whichlens.com - Lens Reviews

By JonHuang

There are a lot of things more expensive then photography. Yachts. Flying. Custom cars. Politics. There are a lot of things cheaper too--a rear lens cap costs $6.99, when it's no more complicated than the screw lid to a peanut butter jar. You know it; slap the word "photo" on something and it costs five times as much. Add "professional" to double. Add "digital" and roll a 6 or greater to quadruple word score!

The point of the matter is, there are non-photographic things that can be adapted for photographic uses. Here are my favorites.

Cleaning





1. Microfiber lens cloth
   A thick, fluffy but lint-less cloth is the basics in lens care. But why settle for an expensive 6x7 square when you can get a towel meant for electronics for a little bit more (or often, the same or less). Cut with scissors. If you really want to save, go for the giant car towels.

   OR







2. Eclipse cleaning solution
   Please please please don't use tap water to clean your lenses. You'll leave ugly mineral deposits that can solidify on the glass. Alcohol can cause fogging and doesn't lift oils. Household glass cleaners will damage the anti-glare coatings you paid so much for. Eclipse cleaning solution is the photographic solution, but the photo industry isn't the only one that has had to clean coated glass!

   OR







3. Sensor cleaning
   I'm not going to explain this one. Suffice to say that a $100 service at your local camera store can be done with a cut off plastic spoon, a rubber band, some cleaning solution, a pec pad and a brave heart. Link.







4. Blower
   The first step in cleaning is the blower (DO NOT use compressed air--it can damage components with too much force or chemical exhaust). But who you kidding, Giottos? This isn't a fancy tripod you're selling us, it's a baby's snot-sucker.

   OR





Bags, cases, and storage





5. Tripod bags
   Tripods are funny shaped things; you can't just stick them in your backpack, that's for sure. What about your golf bag? Too many pockets? Try a baseball bat bag.

   OR

   OR





6. Camera Bags
   Frankly, I'd recommend buying a good camera bag. It's worth it. But if you're really, really cheap I'd tell you that the insulation on soft-sided 6/12/18-pack cooler totes. As a side benefit, a lot of them don't look like please-steal-me-camera-bags. (I found one that looks like a purse below). If you're packing a lot, try a diaper bag--no one will steal that!

   I found these at amazon. They're stylish, but rather expensive. Since we're working the ugly and cheap, I suggest your local target--I bought a clutch of these to hold studio lights at about $3 each.

   OR

   OR





7. Drybox
   Way too many people keep their lenses on the bookshelf. That's a good way to ruin them. Aside from dust, pets, and children playing telescope, the major killer of lenses is mold. Eats the coatings, secretes an acid that etches the glass, making them unrepairable. I keep all my glass in airtight boxes with a silica air dryer. My boxes don't say "Pelican" on the side though--they say things like "25 LINKED SHELL FOR CANON" and "50MM NATO COUNT 200."

   Manufactured for the US government to be hardy, waterproof, and indestructible, Ammo cans come in all sizes and shapes and cost 10% of Pelican boxes, which were manufactured by private industry to be hardy, waterproof, and indestructible. Both are crazy heavy though.

   The most common size you'll see for sale is the 50mm, which is about the size of a few lunchboxes stacked together. If you're buying the large ones, I recommend your local surplus store--shipping can be killer on these. Keep in mind that Ammo cans usually open on one of the narrow sides (like a trash can, not like a suitcase).

   You'll also need silica gel packets. You can buy them online, or you can start saving the ones that come in cracker boxes, dry seaweed, etc. They look like little packets with tiny pink balls in them (they turn blue when they're saturated, heat in over to reactivate). If you pull a packet and it feels like it's full of sand, throw away. That's a different kind of desiccant; one use only.

   OR

   +




Printing





8. Inks
   Don't use generic inks. They have a smaller color gamut, low availability of custom profiles, and will make your prints less than the best they can be. Just pay the outrageous $1 a mL extortion and write to your congressperson.





9. Paper
   Kirkland professional smooth gloss photo inkjet paper, available only at costco, is an excellent deal. Costco used to sell deeply discounted Ilford inkjet paper, but stopped when they started selling their own Kirkland brand. Ilford is made in Switzerland. Kirkland paper is made in Switzerland. People swear that it's the same (except for a slightly different core?), draw your own conclusions. It's under $20 for 100 8x11s.





