A Look at Four Compact Digital Cameras

Now that the prices for good digital SLRs have come down to about $2000 new and $1000 used, there are probably three main reasons for a serious photographer to consider a compact digital camera.

1. They're less expensive than digital SLRs.
2. They use a viewfinder with unlimited depth of field vs. an SLR which normally sees the world at somewhere between F/2 and F/2.8
3. They are small and light enough to fit in a large jacket pocket, allowing them to be with you all the time.

That last advantage is what prompted me to buy a compact camera.  I use a Canon 1Ds digital SLR for most of my personal and professional work.  But I like to have regular daily practice and having a small camera in my jacket pocket allows the camera to be with me almost all the time.  Within the normal routines of my daily life, taking care of a young daughter in the winter, in and out of cars constantly, in and out of snow and rain, etc.. I would need several extra hands to have a normal SLR with me throughout the day.  

To the best of my knowledge, no one is yet making a serious compact digital camera with a viewfinder like a Leica M7 and the AF/processing speed/image quality of a Canon 1D.   That combination would yield an excellent camera and I imagine someone will build it in the next few years.  Imagine...a big, bright, sharp finder and cat-like reflexes (instant focus/exposure lock, instant shutter response).  That combination exists in some professional DSLRs but they won't fit in a pocket.  

What I was looking for was a compact camera that came closest to this ideal.  My priorities were response speed, viewfinder usefulness, manual control and image quality.   I also shoot extensively in existing light and so ISO 400 performance was important as well.  Since I can’t make a picture I can’t see, the viewfinder is critical.  The Leica M series film cameras certainly set the standard here with large, bright and sharp viewfinders that provide a photographer with the kind of information he or she needs to make a strong picture.  I ruled out the EVF (electronic viewfinder) cameras after trying them.  For me, using their finders was like looking at a grainy television picture through a cardboard tube.  Some of the EVF cameras, such as the FujiFilm FinePix S602Z, have sharper displays than others but overall, they lack the clarity of a good SLR or viewfinder/rangefinder camera.

My research lead me to compare four digital compact cameras: The Leica Digilux 1, the Canon G2, the Nikon CP5000 and the Olympus C5050Z.   There are other options of course, but I found that these seemed to be the top performers among the cameras that would fit in a jacket pocket.  Note: The Canon G2 has largely been replaced by the Canon G3 but the newer camera's finder is partly blocked by its lens, a design weakness that again demonstrates how little attention manufacturers are paying to these finders. That problem took it out of the running for me right away.

First off, none of these three cameras are as fast as a good digital SLR.   People often discuss "shutter lag" with these cameras but in fact the real lag is the focus/exposure/preprocessing times.  All three of these cameras are slower than the best digital SLRs  from the time one half-presses the shutter to the time the green "focus/exposure set"  light appears.  Once the magic green light is on, all three respond quickly to the shutter button.  So making them responsive to the moment of the picture requires some technique.  I can bring my Canon 1Ds, which is essentially a state of the art DSLR as of this writing,  up to my eye and within milliseconds it locks focus and trips the shutter.  The lag time on the 1Ds is so miniscule that I couldn't reliably measure it.  My own reaction time was more measurable than the camera's lag.  By contrast, the Olympus C5050Z camera reviewed here, which is among the fastest compacts available,  has a total lag of 1.84 seconds in autofocus mode and 1.59 seconds in manual focus mode when using a 32 MB Olympus XD card.  That's not perfect but it's faster than older compact models.  Three of the four cameras here include fairly fast (about F/2 to F/2.5) zoom lenses.  The Nikon lens is slower but zooms wider than the others.

Fourth Place: Leica Digilux 1

Leica has been much maligned for having let Panasonic build the Leica Digilux 1 but they at least addressed the compact camera speed issue better than most manufacturers; the Digilux 1 locks focus and exposure quickly and has very little shutter lag.  In other words, it's closer than most compacts to having the reflexes of a good film camera or DSLR.   Unfortunately, I did not have the opportunity to measure it's exact lag times.  The camera is much faster in operation than most but has several weaknesses.  

