Friday, August 28, 2009
When I first saw the specifications and features for the G11, I was disappointed. I’d already predicted the specs and features for the G11 and my prediction was substantially wrong. After a few days, I became more interested in the G11 and, in the previous post, took a positive attitude about the G11. In this post, I get to vent my frustrations and unfulfilled wishes for the G11.
No CMOS! I had predicted a CMOS sensor for the G11 not because of my expertise in sensor design but because of my faith in Canon and because Canon has built a manufacturing plant specifically for compact camera CMOS sensors. Canon uses CMOS sensors in all its DSLR cameras and I wanted their demonstrated expertise to be applied to the G11. Now I cynically take the opposite view: Canon fully expected to make a high quality, low noise CMOS sensor for the G11 but, in fact, could not do so. As a result, Canon had to hurriedly choose a CCD sensor from an outside source (widely believed to be Sony) and to accept the compromised performance of that sensor in their new G11 camera. (Cowardly Disclaimer: This is my own, personal conclusion and viewpoint. I have no data or information whatsoever to support my belief.)
No large sensor! Along with the CMOS sensor, I also predicted a slightly larger sensor for the G11. Now I think that Canon probably planned to use the same size sensor, albeit with fewer pixels than the G10, all along. I can understand their (assumed by me) reasoning but the G10/11 is increasingly being compared to its Micro Four Thirds (MFT) competitors and MFT cameras have larger sensors than do the Canon G series cameras. I did not expect an MFT size sensor for the G11 but fully expected one that could be advertised as “larger” -- meaning in comparison to the G10.
No HD video! This is the obvious indicator of the compromised sensor. Everyone expected HD video in the G11. Granted, not all users want video but there is an obvious convergence of still and motion photography underway in the digital cameras. The G11 is even being promoted by Canon as a photojournalist’s camera. Come on! The lack of HD video will probably be the first item on anyone’s list of surprises about the G11.
No voice recorder! Apparently this feature, available on my G9, has been omitted from the G11. No voice recorder in a photojournalist’s camera?
No superwide! The 28mme focal length is a reasonable choice for the wide end of the zoom but where is the optional conversion lens? The G11 lens is (apparently) the same as the G10 lens and Canon has not offered a wide angle conversion lens for the G10.
There, now I feel better – and my list of complaints is not as long or as serious as I thought it would be.
Tuesday, August 25, 2009
As it became apparent that August 19 was really THE DAY for the G11 announcement, the rumors could be taken a bit more seriously; however, the most accurate rumor was certainly a disappointment. I was disappointed not only because I’d made my own prediction and now had to account for it but because the G11 did not appear to be the exciting upgrade to my G9 that I’d wanted. I decided there would be no G11 in my future.
Now that the smoke has cleared (well, actually the discussion forums, such as the Powershot forum at DPReview are really heating up), the G11 really does have some features and potentials that interest me. Here’s my personal list – in some kind of meaningful order – as gleaned from the published specs and the G11 manual.
1. Wide angle lens (28mme). Not as wide as I wanted (24mme) but an improvement over the 35mme of my G9.
2. Articulated display – a “real” G series to many people. I liked that feature on my old G3.
3. Wired remote. Although I use the Franiec cable release on my G9, I like the idea of the wired remote – and I already have one!
4. Improved dynamic range and higher usable ISO. Although data and comparisons are not available at this time, I have no doubt that the G11 will have less noise and increased dynamic range than the G9 and, for that matter, the G10.
5. Intelligent contrast correction. This in-camera processing option might help to save some post processing for those occasions when I don’t post process from RAW mode.
6. Decreased shutter lag. Actually this is just implied and is probably only for Quick Shot Mode.
I’m curious about these features but wouldn’t buy the G11 just to get them:
- Quick shot mode
- Registering commonly used menus
Whereas my first impression of the G11 was “Not for me” now I’m on the fence and trying to make up my mind. I’ve gone so far as to put my name on a waiting list but frankly am happy that the G11 is not available for ordering at this time.
Sunday, August 23, 2009
I’m one of the complainers about shutter lag with my G9 – even in manual mode – and, in the midst of a debate on a DPReview forum decided to conduct my own tests. But how to do so? At first, I imagined all sorts of timer circuitry, switches, etc. but since I don’t really know how to do that sort of thing, finally decided to try simply taking a photo of my computer screen. This crude approach actually worked surprisingly well – sort of.