Misc





10. Monopod
    I don't recommend that you take a stick or cane and shove in a 1/4" threaded screw. That would probably work though--did you know that your tripod mount takes the same standard screws you see at the hardware store? Lots of possibilities there.

    Anyway, if you're looking for mobile stability on the cheap, try picking up a 1/4 inch eye bolt and filing off the tip. This screws into the bottom of your camera, making a ring to thread a goodly length of rope into. Tie a loop at the end of the rope and step on it, pulling the rope taut. This should make a triangle between the rope, your body, and your forearms (brake your elbows against your body). With some practice, this should give you a bit more stability. I'll MSPaint a picture sometime.





11. Remote triggers(canon)
    If you're using a Canon 300D or 350D, your cable release trigger is a standard 2.5mm audio jack. You can roll your own cable release, or use certain cheapo handsfree sets. Or even your PDA or universal remote, for the wireless.

That's it for now. I may do a future article like this about substitutes for the photo studio environment. Happy shooting.

An E-TTL II review by jonhuang

Table of Contents

Photos and Introduction
Flash Terminology
Feature Comparison
Flash Coverage at 75mm
Autofocus assist
Bounce Flash
Flash Power Measurements
Studio Tests of ETTL exposure
Consistancy Analysis
Outdoor Fill Flash tests
Ergonomics
Timing Notes
Conclusion

Photos

I’m a big fan of third-party gear. More often than not, you can get nearly same performance—sometimes, better performance—at less than half the price. Sometimes, you don’t.

Sigma is the most prolific of the third party vendors in the Canon EOS system. They’re especially known for their wide-angle lens options, some which were the only offerings in their class. But while there are many third party lenses available for Canon bodies, there are very few reverse-engineered implementations of the modern E-TTL system. For most consumers, there’s only one worth considering:

...which is pretty much a direct clone of the old canon flagship flash, the 550EX Speedlight, only recently (barely) surpassed by the 580EX.

This review is intended to do a side by side comparison of these two flash units. In particular, there are persistent rumors that the DG Super has inconsistent results from the 550EX and is prone to underexposure, a grave charge when working with the limited exposure latitude of digital or slide film. We'll try the flashes under some controlled circumtances.

This review will also give my personal impressions from having used each for a number of months. It will not cover the remote slave functions or the stroboscopic setting, as these are uses for which the average user will not often see.

Incidentally, the rubber bands and Velcro (on the Sigma) in the picture are for me to attach bounce modifiers and cards to. Not Included With Product.

Terminology

Some terms that will be used throughout the review.

Coverage

The area covered by a flash’s light. This is usually given in lens-equivalents. e. g. A 50mm flash coverage lights up everything seen by a 50mm lens on the same camera.

Flash zoom

Modern flashes have a motorized zoom that uses a fresnel lens to focus the light from the flash. When you’re using a telelphoto, the flash zooms in, when you’re using a wideangle, the flash zooms out. This increases the range of the flash by only covering the area in the photograph. Both of our the flashes today have motorized zooms and a pull down diffuser that increases coverage to 17mm.

E-TTL or ETTL

Canon’s current flash metering system. All the digital SLRs can only use this system that uses a short “preflash” pulse of light to meter the scene.

ETTL2 or ETTL II

A revised ETTL algorithm used in the 20D, Rebel XT, 1D mkII, and 1Ds mkII cameras. Mostly the same as ETTL, but takes focus distance into account. This review was written using ETTL2.

Flash Compensation

The default power of the flash can be adjusted in 1/3rd stop intervals. This is the flash equivlent of exposure compensation.

