What you will find in this review is the camera's specs, annotated with the discussion of their significance and real meaning, plus my own, often subjective and opinionated, impressions.

The article is addressed to the advanced amateur who would like to gain some knowledge of the camera beyond the bare specs or the description of control placement.

The review, originally posted in December, 2000, is based on production-run cameras I bought through regular retail channels. At the time of this update (June, 2004) I have used the E-10 for more than two years before selling it to a friend; E-20 — for almost three years, and the text is still being updated from time to time.

(If you have to know: the pictures at the right were also taken with the reviewed cameras, about 1s exposure at F/8, diffused daylight.)

The finish is first-class. Handling this camera is a pleasure.

The few plastic parts do not feel cheap, and the grip is additionally covered with a soft, textured layer. Very nice.

The tripod socket is well-positioned and solid.

Top

• Dust and humidity protection: a fixed-lens design allows for sealing the optical sensor from elements.

  This argument does not convince me in 100%. E-10/E-20 uses a non-movable, double-prism beam splitter instead of the mirror. The CCD array is safely located behind it, while the front face of the splitter is less sensitive to cleaning damage than the CCD surface itself.

• Dimensional rigidity: with the sensor being 1/4 of the (linear) size of a film frame, this is more critical than in 35-mm cameras. Fixed-lens design alleviates any problems related to interchangeable lens mount precision and wear.
• Sensor requirements: the CCD light sensitivity varies greatly with the angle of incoming light (one more reason for the inverted telephoto design, and the hefty lens size). Therefore using lenses not specifically designed for this camera might result in inferior results.

  Again, this is not the best argument. Agreed, using lenses designed for a film-based SLR is out of the question. Most importantly, they are designed for a different frame size, with different sharpness requirements (that's why I'm not so hot Canon EOS D30/D60). Still, I would like to see a camera body with its own system of 3-4 interchangeable lenses designed specifically for it.

Especially for a non-interchangeable lens, the 35 mm short end is painfully limiting.

• F/2.0-11 (at 35 mm)
  
• F/2.4-11 (at 140mm)
  

For the same reason I would avoid using apertures smaller (F-numbers greater) than F/5.6, especially at the short end of the zoom range, except for close-up work.

The lens uses the inverted telephoto (retrofocus) design, with its entrance pupil much closer to the image sensor than the physical location of lens elements. In 35-mm SLRs this design is used for wide-angle lenses only, to make room for the swinging mirror. Here, I believe, it serves two additional purposes: making the light strike the CCD at smaller angle (as measured from perpendicular), and allowing the lens construction to be physically larger, to make the mechanical tolerances more acceptable.

While the lens is very nicely made and appears really sharp, it displays some barrel distortion at the wide end (not noticeable at 70 mm and above), as shown in this picture, taken at EFL=35 mm ("EFL' stands for "Equivalent FOcal Length".

The lens is bigger than an average 35-140 mm zoom on a 35-mm camera. This is partly due to the inverted telephoto design (see above), but clearly the lens was also designed with the optical front-end attachments in mind (mind it: the 3x, 420 mm-equivalent telephoto attachment causes only 1/2 F-stop aperture loss; now, try to buy an F/2.8, 400 mm lens for an SLR at less than $2000!).

Both zooming and focusing are internal; i.e., do not cause front element rotation or any changes in the overall lens length. A great plus, allowing for use of sculpted hoods, polarizers and graduated filters.

This is not a parfocal design, where the focus setting stays put while you zoom, but rather a varifocal one, like in almost all digital camera lenses. When you change the focal length, the focus setting changes as well.

Or, rather tries to change, as the camera logic immediately activates the focusing micromotor and adjusts the focus setting to the new focal length. This works in both auto and manual focus modes. In other words, although the lens is of a varifocal design (optically), it pretends to be of a parfocal one (operationally). Smart.

Top

To avoid mechanical complexity (and to facilitate LCD monitor use), Olympus uses a beam-splitting prism in the optical path. Some of the image-forming light passes through it to the CCD (this is what creates the recorded image, also used for the LCD viewing), and some is reflected under right angle to the horizontal groundglass above the prism. An image is formed on that groundglass, and it is viewed through the eyepiece, like in a "regular" SLR.

To allow for eye-level viewing of the groundglass image, there is a "mirror prism" above it (as opposed to a solid-glass pentaprism used in "classic" SLRs).

