Digital Photo Processes

-- Introduction -- Lens & image -- A/D conversion -- Lumix & Leica --  Sharpness -- CCD noise1 -- CCD noise2 -- 5 years later -- 

Comparable Lumix and Leica camera models

From its physical appearance and published specifications it appears that the Panasonic Lumix DMC-LX2 is very similar to the Leica D-LUX 3 and the Lumix DMC-FZ50 is similar to the V-LUX 1. The cosmetics are slightly different. The Leica versions have hard edges on the sides of the camera body. In the case of the D-LUX 3 this actually gives a firmer grip on the small body.  It is claimed on the Internet that JPG processing in the Leica model is less aggressive, but presumably there is no difference in the RAW output compared to the Panasonic version. Some of my tests showed a difference in the RAW output appearance between the FZ50 and LX2. Thus I wrote an email to Leica USA with the question: Before purchasing the Leica D-LUX 3 camera I need to know if the RAW files are truly raw data or if some processing has already been applied that smoothes the output from adjacent pixel locations. This apparently has been done in the related Panasonic DMC-LX2 as I noticed from analyzing its raw output in Lightroom and comparing it to the Leica V-LUX 1 output.

I received the following response:

The proprietary firmware in the Leica D-Lux 3 is designed around our preferences and is therefore different than that in the Panasonic Lumix model.  However we do not have any specific details available regarding how the Raw captures are handled in any of the cameras.

There is a considerable price difference between the D-LUX 3 ($600) and the LX2 ($377) and even more so between the V-LUX 1 ($849) and the FZ50 ($466). So what am I missing besides Leica's red dot and a one year longer warranty when I buy the Panasonic model?

In October of 2007 I had a private tour of the Leica M8 manufacturing facility in Solms, Germany. It had been arranged by a relative who used to work in the Leitz Wetzlar microscope development group. It was a most impressive introduction into Leica's dedication to product quality, to the tooling, testing and calibration which are necessary to guarantee the consistency of a superb design in production quantities. 
I also gathered some hints about the differences between Leica and Panasonic products. The lenses were designed and specified by Leica but are being manufactured in Japan. I saw in Solms how lens elements for the M8 were individually ground, polished and tested. This is a slow process even when rows of robots are used and it may take up to 40 minutes for a single, critical element. In contrast Panasonic uses mass production techniques where multiple lens elements are mounted to the same grinding and polishing head and they are all worked on simultaneously. This leads to some variability between element finishes. Furthermore, not every lens element is tested. Instead samples are taken and measured to keep the manufacturing process under control and within specifications. Thus one can expect some variability in the lens performance. See the Sharpness page. I could not find out if the Panasonic built lenses, when used in a Leica product, would have to meet tighter specification than what is shipped in a Lumix. I only learned that if Panasonic would ship a Leica branded product that does not meet Leica's specifications, it would have expensive consequences for Panasonic. Thus I assume that Panasonic keeps tight control on their lens production and will not use a potentially costly process of selecting lenses for Leica. More likely they will assure that every lens meets Leica specifications and tolerances.

The data processing firmware is designed around Leica preferences. So there is a difference, but it only carries an initial development cost. It adds value though.

The remaining critical element for picture quality is the light sensor, the charge-coupled-device or CCD. There is usually some variability in CCD performance due to imperfections in the semi-conductor material and chip manufacturing process. It shows up also in its noise performance. Thus one can assume that each CCD has been tested for acceptability and that possibly they are sorted according to their application. I have found from testing the four cameras that the FZ50 and V-LUX 1 sensors show very similar noise behavior and for the FZ50 sensor actually somewhat lower noise than for the V-LUX 1. The LX2 is noticeably noisier than the D-LUX 3 but slightly less so than the FZ50. The D-LUX 3 exhibits the lowest noise of the four cameras. It leads me to assume that it uses a selected CCD to meet Leica specifications. My tests are on the CCD web pages. 

So, if I am correct in my observations, what do you get when you buy the Leica V-LUX 1?
You get a camera that meets the Leica specifications and that is guaranteed not to exceed their tolerance limit. The FZ50 might have a higher tolerance limit on the CCD. If so, you might get one which is noisier than the V-LUX 1. In my case it appears that I tested a FZ50 which is noise-wise below the V-LUX 1 limit.

The D-LUX 3 appears to use a CCD with tighter specifications and if so will provide better image quality in low light portions of a photograph than the LX2. Both cameras use a CCD design that gives a different noise pattern than the CCD in the V-LUX 1 or the FZ50.

These are my conclusions about the camera differences. Since they are based on testing one sample from each model they are statistically speaking not solid. I just wished that Leica would be more forthcoming about what is really going on. Since the company is usually very committed to quality I will assume that this is also reflected in their compact digital cameras, and not just in the M-series. It would be valuable for a potential buyer to know what is special about a D-LUX 3 or a V-LUX 1 and other models that seem to have a lower cost "equivalent". Hopefully it is not only a Leica brand mystique.

----------------------------------------------------------------------------------

I had written to Leica in Solms about my comparison tests in March, 2008, and that I would appreciate their comments and corrections, hoping to add those to my web pages. Meanwhile their CEO, Steven K. Lee, has been fired. Presumably the company is in some turmoil. Thus it might take a while. Well, I never got a response to that message.