Features

	Canon 550EX	Sigma EF-500 DG Super
Guide Number	138' at 50mm	132' at 50mm
Coverage	17-105mm	17-105mm
Power	Full to 1/128 (8 steps)	Full to 1/128 (8 steps)
ETTL	Yes, native	Yes, reverse-engineered
Bounce	Yes	Yes
Swivel	Yes	Yes
Zoom	motorized (fast)	motorized (slow)
Recycle Time	4-8 seconds	4-6 seconds
Power	4x AA batteries	4x AA batteries
Dimensions	3.2 x 5.4 x 4.4"	3 x 5.5 x 4.6"
Weight	14.24 oz	11.8 oz
Remote Slave	Yes, native	Yes, reverse-engineered
Master Mode	Yes, native	Yes, reverse-engineered
Stroboscopic Mode	Yes	Yes
Power Save Mode	Selectable	Yes
High Speed Sync	Yes	Yes
Second Curtain	Yes	Yes
Autofocus Assist	Yes, 45 point	Yes, center
Built-in diffuser	Yes	Yes

Feature wise, the flashes are nearly identical. A number of the differences given above are pretty much negligible—the power difference is tiny; Canon’s native implementations are most probably the same as sigma’s, and the sizes are within a few cubic inches of each other. Keep reading!

Controlled Tests

Never trust the box specs. Have you ever had a battery last its specified interval or cared about how much a camera weighed without batteries? As far as I’m concerned, the only box specs that have never let me down are the expiration dates on yogurt.

With the usual caveats about “different units may vary,” “I’m not an expert,” and “these results may be totally wrong,” I slotted four fresh batteries into each unit, grabbed my 8x10 kodak color grey card and did a few consumer report-esq tests.

Coverage at 75mm (120mm equiv)

The first set of tests tested the evenness, consistency, and color of the light and flash metering.

Tests were conducted with a 20D at ISO100, f7.1 daylight white balance. A Tamron 28-75 lens with the zoom taped down at 75mm was aimed at an 8x10 grey card placed just far enough to fill the frame, as was the flash (automatic zoom, E-TTL Evaluative, Flash compensation 0). This situation is fairly typical for a head shot in a darkened room.

The card is clamped down, the camera affixed to a tripod. In this situation we expect the flash to illuminate the card evenly at about 13% grey.

No flash (f7.1, 5s)

Here you can see the card without a flash, shot at f7.1 aperture priority (5 seconds). The histogram will show that the camera exposed the card correctly—that’s what we want to see in the next two exposures. Gaussian Blur of 50, desaturated to remove reddish cast from long exposure.

Canon 550EX (f7.1, 1/250s, +0fc)

The 550EX. Even exposure, minimal vingetting at the bottom, a white balance that’s a little bit colder than sunlight (normal). Overall, a pretty good result.

The sigma, right off, displays a disturbing amount of light falloff on the bottom of the frame. The overall exposure is correct, but a result like this would make an uneven exposure (though this may not occur in real-world situations or when the subject was farther from the flash—see the studio trials for more details).

Color wise, it’s pretty much identical to the Canon. Since both are Xenon-strobe based, as are virtually all modern flashes, the color temperature should be consistent, barring discoloration of the plastic lens elements.

Let's look at this a little more in-depth:

Here are the histograms from the three exposure tests. You can see the ideal exposure on the No Flash, long exposure graph on the left. A single color, totally neutral grey, placed at about 13% grey. The canon and the sigma both have a slight blue bias, and peaks clustered around the 13% mark, but the Sigma is more spread out--reflecting the shadows from the lower-area falloff. Also note that the Sigma is also brighter than the canon in some areas; this is due to it trying to maintain a 13% average, which it does. The implication here is that the issue may not be a metering or communication problem, but something physical from the alignment or coverage of the flash.

Autofocus assist

I also observed that while the canon’s autofocus assist light was right in the center of my frame, the Sigma’s assist light flashed higher up on the grey card (though both were adequate for focusing in the dark). For the 20D at least, the sigma might be aiming too high. This would not be a concern if the subject was farther away from the flash, but at the 3 feet or so of my test, this was definitely an issue.

Bounce Flashes

Most of the time, it’s preferable to use the flash to “bounce” the light off a wall or ceiling. This provides a larger, softer light that is overall more attractive and looks less like your flash is powered by a small atom bomb. For the sake of science, I tried a bounce flash with both flashes. Both resulted in even, well lit and identical grey cards. Too boring to post the pictures, move along.

Flash Power

Breaking out my Sektonic light meter, I then proceeded to test the raw power of the flashes. For this section, I clamped the light meter exactly 1 foot away from the flash (not from the lens) and triggered it in manual mode with the test button. Flash zoom was set to the maximum of 105mm. This eventually resulted in a fantastic headache from all the close range flashes, whereupon I lay down for a while before resuming.