These two design decisions (splitter and mirror "prism") are quite meaningful. First, the splitter causes a loss in the amount of light reaching the image sensor. This results in lowering the effective ISO-equivalent sensitivity. It may be the reason why the E-10's CCD is rated at 80 ISO, and not 100, like in most other cameras.

Second, the mirror "prism" (I'm stubbornly using quotes, as it is not a prism at all) usually results in darker viewed image than a pentaprism — in addition to the fact that we are already using only the part of the light diverted from the sensor.

This is why I expected the viewfinder to be rather dark. Surprise: the viewed image is quite bright and useful in both composing the picture and focusing. The finder is comparable in brightness to that of two Minolta AF SLR bodies with similar focal-length (although, obviously, darker) lenses I'm using; one has a pentaprism, and another — a mirror "prism". It is no match, however, for my old XD-5 which uses a pentaprism and an Acute-Matte screen, but this is true of almost all present-day AF SLRs.

There are two obvious reasons for the viewfinder brightness being greater than expected:

• Brighter lens: almost two stops brighter (that's a factor of 4!) than an equivalent 35-140 zoom on a 35-mm SLR — F/2 as opposed to F/3.5 or F/4 (unless you buy something really expensive)
• Smaller finder magnification (see below).

Note on terminology: some reviewers refer to the E-xx cameras as not "true SLRs", some because of non-interchangeable lens, some because of the beam-splitting prism as opposed to mirror. Wrong on both counts: an SLR (single lens reflex) is a camera with optical viewing through the lens, and it does not matter whether the mirror plane is semi-transparent or not. As a matter of fact, a number of 35-mm cameras used the same approach (with the venerable Canon Pellix being perhaps the first). The acronym "ZLR" (zoom-lens reflex) is just a marketing trick, as "ZLR" sound better than "non-interchangeable lens SLR".

(Some reviewers report a 95% viewfinder area coverage, not understanding the difference between linear size and area. If I cut an image in half vertically and horizontally, I end up with 25%, and not 50% of the image. Something is wrong with our school system.)

With an almost full coverage, and a bright image, I found myself using the optical viewfinder almost all time, with the LCD monitor useful mostly for reviewing pictures already taken (plus some close-up work).

• at 35 mm: 0.42x
• at 50 mm: 0.6x
• at 140 mm: 1.6x

• Autofocus status
• Flash status
• Macro focus mode
• Shutter speed
• Aperture
• Metering mode
• Exposure override
• Exposure lock
• White balance

The information is complete and not distracting; this is clearly a photographer's (as opposed to gadget lover's) camera.

Top

We have to keep in mind, however, that no other digital SLR, regardless of price, offers live preview of the image before the picture is taken.

• E-10: 100%
• E-20: 90%

The E-20 also allows to choose between two display modes: "raw" or corrected with brightness and color balance as applied to the image.

Top

Originally I suspected that this shutter is a leaf, in-lens type, like it used to be in most SLR with non-interchangeable lenses. This was consistent with two blackouts in my E-10's viewfinder: at the beginning and at the end of the exposure (after all, in order to open the shutter for exposure you have to close it first, as it usually stays open for viewing).

As it turned out, when the E-10 firmware was upgraded, the first blackout disappeared; it also is not present in the E-20! Both cameras show a blackout only at the end of the exposure (although, of course, the viewfinder darkens at the start, as the aperture is closed down to the working value).

The only way to explain the absence of the first blackout is to assume that the shutter is not only behind the lens, but also behind the beam splitter — unless I'm missing something here. In this case, the remaining blackout could be explained only as a visual effect — although how it is achieved (certainly not by closing the aperture fully), remains a mystery. If anyone knows more on the subject, please let me know: I hate not being able to explain how things work...

• E-10: From 8s (manual) or 1s (all auto modes) to 1/640s
• E-20: From 60s (manual) or 2s (auto modes) to 1/640s (full-resolution, interlaced scan mode)
• Adjustable in 1/3 EV increments
• E-20: shorter times available in the progressive scan (reduced resolution) mode, see below.

There may be two reasons behind this limitation. Leaf shutters (as opposed to focal-plane ones) rarely reach speeds higher than 1/500s, and at higher speeds they suffer significant light losses (as for a fraction of the nominal exposure time the shutter is only partially open). Using a focal-plane shutter in the four-step mode mentioned above would be messy. I suspect that some digital cameras claiming higher "shutter speeds" do not have a shutter at all, using the "electronic shutter" process mentioned above.

The second reason is related to interlaced scan regime of the CCD, discussed below. Faster shutter speeds would be too short to accommodate the time shift between odd and even lines of pixels.

