on other sites More sharing options...
jgmb Posted October 25, 2006 Share #29 Posted October 25, 2006 (Peter, great minds think alike... as I was composing my post to suggest the Foveon sensor, you beat me to it!) My wish for the D-Lux 4 might more properly be classified as a wishlist for a whole new class of niche camera, positioned in between the D-Lux 3 and the M8. My dream camera would be a compact digital version of the Hasselblad XPan, but with 16:9. May I propose the "Leica Lux-Pan"? It would have the following features: 1) A much larger 16:9 sensor, proportionally same size as APS at the very least. 10-12 megapixels 2) Even better, I would love to see the Foveon X3 sensor paired with a sharp Leica lens! That could be the ultimate in accurate color rendition. 3) While zoom is nice, I would sacrifice it for a fixed lens of higher quality, perhaps a 28mm Summicron optimized for the larger 16:9 sensor. 4) While such a sensor would necessitate a slightly larger body, I would accept a size similar to the Leica CL with a rugged titanium frame, and same build quality as the M8. All black! 5) Hotshoe with an optional optical viewfinder (like the Ricoh GR)... or built-in optical rangefinder with 16:9 frame 6) OLED 16:9 finder or high-definition LCD (double the number of pixels) 7) 16 bit RAW color capture. No AA filter (not needed with Foveon sensor) 8) Larger RAW buffer 9) Auto-stitching alignment software to allow precise wide panoramas. 10) Tripod mount centered on optical axis. This would make the ideal compact landscape and travel camera: a fixed-lens "baby M8" or "digital XPan." For all the above, I would gladly pay $1,500-$2,000. Jeff Link to post Share on other sites More sharing options...
peterb Posted October 25, 2006 Share #30 Posted October 25, 2006 Farnz, I hope I can explain this right. In the Bayer model of a chip, the entire surface of the chip is made up of three kinds of sensors. Those which record red light, those that record green light and those that record blue light (the three components of light in its inifinite combinations of hue and intensity). Since the entire light collecting surface area in this approach has to accommodate all three sensors at once, chip companies have gone overboard trying to pack as many sensors as possible onto the surface. Their reasoning, which is makes perfect sense, is that the 'finer' (or smaller) you make the sensors the more accurate the representation you'll have of each ray of light that hits it. The problem is, no matter how infintessimally small you make the sensors, each sensor still can't account for the accuracy of the color it's NOT receiveing. That is the light that falls on the sensors next to it, that are gathering information for their OWN colors. So for example given a red sensor no matter how small still will not be gathering the 'red' information in the light that's hitting the green and blue sensors that are sitting next to it. So how is this achieved? Through software. And the forgiveness of human vision. The software taking in the information of a red sensor, then, through some algorithm, predicts the red information that would have been received in the other two places where the other sensors are. Helping the software is, of course, information provided by the next NEAREST red sensor, for example) and then calculating the levels of red determine that the gap between the two sensors (where sensors are gathering most likely blue and green information) and figuring out what that 'red' is. In the end, you look at the picture made up of all those tiny little dots of red blue and green and your brain mushes them all together to help you perceive the right color. (Or so you think. Or, as they say, close enough for government work.) So while the end result is probably fine for the most part, you still don't have an accurate read of any one wavelength since you're only gathering the information from only one third of the sensors on the surface of the chip. As manufacturers make the sensors ever so tinier to increase the number of sensors among each primary for a given surface area two problems arise. One, by making the sensors smaller in the attempt to achieve a finer read of any part of the image, each sensor gathers less light since in order to make a sensor smaller the surface area which light will hit it will also become smaller. With less light hitting smaller sensors Mr. Einsteins' photoelectric effect will produce an inherently smaller signal. That means what little signal is produced will have to be amplified somewhat. And that is where the noise comes in. But no matter how many sensors/pixels you pack on a traditional Bayer chip, the total number of pixels for any one color will be a third of the advertised number in the ads. A 6 mp chip for example means that there are 2 million red sensors, 2 million green sensors and 2 million blue ones. With the Foveon approach there are THREE layers of sensors, one on top of the other as in film. Each one responsible for one primary color. As a result EVERY square millimeter on the chip's surface is accounted for in any one wavelength that is to be captured. No interpolating. No fudging. No kidding. On top of it all, with the Foveon approach, each layer can be designed for maximum resolution, maximum light gathering and minimal noise. So, whereas on a traditional 6 mp chip, say, there would be in actuality 2 million blue, 2 million red and 2 million green sensors all packed into a tight array on the chip, in a Foveon version of the chip, assuming you used the SAME sized pixels, you could create three layers, EACH with 6 million sensors on it or a total of 18 mp! Foveon, however, uses a smaller number of very LARGE sensors on each layer. So you have larger, light gathering sensors with less noise produced (theoretically). And the image information is not any less resolved than would have been obtained using the Bayer arrangement of sensors. Hope that helps. P And Jeff, GMTA!!!! Spot on. P Link to post Share on other sites More sharing options...