Both flashes were adjustable from Full to 1/128th power in 1-stop increments.
The Sigma has a rated maximum guide number of 132
The Canon has a slightly higher rating of 138

The measured timings from 1/128 to 1/1:
Sigma: f9, 13, 16, 22, 36, 45, 64, 90
Canon: f9, 13, 18, 25, 36, 45, 64, 90

Note that both the Canon and the Sigma measured the same maximum and minimum, had a slight (<1/3 stop) variations in rated power at each stop, and were slightly under their rated power range. Neither was perfect, but neither was notably better in this category.

Studio (Dark) Tests

The primary use of flashes is to make light in dark places—surprised? What’s more important is that flash exposures can be strongly affected by the color of the scene—much like taking pictures of snow without exposure compensation results in grey snow, taking flash pictures of snow without flash compensation results in more grey snow.

I live in texas, we don’t have snow. I do have some seamless paper backgrounds, however. Canon on the left, Sigma on the right.

Note the underexposure caused by my decision to not use flash compensation caused dear Ashlee to lose her natural sexy blush in the pallor of underexposure in the first row of flash pictures.

The black background pictures in the first row are the same story, only a little overexposed. It’s interesting to note that Ashlee isn’t very overexposed at all, considering. I’m guessing canon engineers expect a lot of flash pictures to end up with black-hole backgrounds and programmed in compensation for it.

Flash was bounced off the ceiling at 90 degrees in the third row of flash pictures. Better pictures, identical (though still underexposed). A white card attached to the flash with a rubber band might have provided a better fill in those plastic eyesockets.

The important thing to note is that exposure is pretty consistant regardless of what flash is being used. How consistant? Glad you asked.

Consistancy

What the hell is that?! Basically, I took the sigma picture, overlaid it on top of the canon picture, set it to show only the difference between the two pictures, inversed it, and then doubled the contrast.

Basically, it's a map that shows the difference in exposure from the Sigma and the Canon during the white-background trials. Keep in mind that I doubled the contrast, so the actual difference is only half of what you see.

You can see that the Sigma was a bit more underexposed than the Canon on this trial; furthermore, there is a shadow gradient from the top of the frame to the bottom, consistant with the grey card tests. Note that the degree of this shadowing is much lower than in the grey card tests, and can not easily be seen with the naked eye. Still, it's definately there. A concern if you're doing a lot of closeups or extremely color-sensitive work. Otherwise, I wouldn't worry too much about it.

Outdoor Fill Flash

In the bright, point-source spotlight of the sun, photographs often show extremely dark shadows under the nose, cheecks, eye-sockets, and chin. This is all and good if you’re filming a tribute horror flick or yet another oh-my-god-this-old-person-has-so-much-character-in-their-wrinkles shot, but most people look better with a little more detail. “Fill flash” is the use of a weaker flash to “fill in” the shadows with some more detail.

Incidentally, ETTL will automatically reduce the power of your flash by a stop or so when you’re shooting in bright sunlight, since it assume you’re trying to make some fill. A lot of people don’t know this, including quite a few people who should. If you change your ETTL mode to “center-weight,” (not an option on all cameras) this won’t happen.

My girlfriend was busy hitting things in the backyard with a 10-foot bamboo pole, so I scurried out to take some unattractive pictures of her to demonstrate this.

As you can see, the results are functionally equivalent. Note that the “center-weighted” option at a flash compensation of 0 causes the flash to be too strong, making a flat image.

Ergonomics

If you’re shooting outside the studio, ergonomics matter. When using the flash to cover an event, I find myself continuously changing settings:

Exposure compensation to adjust to changing light and fill flash ratios.
Flash zoom for increased range or to create desired effects: long zoom to make a spotlight vignette or a wide zoom for additional fill from nearby walls.
High Speed Synch needs to be triggered when shooting with high shutter speeds.
Second curtain flash for when I’m doing a slow-sync flash.
Manual mode is something I find myself using quite often, so that I can reproduce a consistent setting or not have to worry about lining up the subject on a focus point when hitting the pre-flash.