I was disappointed to find that it's finder is small and not very sharp; one would hope that the company who gave us the "M" series would design a "photographer's finder" for the Digilux but alas they did not.    It  can't display a histogram after capturing a picture and that removes a very useful exposure tool from the digital photography process.   Shutter speeds and lens openings are not shown in the top LCD display so one has to turn on the large LCD in order to check and change exposure settings.   The camera does have a large and sharp, albeit uncoated and thus reflective, 2.5" LCD screen with a nicely designed folding hood but the design of the camera almost forces one to use this LCD screen instead of the optical finder.  Some may like that method, but I find it to be a slow and alkward way of working.  The camera also exhibits excessive noise in its ISO 200 and ISO 400 files and is unable to record a RAW file; in both respects it is outpaced by competitive cameras. With a better finder and a camera top display of exposure settings, it might be one of the best options available despite the image noise.  But without an excellent findn't.

Crop from 100% sample of Leica Digilux 400 ISO capture, from original file  (tripod, 1/7 second, F/2.5)  Courtesy of Benny So

 

Third Place: Canon G2

The G2 is another camera I really wanted to love.   My current Canon digital SLR was preceded by an excellent Canon D30 digital SLR.  I also have owned several of the excellent rangefinder cameras Canon produced in the 1960s and 70s.  Finally, I really like the Canon CRW which is a mature RAW format supported by a wide range of viewing and processing software.  The G2's general response speed was similar to that of the Nikon, which is to say slower than that of the Leica and Olympus, and it seemed well-designed in most respects.  Unfortunately, I did not have the opportunity to measure it's exact lag times.  It has probably the worst optical finder of the group: tiny, blurry, prone to easy flare, etc.  With a better finder, it could be a much better camera.   The ISO 400 files it produced were also not as good as I expected, despite being captured in RAW format which gave them their best possible quality.   It's part metal, part plastic case also felt more vulnerable than the others.  This has been a very popular camera but I found it to be serious eclipsed by the Olympus in most respects. Canon is currently producing superb DSLRs and certainly has the resources to produce a serious digital rangefinder camera.   As of yet, they haven't chosen to do that.

Crop from 100% sample of Canon G2 400 ISO capture, from original file (1/160, F2.8)

 

Second Place: NIkon Coolpix 5000

Nikon is another company that has produced excellent film rangefinder cameras in the past.   That experience shows to some extent in the Coolpix 5000 which has as good an optical finder as I've seen in any compact digital camera.  It's still not as good as it could be but it's larger, brighter, sharper and thus more useful than most.   It's general speed of operation is similar to that of  the Canon which makes it slower than the fastest cameras reviewed here.   It also allows one the option to capture in a 2 by 3 ratio frame (like traditional 35mm negative) instead of a 4 by 5 ratio frame (like a traditional view camera negative).  The  LCD display accurately shows the 2:3 ratio image but the optical finder's etched frame for that format is not all close to what the camera captures.  In order to accurately compose the image in a 2:3 ratio, one would need to use a home-made mask over the optical finder.

Unfortunately, the camera's maximum lens opening varies from F/2.8 at wide angle to F/4.8 at telephoto which means it needs a stop or two slower shutter speed than the other cameras listed here (most of which have lenses that open to about F/2 at wide angle and  about F/2.8 at telephoto) to get the same exposure in low light.    To it's advantage, it's the only camera here to have an ISO800 option, gaining back one of the stops it loses in lens speed, but it's ISO 800 files are noisy.  (Note: I did not have the opportunity to make an ISO 400 sample file from the CP5000.  If anyone would like to send me an original JPEG @400 ISO from this camera, I'll include it.)  Given the large amount of DOF these little lenses have, I'd prefer F/2 @ ISO 400 to F/2.8 at ISO 800.   Although it's slower than the others, it's zoom lens goes as wide as 28mm (35mm camera equivalent), vs. the 35mm or so wide zoom limits of the others.  That can be a real advantage in some situations.  Nikon, like Canon, also has a well-established RAW file format called NEF which is well-supported by a variety of third party software.     If one can live with the slower lens and slower response times of the camera, it does have much to recommend it with a decent finder, a wide lens and a versatile RAW format.