rvincent Posted October 25, 2006 Share #31 Posted October 25, 2006 Jeff and Peter - count me in. When I read about the Sigma DP1 with the Foveon chip I immediately thought how sweet it would be to have that camera with a Summicron lens and hotshoe (for a real optical viewfinder). My hope is that the M8 is so successful that Leica will have the confidence (cash) to build a real compact digital camera. Cheers, Robert Link to post Share on other sites More sharing options...
Kirk Posted October 25, 2006 Author Share #32 Posted October 25, 2006 I haven't seen anything to suggest that the Foveon sensor can actual produce the results they claim... especially at 14 megapixels. What about a version of the Kodak sensor that is in the M8? Link to post Share on other sites More sharing options...
jgmb Posted October 25, 2006 Share #33 Posted October 25, 2006 Kirk, they may only be tests taken with a prototype Foveon X3 sensor, but check out these photos: http://www.sjphoto.com/web-special/index.htm If they can get this kind of clarity and sharpness with a Sigma lens, imagine the results with a Summicron. And yes, I would gladly accept the M8's Kodak sensor if they could make it in 16:9 format. Link to post Share on other sites More sharing options...
Kirk Posted October 25, 2006 Author Share #34 Posted October 25, 2006 Thanks for the link to the foveon image samples. I am impressed! Link to post Share on other sites More sharing options...
albertwang Posted October 25, 2006 Share #35 Posted October 25, 2006 Foveon is improving quite a bit. The DP-1 looks like the potential Minilux here. Sean will be testing it hopefully! Link to post Share on other sites More sharing options...
farnz Posted October 25, 2006 Share #36 Posted October 25, 2006 Peterb, Thanks for your excellent explanation! Without wishing to sound over-cynical it appears that the non-Foveon manufacturers are also 'inflating' the number of pixels their sensors contain so we are comparing apples with apples which is what I was unsure about. Farnz. Link to post Share on other sites More sharing options...
robert_parker Posted October 25, 2006 Share #37 Posted October 25, 2006 Having had D-Lux 2 and now got D-Lux 3: 1) Provision for an optical viewfinder, maybe via hotshoe 2) Care not to loose any build quality 3) Slightly increased form factor would be tolerable but front to back 'thickness' is critical if it is to go into a pocket. Weight is also important. 4) How about an interchangeable lens ? 5) Not over bothered about 16:9 Link to post Share on other sites More sharing options...
robert_parker Posted October 25, 2006 Share #38 Posted October 25, 2006 Just noticed that there is an unused D-Lux 2 going at The Classic Camera in London - Eeeek!! The Classic Camera - The Leica Shop Link to post Share on other sites More sharing options...
zymonk Posted October 25, 2006 Share #39 Posted October 25, 2006 1. Add threads to the lens. Filters can be usefull and fussing with a lens cap can be a big drag. 2. Improve the noise. 3. One of my biggest gripes with my LX1 after using it so extensively turned out to be with the flash. I'm not talking about the size, power, coverage ect... which I think are fine for it's size. My complaint is with the setting options and specifically with Slow Sync. Why is it only available with red eye reduction and why oh why is Slow Sync not available in shutter priority and manual modes? This totally defeats the purpose of a Slow Sync mode and limits creativity. If you can't control the shutter speed you can't control the blur. Urrrr. Link to post Share on other sites More sharing options...
OttoG Posted October 26, 2006 Share #40 Posted October 26, 2006 Why is it only available with red eye reduction and why oh why is Slow Sync not available in shutter priority and manual modes? What am I missing here? In shutter priority and manual, you by definition have slow sync just by choosing the desired shutter speed, don't you? That would presumably solve the red-eye reduction problem too. Or is the camera able to sync on the end of the shutter cycle (like second-curtain sync on SLRs) in the other modes but not these two? I never played around with slow sync yet on the DMC-LX1. I used to use it with compact film cameras, but with digicams, I tend to find the results disappointing compared to adjusting white balance and doing a long flashless exposure. While I guess this should be a new thread, once again, does anyone have experience with external flashes for this camera line? Link to post Share on other sites More sharing options...
Recommended Posts
Archived
This topic is now archived and is closed to further replies.