This may exclude the E-10 from, or limit its usefulness in, some areas of photography (like sports). More importantly, the F/11 and 1/640s combination is quite close to the "sunny 16" exposure: in full sun, medium latitudes, expected exposure at F/11 and ISO 80 would be 1/160s. This means the fastest shutter speed is only 2 EV above the average exposure under such conditions. What if it happens to be brighter, or if we want to underexpose? A 4x, neutral-density (or polarizing) filter may have to be used. (Why at least doesn't Olympus drop the gain to get down to lower ISO speeds?)

This camera uses a CCD with interlaced scan, where odd and even lines of pixels are out-of-phase by 1/1000 s or so (remember the Sony Mavica's infamous "field" and "frame" modes?). This means that an "electronic shutter" would collect the image information from every other pixel row only.

Let's eat this frog and call it a feature, said Olympus. And, indeed, in the progressive scan mode the shutter speed range extends all the way up to 1/4000s; additionally, in manual and shutter priority modes you can set the super-high speed of 1/18,000s. This looks good in specs.

The price paid is lower image quality (although the picture is still saved in the nominal, selected resolution). The "jaggies" are quite prominent, especially on contrasty, slanted, lines. Contrary to some reports (and to hints by Olympus), this is not the same as halving the megapixel count; a cursory examination of my samples shows images clearly less acceptable than those from a two-megapixel camera. (Actually, 1.25 MPix would be, I believe, a better estimate.)

Still, this is better than nothing. In some applications (sports, birds in flight) the quality loss caused by interpolation may be more acceptable than that due to fast-moving subject or to camera shake. Consider the progressive scan a thrown-in freebie (it is done entirely in the software, anyway), and don't use it if you find it unsatisfactory. Myself, I'll have to be badly pressed to resort to it.

Top

Some cameras (including less expensive models by Olympus) have up to five-frame bracketing, but this is an overkill. With the accurate exposure metering in the E-10/E-20, three frames is just fine, and having the offsets in 1/3 EV (rather than 1/2 EV) is of some advantage.

In most cases when I have to resort to autobracketing, I set the exposure compensation at -1/3 EV, and the bracketing amount to 1/3 EV, thus getting three frames, exposed at -2/3, -1/3, and 0 EV compensation.

The buffer capacity does not depend on your image storage size and quality settings (full raw images are buffered in any case), except for the progressive scan mode in the E-20 when the number of buffered frames increases to seven (and the sequence speed increases as well).

These numbers seem to indicate that the buffer capacity is 32 MB (E-10) or 64 MB (E-20). Doubling this number would be awfully nice.

• Wireless: included infrared;
• Wired: optional, electronic.

The remote is also useful for long exposures with use of a tripod, to avoid camera shake.

Top

• ESP matrix
• Center-weighted
• Spot

Center-weighted mode just averages the light over the whole image area, with greater weights attached to the center. This used to be the standard on most SLRs until about ten years ago (when matrix systems emerged).

The spot metering is for critical work, when we exactly know which fragment of the image do we want to be shown at the standard (18% reflectance) brightness. In such cases we just measure the exposure at that fragment, lock it, reframe, and shoot.

Overall between these three modes we find everything we need. One possible extra feature would be to allow to lock the exposure and then point the spot at various fragments, reading out the difference between that fragment and the locked exposure. Some film-based Nikons and Minoltas I've used have this feature, and is some cases it comes handy.

Anyway, I'm very pleased with the exposure system. Even difficult scenes (snow, light sources in the frame) are usually exposed well, nicely fitting into the CCD's tonal range. Inspecting the brightness histogram (whether in the camera or during postprocessing) shows that the cameras are great performers here. My digital pictures, taken with the E-10 and E-20, are in most cases exposed better than those taken on slide film with respectable, modern SLRs.

A side-by-side comparison of the E-10 and E-20 has shown that both always come up with exactly the same exposure reading (I've tried this on 20 frames or so), quite impressive calibration.

Top

• Program (full auto)
• Aperture priority
• Shutter priority
• Metered manual

The program understandably, tends towards wide apertures. As the light increases, it first keeps the lens wide open until the shutter speed reaches a "handholdable" speed (1/30s to 1/125s, wide to tele). At this moment it starts changing both shutter speed and aperture at the same rate, until the latter reaches F/4. This value is then kept constant while the shutter speed increases to 1/640s, from which moment any further adjustment is done by closing down the lens.

(In the progressive mode, the E-20 stays at F/4 up to 1/4000s, only then closing the aperture down as needed.)