Having used both flashes in such circumstances, I find that the Canon is unequivocally superior in its user interface. Buttons are laid out logically, and there are only a few modes to toggle through, thanks to a switch that sets the flash into single, master, or remote slave modes.

The Sigma, on the other hand, has a confusing and frustrating layout. Changing modes is especially difficult, as there are a large number of modes—but unlike the Canon, there’s no title on the LCD to tell you which mode you’re in. You have to figure out what mode you are in based on what items are visible on the screen, this usually results in valuable time being wasted while cycling the mode button. Some of these modes are rarely used, like the remote modes; some of them are useless, like the TTL mode for older canons. You have to cycle though all of them though.

Such small touches are common. Battery loading on the canon is easier since all the batteries are in a single line, rather than a 2x2 box configuration. The high-speed sync on the Sigma resets to normal after every shot, a major inconvenience. The power switch on the 550EX can also be put in a “Save Energy” mode so that the flash sleeps after a period of nonuse (the Sigma is always in an equivalent energy save mode).

Timings

I didn’t do a formal study on shot timing, but my overall impression was that the 550EX recharged slightly faster than the Sigma. Other people have reported differently--this is a point that is more conditional on your batteries than the flash in question.

What is painfully notable, a major point for me, is the relative speed and noise of the flash zoom motors. The 550EX is substantially faster and quieter; unlike with the Sigma, I never had to wait to take my shot until the flash stopped zooming. When adjusting zoom manually, I could cycle though the settings to the one I wanted twice as quickly on the 550. If you have a USM zoom lens, you can find yourself aimed, focused, ready--and waiting on the flash to catch up to you.

Conclusion

So which flash is better for me? The 550EX. Which is better for you? Depends. While the 550EX is better in build, user interface, autofocus assist, and a few other things, both flashes perform nearly identically power, features, spread, and ETTL metering—and the Sigma costs substantially less.

Generally, I’d recommend the Sigma to the casual user on a budget or to someone considering it’s price equivalent, the 420EX. If you can afford it however, the 550EX will not let you down.

If you liked this review, please consider buying from our sponsor below:

SIZE
The 350D is tiny. The 20D seems like a behemoth in comparison.
350D Dimensions (WxDxH) : 4.98 x 3.71 x 2.63 in./126.5 x 94.2 x 64mm
20D Dimensions (WxDxH) : 5.7 x 4.2 x 2.8 in. / 144 x 105.5 x 71.5mm

GRIP
I can easily see it being too small for some people. For myself, with smaller than average male hands, it is just the right size. I used to use a Pentax *istD which previously held the crown for smallest DSLR and I found it comfortable. Going from a 20D grip to a 350D, there is a very dramatic difference. If the 20D grip is the right size for you or even too small, then the 350D grip will probably be quite frustrating to use. The 350D grip is shallow and short, you won’t necessarily be able to wrap your pinky finger around the grip. I find the best way to hold the 350D is to wrap four fingers around the camera, and let the pinky support the camera from below rather than try and squeeze all four fingers together on the grip.

WEIGHT
350D – 17.1 oz. / 485g
20D – 24.2 oz. / 685g
The 350D is extremely light. I can easily hold it and operate it with one hand. On the 20D, the controls are a bit too spaced out and the camera is too heavy to use easily with one hand. Of course, this is a minor point, you should be holding your >$1000 camera with two hands just like they taught you in driving school. Both are not so heavy or light as to make them unusable or uncomfortable.

BUILD
Build quality on the 350D is good. Fit and finish are fine, even at normally iffy areas such as the battery door and compact flash door. The surface finish is a slightly rougher grain than the 20D. The quality is on par with the 20D. 20D has the advantage of magnesium alloy body but the plastic 350D can survive a bit of abuse as well. Quite simply, weight != quality. The 350D holds its own.

BALANCE WITH LENSES
With most of my primes, the 15mm fisheye, 35mm or 50mm, the 350D feels well balanced. In fact, such a lightweight configuration gives it an almost point and shoot feel. With larger lenses such as the Canon 16-35L f2.8, the feeling is reminiscent of the Sony F707/717 series. The configuration of a large lens with a small body looks a little comical, but actually is quite comfortable. It just means you are support the majority of the weight with your left hand and balancing it with your right hand rather than equally distributing weight in the left and right hands like you would with the 20D. The 20D is well balanced with most lenses and is also very comfortable to use, if a little bit heavy.