First Place: Olympus C5050Z

At first it seemed that the Olympus was not appreciably faster than either the Canon or the Nikon to focus or set exposure.  Later I realized that the camera is particularly sensitive to recording media.  I had initially tested the Canon, Nikon and Olympus with my usual 1GB IBM Microdrives.  In later testing with the Olympus-supplied 32 MB XD card, a Lexar 4X 64MB compact flash card and a Sandisk Ultra 256 card,  I discovered that the solid state media gave the Olympus a huge speed boost.  One would expect a difference in write times and review times but the recording media had a huge affect on shutter/exposure/focus/white balance lag.  I have not yet had the opportunity to test the Canon and Nikon using solid state cards and so cannot say what impact that might have on the operational speed of those two cameras.    It is interesting to note that the two fastest cameras in this comparison, the Olympus and the Leica, both use the latest miniaturized storage media: XD for the former and SD/MMC for the latter.   It would seem that at least the Olympus, and perhaps the others, do some brief pre-writing to the storage media before the shutter can be released.

Having no access to sophisticated test equipment, I evaluated the C5050Z's total lag time by photographing an "on monitor" stopwatch and subtracting the start time from the time shown in the resulting file.  For example, the review file shown in the camera's LCD screen might show 6.95 seconds.  I would subtract the start time of 5.0 seconds from the capture time of 6.95 seconds and get a total lag of 1.95 seconds.   My own reaction time is a potential confounding variable with this method but by doing twenty trials for each card type and taking an average of the trials, that variable is mostly eliminated when looking at comparative speed.  That is to say, my trigger response (be it fast or slow) affects all of the card tests more or less equally once the results for each are averaged.

Olympus C5050Z Lag Tests

1. Camera was set to auto-focus mode, shutter priority, 1/60 second (which works well for photographing a monitor), auto white balance, super high quality, 2560 x 1920 pixels, ISO 400.  Camera to subject distance was about 14".

2. Camera was set to manual-focus mode at 14", shutter priority, 1/60 second (which works well for photographing a monitor), auto white balance, super high quality, 2560 x 1920 pixels, ISO 400.  Camera to subject distance was about 14".


Media

 Total Lag: Auto-Focus Mode Total Lag: Manual-Focus Mode RAW File Cycle Time*
Sandisk Ultra 256MB card

1.92 seconds

1.56 seconds

8.39 sec.

Olympus 32 MB XD card 1.84 seconds 1.59 seconds 10.45 sec.
Olympus 256 MB XD card 2.43 seconds not tested 10.98 sec.
Lexar Media 4X 64MB card 2.41 seconds 1.73 seconds not tested
IBM 1 GB Microdrive 3.92 seconds 3.22 seconds 9.61 sec.

*This test measured the time it took for a RAW file to be saved and the shutter tripped again for the next picture.

In both manual and auto-focus mode, both the 32MB XD card and the 256MB Ultra CF card make the camera more than twice as fast as it is with the 1-Gig MD.  In actual use, that difference feels gigantic - the difference between getting or missing the moment.   Surprisingly, the 256MB XD card is not a strong performer at all, turning in slow times for both total lag (auto focus mode) and RAW file cycle time.  On the other hand, the MD, once it starts up, is actually quite fast to store a RAW file.


The Olympus  has various distracting sound and screen features which, fortunately, can be switched off.  But  it also tied with the Nikon for having the best finder of the group.   Since I can't make a picture I can't see, this was an important advantage.  Knowing that none of the finders were up to snuff, I wanted to get the best possible.  Like the Nikon and Leica, the Olympus has a rugged, mostly metal body that fit my hand better than the Canon or Leica.  It also will record in RAW mode in Olympus's proprietary ORF format but this is not as mature a format as Canon's CRW or Nikon's NEF and is not yet well-supported by third-party software. The included Camedia 4 software does a terrible job with ORF conversions, adding artifacts and not allowing much control over the conversion process, but Adobe's new Camera RAW plug-in (for Photoshop 7) does an excellent job with C5050Z RAW files.   