I searched for an on screen timer and downloaded XNote Timer . This is a basic and free version of their more comprehensive XNote Stopwatch. The image above is one of the several hundred shots I took of XNote Timer from my LCD. (I know, I know – sort of sad, isn’t it?)
I was especially curious about differences in shutter lag with and without Image Stabilization (IS), the effect of the half-pressed shutter button (sometimes called pre-focus) and the quickness of all manual mode with IS off, display off, etc.
Both Imaging Resource and DPReview publish shutter lag values but their test results did not completely satisfy my curiosity. According to Imaging Resource, the G9 shutter lag is from 0.57 seconds at wide angle to 0.69 seconds at telephoto. When prefocused, the shutter lag is 0.088; with manual focus (but presumably auto exposure) the shutter lag is 0.58 seconds. According to DPReview, the full press lag is about 0.6 seconds; the half press lag is about 0.12 seconds with the LCD turned on but decreases to about 0.05 seconds if the LCD is off. I don’t have any particular reason to dispute these values.
No, the shutter lag of my G9 was never the 40.63 seconds in the picture. After starting XNote Timer, I took a picture every 10 seconds. (I watched the computer monitor – not the G9 display.) So the example represents the 4th shot and the delay was 0.63 seconds. I usually took a dozen shots (2 minutes worth), discarded any obvious long or short times and averaged those remaining. Here’s what I found.
Base Case (my typical setup): Av mode, RAW+JPEG, display on, evaluative metering, small flex frame focus, shot only IS on, AWB. Full press lag = 0.61 seconds.
Same as Base Case (above) but timed from shutter button half-pressed: Lag = 0.15 seconds! In spite of the crudeness of my method, the improvement in shutter lag was obvious. Additionally, I found that the half pressed shutter button reduced shutter lag time to about 0.15 seconds no matter what options were selected in the menu or function settings.
Base Case but with manual exposure, manual focus, IS off. Lag = 0.23 seconds. This test was done with a full length press of the shutter button. Again, lag was reduced to about 0.15 seconds when the shutter button was half pressed prior to completing the press.
In some tests, turning IS off shaved about 0.1 seconds from the shutter lag but other times there was no measured effect on the average time. All these tests were hand held and perhaps I was more steady some of the time. I’d guess that IS probably adds about 0.1 seconds to the lag.
Interestingly, I found that by turning off RAW and using JPEG only, setting white balance (instead of using auto white balance) and reducing resolution to M3 with Normal compression reduced shutter lag by about 0.2 seconds from the Base Case. This was unexpected but perhaps not really correct because by that time I was getting pretty good at taking the shutter lag test. As I got better and better at taking the shutter lag test, I even recorded some negative shutter lag! That is, I jumped the gun a bit while anticipating the 10 second interval. As my own lag improved, the first series of tests had to be repeated.
So I learned some things about shutter lag. The very best technique is, as expected, to first press the shutter button halfway to obtain focus, exposure, set the IS, etc. Complete the shutter button press on seeing/hearing the G9 confirm that it is ready. Full manual mode (manual exposure and manual focus with display off, IS off) is better than I’ve been crediting it.
Granted, this technique isn’t great but it was an instructive exercise and perhaps will be useful to others. My own reflexes are not particularly quick so these speeds should be easily beaten. Try a few variations and learn more about shutter lag.
Wednesday, August 19, 2009
Canon’s G11 was announced today. The announcement was expected but I was very surprised by the feature set of the G11. I’d made my own G11 prediction a few months ago but didn’t exactly hit the mark. Well, taking a more positive attitude, I did get a few things right: the name G11; same form, appearance, controls but a little larger; same focal lengths and apertures; fewer pixels but higher ISO; highlight recovery option. Perhaps not so terrible after all.
On the down side, I completely missed the mark with my prediction of a 12MP CMOS sensor manufactured by Canon that would be slightly larger than the G10 sensor. The G11 uses a 10MP CCD sensor and, at this point, the manufacturer is not known. The G11 sensor is the same size as the G10 sensor.
Even more surprising to me is that the G11 does not include HD video. Actually, I’m shocked at this omission. It never even occurred to me that HD video would not be in the G11 feature set. The omission of HD video in the G11 is probably a serious marketing error by Canon. (Having said this, I must admit to using movie mode very infrequently on my G3 and G9 but was looking forward to trying HD video with my next camera. Also, I’m well aware that the need for HD video in a compact camera is a much debated topic among photographers.)
The G11 even has the articulated display screen in spite of my prediction that it would not. However, I’m happy to have been wrong!