This usually cannot be done with an image-editing program: once some of the scene brightness range gets cut off from the recorded image and nothing will restore it.

Enhanced sensitivity is useful in low-light situations but comes at a price: greater image noise, quite visible at ISO 320.

I am disappointed, however, that Olympus did not include an option to reduce the exposure index to ISO 40 (or even ISO 20!). This would be useful in full daylight situations, when we still want to reduce the depth of field by using wider apertures. With the top shutter speed of 1/640s, most daylight shots use the aperture of F/5.6 or F/8; being able to open the lens a stop or two wider would be awfully nice. As opposed to some of my other gripes, there seem to be no technical reasons behind this one — just an oversight.

• Auto
• Seven presets
• Quick reference

For perfectionists, the "quick reference" setting allows you to put a sheet of white paper in front of the camera, and say "this is what I mean to be white". The camera will then compute and store the while balance, which now can be recalled as a preset. Very nice.

The preset values are: 3000K (incandescent light), 3700K (same, warm tint desired), 4000K and 4500K (roughly: fluorescent light), 5500K (sunlight), 6500K (overcast), 7500K (open shade). This should be enough for any use.

Top

Obviously, an active system (casting a beam of infrared light on the subject) will perform well in low-light situations and with subjects lacking vertical detail, necessary for passive systems to work well. (Actually, in such cases the camera will use only the active system.)

This should, in principle work very well, and usually it does. The E-10/E-20 has, however, some problems focusing in low-light situations (at EV4 or below). Worse, it will often confirm the image as in-focus while it is hopelessly off the mark. After some time you develop the habit of verifying the focus in the viewfinder and giving the autofocus another chance if you find it confused; this will save you some wasted frames.

A major complaint: the camera uses a single, centrally located, AF sensor, responding mostly to vertical lines. This is at least two generations behind the modern SLR AF technology. You have to remember about it when taking pictures of off-center subjects, and you may have to tilt the camera if the subject displays predominantly horizontal patterns. At least a set of three AF sensors with at least the center one being "crossed" would be a reasonable expectation in a camera of this class and price. Just so-so.

On the positive side, the focusing is done with a motor built into the lens (like in Canon AF SLRs). It is very fast, and the motor is barely audible, much quieter than in most 35 mm cameras.

• 0.6 m (24 in.) to infinity in "regular" mode;
• 0.2 m (8 in.) to 0.6 m in "close" mode.

At the minimum focusing distance (which, as it turns out, is measured by Olympus from the front of the extended lens), the field of view is about 80x60 mm for the longest lens setting, and 232x174 mm for the shortest (my measurements, estimated ±2% accuracy).

What I find quite irritating though, is the necessity to switch between "regular" and "close" focusing modes at about 60 cm distance (mercifully, there is a slight margin, maybe 10 cm, around that value). This stems, I suspect, from the inability of the infrared active system to work at close distances (when set to the "close" range, the camera uses only the passive TTL system).

There is a way around this: set the focus mode to the non-blinking attachment symbol. This disables the IR pre-focus; focusing will be a slower, but without a need to switch back and forth.

Mercifully, switching between focusing modes does not require the use of the LCD monitor, and can be done quickly, without even looking at the top control panel.

Anyway, the number of steps (depending on the focal length in use) is large enough to make the question rather irrelevant.

Top

• 0.6-6.3 m at 35 mm
• 0.5-5.2 m at 140 mm

The red eye does not seem to be a problem, because the flash pops out away from the optical axis. The popping is triggered by pressing a button.

The difference in range is explained by difference in the maximum aperture at both ends of the focal range. Despite the manufacturer's claim that the flash exposure would not work properly below 0.6 m, I was getting well-exposed pictures at less than half of that distance, as long as I was avoiding the shortest focal lengths (lens shadow).

In such cases you can have the flash go off at the beginning, or at the end of the exposure. For stationary objects this does not make any difference, but for moving ones it does, as the "streaks" will precede or follow the sharper, flash-generated image, respectively.

For unknown reasons, a forced (always on) flash is available with red eye reduction only in shutter priority and manual modes, but not in others.

Top

• E-10: 4.1 million (as given by Olympus), really 3.76 million or 3.59 megapixels;
• E-20: 5.0 million (as given by Olympus), really 4.92 million or 4.69 megapixels.

(In computer terminology, "mega" means 2^20, i.e., 1048576, not one million; camera manufacturers ignore this distinction, as it makes the specs look less impressive.)