FLASH
The 20D flash is definitely bigger than the 350D. The vertical offset from the lens is pretty similar on both cameras, but I suspect the 350D has a weaker flash than the 20D.

Left 20D, right 350D

BATTERY
The 350D takes NB-2L batteries with a capacity of ~720mAh versus the 20D’s BP-511/512 batteries with a capacity of ~1350mAh. The NB-2L is lighter than the BP-511 and 512. So far, this reduction in battery life has not been a problem, on a fairly hectic 4 hour shoot, I shot 150 pictures in RAW and checked the LCD regularly for preview information. The battery did not indicate it was at low life, so you could expect the battery to last at least a full day of on and off shooting. Changing settings on the 350D requires a lot of referring to the main LCD where information is displayed. At first I thought this would make a big difference in power consumption, but it appears to have had fairly limited effect. The 20D can go for a day or two of on and off shooting and it uses a low power LCD to display most camera information, so it is still the absolute winner in terms of battery life.

Operation:

LCD
The 350D has a rear mounted status LCD rather than a top mounted one like the 20D. It can be illuminated just like the 20D LCD and displays the same basic information, albeit in a slightly different configuration. However, it is does not show ISO or flash exposure compensation on the status LCD, those are shown on the main LCD instead. When changing ISO, white balance and metering mode, the changing states are shown on the main LCD rather than the status LCD. A little different, but not all that troublesome since the two LCDs are adjacent to each other. The main LCDs seem to be identical in construction. Similar brightness and size, around 1.8 inches in size.

Left 350D status and main LCD, right 20D status LCD

CONTROLS
When adjusting settings such as aperture, shutter and exposure compensation, the 20D has the front control wheel where the index finger rests, and a rear spin wheel below where the thumb rests. The 350D also has a front control wheel, but lacks a rear spin wheel. Instead, the exposure compensation button must be held down while the front control wheel is rotated. I prefer this to the 20D’s spin wheel since it allows for easier one handed operation, but ultimately, a second dial on the back where the thumb normally rests is what I really want to see.

On the 350D, Control of ISO, white balance, auto focus mode, and metering mode have each been assigned one of the buttons on the five button cross arrangement. Each can be accessed by pushing a single button. Successive pushes of those buttons will scroll through the list of options available and displayed on the main LCD. Spinning the front control wheel or pushing the up or down buttons once you are in a control screen has a similar effect. Shooting drive mode can be changed by pushing its button to scroll through its options, but it is separate from the five button cross arrangement. Flash compensation can only be accessed through multiple button pushes and navigating through menus on the main LCD. On the 20D, more controls are accessible from one button push. You can access auto focus mode, white balance, shooting drive mode, ISO, metering mode and flash compensation all from one button push followed by spinning either the front or rear wheels to navigate through the options. This method of control is more consistent in design, but when it comes to actual operation of the camera, it doesn’t make too much of a difference.

350D five button cross arrangement

AUTOFOCUS
Autofocus point selection is another major difference of the 350D compared to the 20D. The 20D has 9 AF points in a cross formation. It has a small 8 way joystick to quickly choose the appropriate point. The 350D has 7 AF points which are selected by pushing on the five button cross arrangement to move from point to point. It is a less elegant solution and can be frustrating if you are trying to adjust focus point while keeping your eye on the viewfinder. It is too difficult to operate the five button cross by feel alone.

The 350D also lacks the enhanced central AF point of the 20D. When using lenses faster than f/2.8 and in low light, the central AF point on the 20D is able to AF with greater accuracy and consistency than the not-so-enhanced central AF point on the 350D. Real world tests have shown that the 350D is still very capable of achieving focus lock in low light situations, but it does not have the consistent accuracy of the 20D. In situations where the 20D would be able to get a perfect, quick focus lock, the 350D will only get it right 90% of the time, and may have to hunt slightly before it gets it. With low light and moving subjects, a doubly hard situation, the 350D lags further still. In daylight or well lit conditions, the 350D is on par with the 20D.