Other features that I mis-predicted include: launch date (I thought September); price (a good thing!); new lens (apparently is the same as the G10); pixel binning at ISO 12800 (instead of 6400) – sounds interesting; new conversion adapter (will use the G10 adapter); RAW files from AUTO mode.
The G11 may also have some minor tweaks to shutter lag and flash performance. There’s a new mode, “Quick Shot”, that sounds interesting but is not described in detail.
Canon’s launch information leads me to believe that ISO 400 and possibly 800 might be routinely usable with respect to noise. Of course, noise and noise reduction are greatly debated topics and others may very well disagree. With my G9, I was beginning to feel OK about ISO 400 after suitable processing. The G11 would appear to have the potential to go to even higher ISO speeds.
As always, DPReview has additional information and summaries. I’m looking forward – and impatiently so -- to their review of the G11.
Thursday, August 13, 2009
The Canon G9 is not a wide angle camera. The wide end of its zoom lens has a 7.4mm focal length; this provides the equivalent view of a 35mm lens with traditional 35mm film cameras. At the same time, it must be said that, in one way, a 35mm focal length lens was indeed considered “wide angle” for 35mm cameras back in the day when 50mm was considered “normal”. For many photographers, the 35mm focal length lens was a favorite.
Even so, in a small room the G9 photographer might be wishing for a wider wide angle (shorter focal length) lens. Outdoors, a landscape photographer might also prefer a short focal length lens. Canon got around this short coming by offering an auxiliary wide angle lens for the Powershot G series cameras. The auxiliary wide angle lens attaches to the Conversion Lens Adapter. There is also a telephoto auxiliary lens. Until recently, I’ve not had, used or even seen these auxiliary lenses but I now have the wide angle lens, WC-DC58B.
As can be seen, the WC-DC58B is a large lens. This one is attached to a Lensmate adapter but, of course, it also fits the Canon adapter. The lens is threaded the same as conventional 58mm filters for attaching to the camera; however, the auxiliary lens does not include threads for attaching filters. The auxiliary lens almost completely blocks the viewfinder and the assembled camera+auxiliary lens is an awkward package; however, it does work. The lens is marked ".75X" which means that, when attached, the G9 focal length should be multiplied by .75. Therefore, the widest focal length on the G9 becomes 5.6mm (26mm equivalent for 35mm cameras). The WC-DC58B also seems to work on my old Canon G3. (Please note that the G10 includes a wider angle zoom lens and the WC-DC58B and similar products may not work very well.)
I’ve actually already reported on the WC-DC58B in a previous post but still haven’t used it very much. More later.
Sunday, August 9, 2009
The Strobist, (see my "Blogs I like" list) David Hobby, is challenging his followers with a series of assignments in his second "Boot Camp". His current assignment, the third in this series, is to light and shoot a single room for a real estate sale. Although I follow the Strobist regularly, I've not been an active participant in the Boot Camps; however, I did take on this assignment with the picture below.
A bathroom – even a large one – can be difficult to capture with a single shot. This picture was stitched from 12 shots taken with a Canon G9 fitted with a wide angle adapter (DC-58B) to make a 26mm (equivalent) focal length. The lighting was quite a mix: sunlight, tungsten bulb and flash. In fact, the closet had fluorescent lighting but it was turned off to avoid additional complications. The flashes were a conglomeration of Nikon SB-24 (2), SB-26, SB-28, Vivitar 285, and even an old Vivitar 2600. All flashes were gelled with a full CTO to match the tungsten bulbs over the vanity. The (nearly) tungsten white balance rendered the sunlit window a nice blue. The SB-26 was triggered by its own optical slave; all others were triggered by Cactus PT-4 wireless triggers.
At ISO 80, the ambient exposure was 1/3 second at f5.6. Of course, this ambient lighting was very uneven and the tungsten lights were completely overexposed. By adding Stofen diffused full power flashes in the shower and tub areas as well as a bounced full power flash in the room the lighting was made more even. Half power flashes in the toilet and closet areas lighten those shadowed areas. The final exposure was 1/125 second at f5.6.
Stitching those 12 shots was actually the easy part and was handled automatically, albeit a bit slowly, by Photoshop CS4. Most of my time on this project was spent in cleaning the bathroom, moving items from the vanity, setting up lights and then removing and storing those lights and, finally, replacing vanity items.
Saturday, August 8, 2009
As her reward for sitting so patiently while I experimented with fill flash settings, I got this nice portrait for Hannah. This was taken at 1/640 second and f5.6 using the Nikon SB-28 directly mounted on the G9. The background was a bit less bright than the harsh glare of the driveway and, at 1/640 and f5.6, is about a stop underexposed. I removed the camera from the tripod and dropped down on one knee to get a better perspective.