The Olympus can work in a "shooting priority" mode where it can be set to a recording mode, such as shutter priority, while still allowing one to review pictures on it's LCD by pressing it's  "quick review" button twice quickly.  It isn't necessary to switch into play mode to review pictures and a press of the shutter button returns you to shooting mode.  This is in some ways similar to my Canon 1Ds and is a very useful feature.  I don't know if the other cameras have this option.  It also has a "live histogram" feature which allows you to evaluate exposure graphically on the LCD screen while you're making adjustments.  That's also handy.  When the camera is set to "not reset" it will preserve selected ISO speeds, white balance setting, shutter speed, focus distance (if in manual mode), etc. when it's shut down.  This greatly speeds use.  Like the Nikon, it also allows one the option to capture in a 2 by 3 ratio frame (like traditional 35mm negative) instead of a 4 by 5 ratio frame (like a traditional view camera negative).   I prefer the 2 by 3 ratio and may experiment with that setting.  The  LCD display accurately shows the 2:3 ratio image but the optical finder has no way of indicating the borders of an image captured in 2:3 ratio.  In order to accurately compose the image in a 2:3 ratio, one would need to use a home-made mask over the optical finder. 

The C5050Z also has a button located on top of the camera which can be custom configured to quickly make one of several adjustments.  I set it to adjust ISO so that a press of this button automatically turns on the LCD screen and opens the ISO setting menu.  A quick turn of the command dial resets the ISO and then the screen turns itself back off.  It's fast and direct, just the way these controls should be.

Evaluating the JPEG picture quality of the Olympus really made for a roller-coaster ride.  At first blush,  25% view in Photoshop, they look wonderful.  The same file printed at 8" x 10" on my Epson 2200 also looked great.  But expanding the file to 100% view in PS revealed some ugly artifacts that were at first very disappointing.  But, following a tip I'd read in another review, I did some experimenting with backing off the camera's internal sharpening to reduce the artifacts it was creating.  Backing it off to -5, which hopefully means completely off, made a dramatic improvement in the file quality.  Once adjusted that way, the ISO 400 file was the best quality I'd seen from a compact digital camera.  The Olympus ORF files, if they work as they should, will not have any in-camera processing at all applied to them.  That is the case for the Canon CRW files made by the G2 and the NEF files made by the CP5000, as well.

 

Crop from 100% sample of Olymous C5050Z 400 ISO capture, from original file (handheld,1/40, F5.6), flash on,
sharp -5, contrast -5, saturation -1


The lack of noise and general smoothness of the file is impressive.  It's nothing like a 1Ds file at that ISO but it's pretty good for being such a small sensor.  Rolling back the sharpening (to reduce artifacts) and contrast (to increase dynamic range) is the closest one can get the Olympus JPEGs to RAW.

There are many aspects of the cameras which I haven't addressed here.  I evaluated them in terms of my priorities and it should go without saying that yours may be different.  In particular, photographer's who primarily use the LCD screen to compose their pictures may prefer the Canon or Nikon LCD screens which twist and tilt in various directions as compared to the Olympus screen that only tilts up and down.  Photographers who want the 28mm coverage of the Nikon might trade that for the slower 2.8 max aperture.  

This review does not include detailed technical specifications. For excellent and thorough reviews of the Canon, Olympus and Nikon which include detailed specs. see:

http://www.imaging-resource.com

http://www.dpreview.com

These are the two most thorough sites I've seen, the former includes a "comparometer" which is very useful for comparing two pictures of the same scene made by different cameras.   It appears that the Olympus was used at default settings for the reviews at both sites and as such it's picture quality is worsened significantly by the in-camera sharpening.  The Canon and Nikon don't seem to use such destructive internal sharpening methods at default settings.

The Leica has not been reviewed much but it's near twin the Panasonic Lumix LC-5 has been reviewed at:

http://www.dcresource.com/reviews/panasonic/dmc_lc5-review/

http://www.steves-digicams.com/2002_reviews/lumix_lc5.html

 

Link to my other equipment notes and a test of portable hard drives here.