VIEWFINDER
The 20D viewfinder is approximately 10% larger than the 350D and slightly brighter. The 350D is still perfectly usable, though it does exacerbate the difficulty of checking accurate focus in low light conditions and making manual adjustments. The eyecups on both cameras are interchangeable, the 20D’s eyecup has a larger rubber cup. Both have diopter adjustment.

Left 350D, right 20D

SHOOTING
The 20D shutter seems to fire slightly quicker than the 350D. The mirror blackout between shots is also slightly shorter on the 20D. The 350D is still very responsive, but a small difference between the two is present. The 20D has a faster, longer continuous shooting mode of around 24 frames at 5fps versus the 350D’s approximate 15 frames at 3fps. I say approximate because the number of continuous frames at 5 or 3fps depends on the speed of the card being used. Faster cards will flush the buffer faster as you are shooting, so the number of continuous possible frames increases. Nonetheless, there is a difference in the size of the buffer and for sports and fast action shooters, the 20D may be the preferable camera.

SHUTTER NOISE
The 350D is significantly quieter than the 20D. The sound has a slightly metallic whirr but it is definitely a more discreet camera than the 20D, whose noisy shutter was one of my pet peeves.

HIGH ISO COMPARISON
I used a heavy duty Manfrotto tripod, shot in RAW and converted to low compression .jpgs (quality 12 in Photoshop CS) then composited them into one final low compression .jpg (quality 8 in Photoshop CS) showing 100% crops. The color space used was sRGB, WB used was flash mode. All parameters such as sharpness and contrast were set to 0. Shot setting was f8, 1/200 using a Sigma 500 DG Super in manual mode to vary the light intensity from 1/128 to 1/8. The lens used was a Canon 50mm f/1.8 prime.
ISO1600
ISO800
ISO400
ISO100
Noise levels are very similar. I think you would be hard pressed to distinguish between the two in terms of noise. Even the sharpness is very similar. One thing worth mentioning though … I had to re-shoot two of the 350D images because there was slight mis-focussing and I never had to do this with any of my 20D test images.

MISCELLANEOUS
The 20D can do a timed mirror-lock up exposure. The 350D can do mirror-lock up exposure and timed exposures but not both simultaneously.
The 20D has a number of custom functions that are not present on the 350D, including the ability to enable ISO 3200.

CLOSING COMMENTS
The 350D Pros compared the 20D:
90% of the functionality (noise performance, shooting speed, responsiveness)
90% of the quality (fit and finish)
Much quieter shutter
For me, better ergonomics. This is very subjective and there will definitely be people who think the 350D is too small. However, if you think the 300D, 20D is too big, or liked the feel of the Pentax *istD or *istDS, then the 350D is right up your alley.
The 350D Cons compared to the 20D:
Smaller, dimmer viewfinder
3fps instead of 5fps and smaller buffer to go with it
Weaker onboard flash
More inconsistent autofocus compared to the 20D, especially in low light, moving target situations. Still, the 350D puts in a respectable performance. 85% accuracy compared to 90% accuracy of the 20D.
No ISO 3200, I rarely used this feature on my 20D but some people might
No AF joystick

General 350D Pros
Excellent 8 megapixel resolution
Excellent responsiveness, zero-wait startup, quick shot to shot
Excellent autofocus in good lighting conditions
Good low light autofocus (just not as good as the 20D, arguably the low light leader among entry level, mid level DSLRs)
Good build quality, no flexing or creaking
Light weight, compact size (this could be a con for some people)
Quiet shutter

General 350D Cons
Form factor is not for everyone. For some people, it is simply too small.
Menu is a little fiddly. Some settings require more button pushes than they should. A joystick for choosing AF point would be very desirable.

What else can I say? If the form factor is acceptable to you, then this is a highly recommended camera. It is capable of great things and at $900, it is a most worthy entry level DSLR. Any questions or comments, please direct them to mark@whichlens.com. If you liked this review, please consider buying the camera from Amazon using the links below.

A little archive of my thoughts during the production of this review. I have incorporated most of these comments into the review already, so consider the text below as “process notes”.

---
12-3-2005 PM –
I shot a fundraiser event last night using the 350D with a Canon 16-35mm f2.8L lens and a 20D with a Tamron 28-75mm f2.8 lens with Sigma EF-500 DG Super flash. The lighting in the room was very subdued and when I wasn’t using the flash, I was shooting at mostly wide open apertures with ISO 800 or 1600. I shot around 150 shots with the 350D and 150 shots with the 20D, all of them in RAW format.