--------------------------------------------------------------------------------

Waiting is over!

The Panasonic Lumix DMC-LX3 is the camera that fits my wish list almost perfectly. It was first shown in July, 2008 in Japan and took a while to become available in the US. I received mine in September from J&R at the surprising preorder price of $399.  I have been happy with the camera ever since. In particular, I like being able to use the combination of wide F:2.0 aperture, ISO400 and image stabilization. It results in ambient light photographs with very good image quality without flash. On the other hand the hot shoe on top of the camera provides for using a more powerful flash with TTL settings and bounce. Based on the results of my previous comparison tests I have absolutely no longing for the equivalent Leica D-Lux4, but I greatly appreciate the results of the joint effort between Leica and Panasonic. 

	The LX3 is solidly built and feels substantial. Its LCD is significantly improved over the LX2, has much wider angles of view and even in bright sunlight remains useable. Still, I anticipate that the slip-in optical view finder will be a relief in extreme situations of sun and snow and fast action shots. The optional lens adapter plus a 46 mm to 55 mm adapter ring allow me to use the UV, ND and polarizing filters, and the magnifying lens from my FZ50 in special situations. The 24 mm to 60 mm zoom range is optimally suited for 90% of my type of photography and then there is still the FZ50 for special occasions. dpreview.com and DigitalCameraReview.com have detailed reviews and there is much chat about this camera on the web. The Canon G10, while impressive, is too bulky for me and lacks low light performance.
	The new LX3 is slightly larger and heavier compared to the LX2 (Leica D-Lux3). This just makes the camera more marginally pocketable and I often wear it in a Lowepro Apex 30AW on my belt. Outside Depth x Height x Width, and weight with battery LX3:  54 x 62 x 109 mm, 265 gram LX2:  52 x 58 x 105 mm, 220 gram
	My wife likes to carry a camera in her purse all the time to be ready to take photos of the grandchildren, flowers, friends or landscapes and not be bothered by technology. She loves her slim FX37 (Leica C-Lux3) with its intelligent Auto mode and a 25mm to 125mm zoom range and not the cameras that I handed down to her in the past. I appreciate the image quality of this camera which is more than adequate for its purpose. Outside Depth x Height x Width, and weight with battery FX37:  25 x 54 x 95 mm, 150 gram

In conclusion let me say that I like the engineering of the Panasonic/Leica partnership in the optical and mechanical departments. Their camera designs strike me as being more creative and better trade-offs. Their user interface is more to my liking than that of the Olympus, Nikon or Canon digital cameras that I have owned. Their weakness has been in the sensor electronics and image processing areas. That seems to have changed with the LX3 in particular. 

  LX3 in use

	The LX3 has accompanied me almost daily. On trips to the grandchildren, to the ocean, the city or the country side the camera comes with me in a small pouch that I modified to also carry a spare battery and an optical viewfinder. The pouch goes on my belt and sometimes on a strap over my shoulder. I attached a small loop so that the pouch is easy to pick up and carry by hand or to hang it over the gear shift stick in the car. When taken out of the pouch the camera often slips into a jacket pocket between taking pictures.
	The optical viewfinder is a necessity to me, because the LCD is only marginally useable in situations where the ambient light is very bright or when the right moment for pushing the shutter is decisive. It is a fixed 50 mm equivalent viewfinder with 1:1 magnification so both eyes can be open and I merely see the picture frame added (Voigtlaender Brightline Finder, 50 mm). I primarily use it when the LX3 is set to 60 mm focal length or close to that and it is like working with a fixed focal length viewfinder camera of olden days, except that I also have the benefits of the LCD for information display and composition. The axis of the viewfinder had to be adjusted vertically to meet the camera lens axis at large distance. This was done by loosening two small Philips screws in the viewfinder shoe and adjusting a set screw until LCD and viewfinder frame matched top and bottom.
	I had first bought a 24 mm viewfinder but did not find it useful because of the scaled down magnification.   The pouch (Lowepro Apex 30 AW) comes with a tongue like internal pocket. I cut it off at the base to make more room for the camera. Then I shortened it a bit so that it slips into the mesh pocket, which is on the inside of the pouch lid.  The viewfinder goes into the small top pocket which has a Velcro closure. For the battery I added a loop so that the whole assembly is easily put into the mesh pocket and taken out for changing batteries.

------------------------------------------------------------

FX580, 25mm, 1/640s, f/2.8, ISO125, 12.1Mp, 1/2.33"CCD, 581x581 pixels	Comparison between FX580, LX3 and TX7  Three pictures were cropped about 7 to 1 to show actual pixels on your monitor. Ignore the color differences and note the rendering of detail for each camera. Standard camera settings were used and no post-processing was applied.
LX3, 24mm, 1/400s, f/3.5, ISO125, 10.1Mb, 1/1.63"CCD, 537x537 pixels	TX7, 25mm, 1/640s, f/4.5, ISO125, 10.2 Mp, 1/2.4"CMOS, 534x534 pixels

-----------------------------------------------------------------------------------------------

-- Introduction -- Lens & image -- A/D conversion -- Lumix & Leica --  Sharpness -- CCD noise1 -- CCD noise2 -- 5 years later --