The good news for E-20 owners is that the pixel count gain (as compared to the E-10) is not the 25% expected from the 5-to-4 ratio, but rather 31%. This, in addition to bragging rights, translates into a 14% increase in the print size.

For example, if you were happy with your 9x12-inch prints from the E-10, you will be equally happy with 10.3x13.7-inch ones from the E-20. The difference, if not dramatic, is noticeable.

(This comparison assumes that the bottleneck in picture quality is the pixel resolution, not the lens itself. This seems to be the case here, as the E-xx lens seems to be nothing short of outstanding.)

Importantly, most pictures would benefit from some slight cropping: you are not always able to compose your picture critically in field conditions. The extra 14% gives you some freedom in improving the composition or straightening not-too-vertical lines.

(The E-10's CCD, shown above, is made by Sony; so is probably the one used in the E-20.)

In at least one Web article I found the E-10's image size given as 10.2x13.5 mm. This is obviously wrong, based on an assumption that the image frame diagonal is equal to the diameter of the CCD sensor (2/3 of an inch, or 16.9 mm); it is also inconsistent with the equivalent focal length data given by Olympus. Trust me.

According to the Electronic Imaging Primer by Edmund Industrial Optics (no longer posted on their site), a 2/3-inch sensor has the active area of 6.6x8.8 mm — these numbers are within about 1% from mine.

Top

Uncompressed TIFF files are huge: above 11 MB in the E-10, 15 MB in E-20, full size. JPEGs are much smaller but lose some information in the compression process, depending on the degree. TIFF and JPEG files can be read with any image-processing software.

Now I'm shooting most images at 1:8 (good enough for nice 8x10 prints), only for some applications switching to 1:4, and there is still the lowest, 1:2.7, compression left in case I really need it.

We have to realize that all files at less than maximum resolution are created by some way of averaging a number of neighboring pixels of the sensor. This, most probably, includes some interpolation. (Having a closer look at how a CCD sensor works, however, we have to realize that this interpolation is done even in the "native" resolution: each photosite responds to only one of the RGB image components, ant the two others have to be computed by interpolation from the neighbors anyway.)

Therefore I had to change my initial guess (originally published here) that an image in, say, 1600x1200 pixel mode has to be of slightly lower quality than such an image taken with a camera with a "native" 1600x1200 resolution. This depends on the particular algorithm doing the interpolation: if the rescaling is done from original RGB photosite readings, the result should be better; if it is done from the full-color RGB image (i.e., one first interpolated to the native resolution), the result will be slightly worse. I suspect Olympus uses the first solution; my experimenting with the C-5050Z seems to support that.

E-20: Contrary to what at least two "big" review sites say, the pictures taken in the progressive scan mode are not limited to the "Half" size. The image (consisting of every other scan line of the CCD, thus losing half of vertical resolution) is interpolated to (and saved in) the current resolution, as defined by the SHQ, HQ, or SQ setting.

Switching between modes (referred to as TIFF, SHQ, HQ, and SQ) is done without use of the LCD monitor, using the "press and turn" approach standard on the E-10/E-20. Recording in the RAW format needs a separate setting from the menu.

The three JPEG modes can be defined by the user, each as any combination of pixel size and compression. This is done via the menu system on the LCD monitor.

Usually I have SHQ set to Full/1:4, HQ to Full/1:8, and SQ — to Half/1:8, the last one is still more than good enough for decent 4x6 or 5x7 snapshot prints.

• SmartMedia (up to 128 MB at present);
• Compact Flash type I or II (any size); the MicroDrive is officially supported only by the E-20.

One SmartMedia card (32 MB) is included with the camera; most users will want to buy a 128-megabyte one right away, and also a large Compact Flash card. Trust me.

After more than a year of using the E-10, I'm still not sure whether two card slots is a good idea. Maybe shaving 6-7 mm off the camera depth instead would be better? Anyway, I've got used to this, and have two cards, 128MB each, inserted all the time. With the 1:4 compression ratio, this gives me about 150 exposures, not too bad, especially that 1:8 seems to be fine for most purposes, and this doubles the capacity. For the E-20 I've got a 128 MB SmartMedia, and a 256 MB Compact Flash, having decided against the MicroDrive (higher power consumption, moving parts).

With SM cards, the controlling circuitry is a part of the device which uses the card, as opposed to CompactFlash, where it is a part of the card itself. This, intended to make the SM cards smaller and less expensive, may have a limiting side-effect: when SM cards with higher capacity show up on the market, an older camera may not be able to handle them.