It was a fairly difficult test for the 350D, but I would definitely say it passed. However, I would not go as far as saying it was a 20D in a smaller body. I will list my thoughts in bullet point form. Sorry for not being more organized, I will re-write this review at a later date, but there are people chomping at the bit for this information.

1. The 350D’s battery lasted fine. I shot 150 shots with image preview turned off but checking regularly on the images at various intervals. The battery indicator unfortunately is still the same full life, little life, dead style indicator but at least the camera did not dip into little life territory. The entire event went for around 4 hours and I had the camera on almost the entire time.

2. The 350D’s viewfinder made things a little difficult at times. It is slightly smaller and darker than the 20D’s and in low light, it was impossible to say if auto-focus had accurately locked focus or not. However, the 20D is not significantly brighter or larger, and so I had the same problem with the 20D, just to a lesser extent.

3. The 350D’s auto-focus in low light was better than I expected. I was using the central AF point for most of the shots and had expected a bigger gap in performance between the 20D and the 350D’s low light auto-focus accuracy on that AF point. However, the 350D was not fool proof. There were more misfocussed shots than the 20D, i.e. there was focus lock but the shot wasn’t in focus when it was taken. Also the 350D choked when it came to fast moving objects (kids). However, given the total number of shots and the difficulty of the lighting conditions, I would say the auto-focus performed well, with 90% of the reliability and accuracy of the 20D’s auto-focus (which incidentally isn’t foolproof either).

4. The 350D with a 16-35 2.8L felt fine. I had no issue with stability or comfort. However, as I’ve said in other parts of the review, the size of the 350D suits me well, and better than the 20D.

5. The 350D’s lack of a joystick for selecting focus points was the biggest annoyance for me. I am used to keeping my eye on the viewfinder while reaching with my thumb to select AF point on the 20D’s joystick. My 20D is set to having the joystick change AF point without having to press the AF point select button first. With the 350D, I had to drop the camera from my eye, press a few buttons and then raise the camera up again to recompose which, under the circumstances, was not ideal. Given that I was shooting f/2.8, the depth of field was probably too shallow to focus with the central point and adjust composition. The side points had to be used.

6. The 350D’s responsiveness when it came to writing and viewing the RAWs was fine. I was never held up by insufficient buffer and the RAW files loaded at an acceptable speed for previewing.

7. Switching between ISO on the 350D was a mixed bag. It was convenient to be able to press the ISO button and cycle through ISO settings, but sometimes I forgot to hit SET after I had selected my ISO setting. On the 20D, you go into ISO selection mode and spin the control dial which is a slightly better method. Really though, Canon should be putting an ISO readout into the viewfinder and the status LCD, and that applies to both the 20D and the 350D.

8. On the 350D, having to hold a button to modify the front control wheel from changing shutter to changing aperture was not as big an issue as I expected.

9. The 350D’s discreet shutter was great and attracted a lot less attention than the 20D’s. There were performances during the fundraiser and I felt much more comfortable shooting with the 350D and not disturbing people in the process.

10. Like a klutz, I knocked the 350D a few times here and there. It survived. Plastic is not rice paper.

---
11-3-2005 AM –
I briefly shot at a small event yesterday and am now fairly convinced now that autofocus is not the 350D’s strongest suit. In a low to moderately lit indoor room, it was hunting for focus lock using the central point only. This was with a Canon 50mm f/1.8 in a room that I have photographed in before with my 20D and the 20D definitely did lock under those situations.

Using the light 50mm f/1.8 prime lens was very enjoyable and I am appreciating better handling with this lighter body and a light lens. The shutter button is pitched closer to the horizontal plane which helped, for me at least, to make a smooth motion when I was releasing the shutter.

I also used my 16-35 f/2.8L briefly. If anyone has ever owned or used a Sony F707/717 they will know the feeling I had. The lens overwhelms the body somewhat, but it seems comfortable and actually offers good stability.

The Adobe RAW converter does not currently support the 350D. I had to use Canon’s supplied software instead to convert RAW files.
---