Tuesday, August 4, 2009
Continuing with the G9 flash fixation, this post is about fill flash. Although previous posts are labeled as flash and exposure compensation, fill flash is, in a way, much the same sort of thing.
First, I set up a very contrasty scene featuring the lovely Hannah . A similar situation – although perhaps not as drastic – occurs when an expanse of sky is included in a scene. The G9 was simply placed in Program Mode (P), ISO 80, widest angle, small Flexizone auto focus and evaluative metering. (For uniform comparisons, all pictures herein are the in-camera JPEG.) Poor Hannah! Although she enjoyed the shade and was not squinting at all, the bright concrete driveway convinced the G9 metering system to use 1/1000 sec and f4 to avoid overexposing the scene. What to do?
Fill flash is the answer here. The question is “How much fill flash?”. Rather than deal with theory, classic ratios, etc., I simply turned on the flash and took another shot. Not too bad! On close examination, several interesting changes took place inside the mind of the G9 (remember it was still in Program Mode, etc. but the flash was turned on). First, the exposure settings were changed to 1/500 second at f5.6; those who follow such details (and all aspiring photographers should) will recognize that these settings produce exactly the same exposure as 1/1000 second at f4. This change was made because the G9 limits its little built-in flash to 1/500 second and at 1/500 second the proper aperture was f5.6. Next, the G9 flash pre-flashes to check the effect and then fires at what it thinks is the necessary power. On second thought, Hannah is actually a bit underexposed. More power is needed.
It turns out that the flash picture was taken with the Slow Sync setting turned OFF so I changed to Slow Sync ON. Duh! Same exposure settings and same result. Next I changed to Aperture Priority (Av) mode and set the aperture at f5.6 to avoid confusing myself. Same results. How about Shutter Priority (Tv) at 1/500 second? Same. Not to be denied my right to force my preferences on the G9, I changed to Manual exposure and kept 1/500 second and f5.6 but forced full power to the flash. (Remember that manual exposure mode also means manual flash adjustments.) Same results – that’s when I realized that the little built-in flash was maxed out. With the flash already firing at full power, there would be no point to increasing the flash compensation setting. So ISO 80, f5.6 and a camera to subject distance of about ten feet is about the limit for the G9. (This would be a Guide Number of 56 but I suspect the GN is really about 40.) Time for a bigger gun.
Next, a Canon 420EX was placed directly on the G9. Again, the G9 was put into P mode. The result was definitely an improvement over the built-in flash. Again, several interesting changes were made by the G9 (not by me). The shutter speed was changed to 1/250 second because that is the limit of external flash unless the external flash can handle “High Speed Sync”. At 1/250 second, the correct aperture setting (considering that bright background) is f8. Actually, the G9 did not particularly like either the 1/250 or the f8 and indicated so by showing both in red on the display. Another point of interest is that the color balance was changed when using the 420EX. The G9 was set for automatic white balance so this was an internal decision by the G9.
The Canon 420EX accomplishes High Speed Sync on the G9 by pulsing rapidly. Unfortunately, this means that flash power is reduced for each pulse. Does the 420EX have enough power at High Speed Sync? I switched the 420EX to HSS and fired away. The G9 responded by changing to 1/1000 second at f4 and the 420EX pulsed away (not noticeable to me). As shown here, the 420EX really does not have quite the necessary power but is close. Also, f4 nicely blurred the background as compared to f8.
Last, I replaced the 420EX with a Nikon SB-28. The SB-28 fires from the G9 but the flash power must be set manually and on the flash. The advantage to using the “incompatible” SB-28 on the G9 is that the G9 ignorantly operates at shutter speeds up to 1/2500 second! Realizing that the “correct” ambient exposure was 1/500 at f5.6, I intentionally set the G9 at 1/1000 second and f5.6 to make the background slightly underexposed. The best setting for the SB-28 was at half power which produced this shot. It’s OK, although the color balance is a bit cool.
Conclusions? Well, fill flash can be very useful and is essentially for some situations. The G9 flash is a bit weak but is certainly convenient. The limits on flash power, sync speed and aperture can lead to false conclusions so be aware of those limits. I much prefer external flash (in fact, off-camera flash). The most understandable setup for me is the G9 in manual mode with the SB-28 also in manual mode; however, this setup often requires a bit of tinkering and trial shots – but I’ll get better at it with practice!