Top

• Four AA cells: NiMH rechargeables recommended, or
• Two CR-V3 lithium battery packs, or
• Two RCR-V3 rechargeable Li-Ion battery packs, or
• Rechargeable lithium polymer battery pack in a handgrip

Olympus recommends AA alkalines for emergency use only. Still, being able to use AA alkalines (although they go very fast!) can be a lifesaver. Try to get a Sony InfoLithium replacement in a general store in Hell Hole, Arizona!

My initial investment in sixteen NiMH batteries and three excellent MaHa chargers from Thomas Distributing paid itself back in three months. (I keep two chargers at home in the States, and one in Poland, where I visit often.)

The lithium CR-V3s (a pair is included with the camera) have a 10-year shelf life), so you may put them into a drawer for some emergency. Note that Olympus warns against using lithium AA batteries in the E10, as they may overheat and damage the camera. (They are OK for many other cameras, including the Oly C-x0x0 series.)

The rechargeable lithium polymer pack, available from Olympus as B-10LPB, requires a special battery holder (B-HLD10, optional) and charger (B-20LPC). In exchange for the money spent ($400 or so), you get a longer-lasting battery which also is smart, reporting its status to the camera computer.

New in 2004: In a recent development, a Taiwanese manufacturer came up with a rechargeable Li-Ion RCR-V3 battery, with the form factor similar to that of the CR-V3: two AA's side by side. While the claims of CRC-V3's delivering more energy per charge than NiMH are grossly inflated, their self-discharge rate is much lower.

In spite of high prices (a pair of RCR-V3's will cost you $50 or more), this may be an attractive option if you keep your camera inactive for longer periods of time.

My C-5050/5060 section contains an article on the RCR-V3, including some test results.

• 200 shots with NiMH
• 500 shots with CR-V3

Top

Top:

• Command dial
• Rotary mode knob
• On/Off switch
• Flash mode
• Size/compression
• Memory card selection
• Control panel with...
• Backlight switch

Front:

• Shutter release
• Quick reference white balance button

Back:

• Command dial
• Exposure lock
• Monitor on/off
• Menu on/off
• Four-way cursor
• OK button for menu (also ISO)
• Protect, erase image and info buttons

Left-hand side:

• Auto/manual focus switch
• Normal/close focus
• Expo. compensation
• Metering mode
• Drive mode
• Flash pop-up button

This side also hosts power, TV, USB, remote and flash terminals, and the LCD monitor release lever.

Don't be intimidated with large number of buttons listed below. It is better to have three buttons than just one, performing three different functions depending on the context. As compared to the Camedia 20x0/30x0 series, the number of buttons has been almost tripled, yet the operation is greatly simplified.

The design follows one principle: to change any setting, press the corresponding button and then turn the command dial until the desired setting is shown in the LCD panel at the top.

Thus, for example, to change the storage mode to SHQ, press the storage button (you have to remember its icon, though), and turn the wheel until "SHQ" shows up in the panel. To switch to the close focusing mode, again, press the focusing mode button and turn the wheel: a small flower symbol will show up in the panel. You can do it even while looking into the viewfinder. The manual sometimes becomes boring, trying to describe these operations step-by-step.

Both command dials can be used interchangeably with any button. Additionally (without any button being pressed) they control the shutter or aperture in the respective automation modes. In the manual mode, though, one controls the shutter speed, and the other — aperture. This is like in most modern SLR cameras.

The mode knob switches between various operation modes, including four exposure metering ones. The On/Off switch is placed on a rotating collar around it.

The release button is very soft, without the clear "half-press" resistance. This led me more than once to an accidental release. On the other hand, once you get used to it, you can trip the shutter really softly, thus minimizing camera shake.

The cursor cluster and OK button are used with menus or while previewing previously taken pictures; in either case when the monitor is on.

Additionally, if the OK button is pressed and held down when the menu is not active, it allows you to change the ISO setting by turning one of the control wheels.

The [Info] button at bottom left has a dual functionality:

• In the shooting it mode toggles the shutter/aperture and distance information shown at the bottom of the screen, duplicating the top LCD display.
• In the playback mode it allows to show picture data (exposure, compression, size, etc.) and the brightness histogram. Pressing it toggles the histogram on/off (only if histogram display is enabled from the Settings menu), while holding down and turning the wheel changes the amount of image data overlaid on the display (none, basic, full). A somewhat inconsistent implementation, I would say, but this is a minor complaint.

Note: contrary to some early reports, the histogram display (which I find extremely useful) can be activated only in the playback mode, after a picture has been taken.

The most frequently accessed buttons nicely fit under the left or right thumb (at least when the camera is held horizontally); it is obvious that much thought went into the right placement.

The only setting for which I sometimes have to access the back LCD display is the flash intensity adjustment. I like using the built-in flash as a fill-in in the daylight, and being able to adjust it externally would be very nice.

• Record mode: 12
• Playback mode: 4
• Setup mode: 6 (E-20: 7)

The menus are usually just one level deep. In the record (shooting) mode, there are only three camera functions which you activate from a menu: time-lapse (interval) sequences, and autobracketing. (As a matter of fact, I would be even happier if they were moved out of the menu system.)

Top

• USB
• TV video

The USB interface is of the newer, "storage class device", variety. This means that when the camera is connected to the USB port of your computer with a proper cable (supplied), its memory will be visible as an external disk drive. Moving files is fast and simple, and does not require installing any drivers or other software. (I'm assuming that you're running the Mac OS, Win98SE, ME, 2000, or XP).

• Sharpening: Soft, Normal or Hard
• Contrast: Low, Normal or High
• Saturation: no adjustment

"Soft" and "Low" correspond to no sharpening and no contrast enhancement, respectively. This is what the advanced user would use, doing all postprocessing on a computer (with PhotoShop or Photo-Paint). Initially I left the settings at "Normal" and forgot about them; now I'm using "Soft/Low" or "Soft/Normal" settings all the time.

Additionally, having a color saturation adjustment would be nice, although not essential.

The Olympus default setting (including saturation) are on the conservative, "less-is-more" side. At the first glance, more strongly manipulated images look more pleasing, and that's why most of the consumer-level cameras apply it more aggressively. On some occasions, however, this leads to undesired effects: highlights are getting blown out (contrast), white artifact outlines appear at edges (sharpness), or detail is lost in strong colors (saturation).

These losses in image quality cannot be corrected in postprocessing; therefore it is safer rather to under-apply the image manipulation, adding only as much as needed when the image is postprocessed. I'm glad Olympus chose its approach as it did, aiming the E-10/E-20 at the more advanced market.

• Audible (imitation SLR sound)
• Visual (viewfinder blackout)

See Shutter above for some discussion of the finder blackout.

This takes a while (up to one minute); during that time the image buffer is not accessible, i.e., you have to wait in order to take the next picture. Obviously, the sequential shooting and autobracketing modes are disabled when the noise reduction option is activated (which is done from the LCD menu).

The noise reduction can be enabled (add will go through the whole long and painful process) at any shutter speed, except that the results seem to be discarded if the camera decides that the exposure is not long enough (1/2s or so). This is not very smart.

More, it looks like the threshold depends also on aperture used. For example, at ISO 320 and 1s, the NR kicks in at F/2.4, but it does not at F/11. This means that the photographer in some cases may not really know if the noise reduction was actually applied to the picture just taken. (If you think that any of this is described in the user manual, you are wrong: these guys were too busy trying to teach you how to insert the batteries.)

I found this feature quite cumbersome and confusing to use. I'm just ignoring it in the E-xx cameras. It just sits there, taking one position on the mode dial.

• Sound annotation
• Digital "zoom"
• "Movie" mode

• Camedia Master 2.5
• PhotoShop 5.5LE (E-10)
• PhotoShop Elements (E-20)

Top

The lens is sharp (hardly any difference between center and corners, quite amazing), and has the flare well under control — with the help of good internal finish and the deeply sculpted hood. The distortion at the wide end will be visible in some applications, however.

Based on my preliminary observations, the amount of in-camera postprocessing applied in the E-10 and E-20 (even at "normal" settings) is less than that in most other digital cameras (including Oly's 3000 series), leaving more room for image manipulation with external software. This is what I would expect from the professional-class camera, although at the first glance it may seem less pleasing than images from many lower-end cameras.

After some experimentation I decided to avoid any in-camera postprocessing in most of the situations, setting sharpness to "Soft" and contrast to "Low". The default "Normal" settings may result in more pleasing snapshots straight from the camera, but the images at "Low/Soft" give better results when submitted to postprocessing on your computer.

A note on comparing images between various cameras: some Web sites have a nice collection of comparison images. The caveat is that often these are not images of real objects "at large", but shots of 2-dimensional, previously taken pictures of such objects. While some aspects of the image quality can be judged from such images, many (like handling of glare, high-contrast transitions, white balance, and more) cannot.

Top

At this moment (mid-2002), the E-10/E-20 line still faces no competition in its class; well, it is a class of its own in the two-dimensional plot of performance and price.

Olympus seems to have caught other camera makers with their pants down, and they were not able to catch up in the following year.

Some offerings from Canon, Nikon, Fuji, and Sigma, come close, but on one hand, they cost significantly more (if you include a comparable lens), on the other — they use standard, 35-mm format lenses.

Some people may consider the latter an advantage. It would have been one, possibly, if these cameras had the image sensor of the size of a 35 mm film frame. Their sensors, however, are about 1.5 times smaller (linear size), and this means that using a 24 mm wide-angle lens gives us the equivalent focal length of 36 mm, hardly an impressive wide-angle performance.

Moreover, every lens involves some compromises. Among others, the sharpness at frame edges can be improved at the expense of that in the center, and both are judged in terms of the confusion circle, usually expressed in terms of the linear frame size. Now, for the 35-mm cameras adapted to digital this value should be 1.5 times less than for film cameras, but I don't think it is. Need I say more?

Of course, I expect these makers to come up with some lenses specifically designed for that breed of cameras. They will be, however, still facing a number of "legacy problems".

As one of my friends used to say: "If you add a big engine to a scooter, it won't become a racing car. It will still remain a scooter, with a big motor attached."

In the price range just below the E-xx there are some nice models out there made by Minolta, Sony, Olympus, and Fuji, but all of them are crippled by one main shortfall: lack of the true, optical TTL viewing. An electronic TTL viewfinder will not replace a true optical one; try them side by side and you'll know what I mean. Please don't send me emails to the contrary.

Top

The bottom line

With the E-10/E-20 Olympus brings a professional-level camera within reach of an advanced amateur — the one who has to have the best (and who can afford it). This is a strong attempt at excellence. The cameras are capable, highly flexible, easier to use than most, and, most importantly, produce high-quality images.

Choosing a tool to be used in your favorite pastime is a highly personal thing. Almost every camera has some good points, and all have some shortcomings. The trick is choosing a model with pluses making your photography better and more enjoyable, and minuses you can either work around or get used to. While I'm highly satisfied with the E-xx cameras, and the results they provide, your mileage may vary.

Keeping the above in mind, here is my personal and, possibly, opinionated recommendation: if you can afford the E-20, go for it. If not — there is no excuse for not at least having a serious look at the E-10, with the image quality high enough to guarantee that the camera will not become obsolete in the next few years.

I have been using an E-10 since '2000, and I like it even more now than then. Having voted once with my own, hard-earned money I didn't hesitate to do it again for the E-20, and yes, I'm glad I did.

Top

Web resources

Reviews

Anyone can grab a camera, run around with it for two days, then rephrase manufacturer's press release and specs sheet, throw in some fluff, and call it a "camera review". Unfortunately, that's what most computer magazines are publishing. Majority of their reviewers seem to have no previous experience in photography (and not too much in computers, either).

We are lucky, however, to have a number of Web sites with people who know what they are writing about (well, at least most of the time). Don't look for the big "consumer sites"; these are not very useful. Stick to specialized digital photography ones like these listed below, run by dedicated individuals.

• E-10 and E-20 at Phil Askey's Digital Photography Review
• E-10 and E-20 at Jeff Keller's DC Resource
• E-10 and E-20 reviewed by Dave Etchells at Imaging Resource
• E-10 and E-20 at Steve's Digicams by Steve Jenkins
• E-10 and E-20 at megapixel.net
• E-10 review (quite technical!) by Alex Tsikulin at Digit-Life.

User pages

These pages, run by photography buffs like myself, can provide quite a lot of useful information from people who actually use the equipment on a day-to-day basis. They are not frequently updated, but still worth visiting.

• Dave Weikel's E-10 Page: TCON-300, FL-40 and the LumiQuest Pocket Bouncer, night shots
• Bo Lorentzen's E-10 Page: shooting techniques, infrared, a good collection of E-10 links
• Bron's E10Pro Site: tips and techniques, image samples, message forum.

Olympus-sponsored pages

Olympus presence on the Web is just an embarrassment; don't waste your time trying to find any information there.

There is one exception: the Olympus Japan site, with extensive sections on the E-10 and E-20. The only minus is that all this is in Japanese. Still, with help of the Altavista's Babelfish translation engine I was able to find there quite a lot of information not available anywhere else. An entertaining byproduct of a visit there is the Babelfish translation itself, with quite a few unintentional comical effects.

Top