New Michael Canon G11 Hands-On Report

[thierrylegros396]

One important thing I clearly noticed is how better the new G11 colors are in comparison with the G10 at ISO800 and above !

More, it seems that the G10 raw files are more heavily processed for color noise.

So for raw user, High ISO G11 files are more usable.

But I also noticed the patern in the wall.

Michael, is it a pattern in the wallpaper or an artifact from sharpening ?!

Have a Nice Day.

Thierry

[madmanchan]

Depends on which pattern you're referring to. As I recall, Michael used LR 3 beta for the images in the article, and luminance NR is not implemented in LR 3 beta.

[Slobodan Blagojevic]

The G10 wall sections used for noise comparison shots appear to be quite underexposed relative to G11. Given that the purpose of the comparison was to compare noise, not metering accuracy, shouldn't a noise comparison be based on equally exposed sections? Underexposed areas are typically noisier, thus making the comparison less accurate. Am I missing something here?

I am also surprised to see much more chroma noise in G11 shots at higher ISO... counterintuitive, to say the least. Is it possible that the LR3 beta profile for G11 is still a work in progress?

[fike]

Some of Michael's recent equipment reviews have seemed a bit brief, but I imagine that with the vast variety of straight review sites out there, he probably doesn't feel like adding more "scientific analysis."  Perhaps a better label for these would be "New Camera Response."

With that said, I too was a bit disappointed to see that there is only about a stop of added low-light performance.  I find myself hoping that the RAW converter IS a problem.  I really expected more from dropping back to 10MP.  

Michael, are you pretty comfortable with the production readiness of the RAW converter you used.

I think that with all the new smaller interchangeable-lens cameras (Olympus and Panasonic) that the G-series may finally be seeing its relevance wane.  It would seem that the S90 may be a better match to this new sensor and provide something the G-11 can't--sleek portability.

[HickersonJasonC]

Michael, I know you long-ago moved from rigorous testing towards subjective analysis. Still, why publish sample images from LR3 Beta, an untested (in the real world) program with obvious limitations. I for one am much less interested in how much actual noise there is in the raw image than I am in the subjective quality of the noise and how it responds to noise reduction. LR3 Beta doesn't even have luminance noise reduction available and the different shooting angles and luminosity values in your noise comparison images make any comparison impossible. Which makes me wonder, why even bother?

And why no mention/update of your recent comparison of the G10 to medium format digital?

[michael]

The cameras were 5 inches apart. The exposures were identical. The processing was identical. LR3B does a great job in raw decoding. I applied no noise reduction of any sort. What you see is what you actually get

Not sure what you found lacking. It's a valid comparison of the two cameras as far as I'm concerned, and I've now received enough emails from others who have done similar comparisons and seen similar results to feel confident that there were no substantial flaws in my methodology.

As for a comparison to medium format, what would you like me to have done or written? Been there done that.

The point of that whole little exercise was to illustrate that small sensor cameras with high pixel counts, when used with great care, and when used at base ISO, and  printed in sizes within their limits, produce very fine results. That's all it shows. Push any limit and they fall apart compared to 35mm let alone MF, but you know that, right?

Michael

[stevesanacore]

I think all these small sensor cameras have reached their limits. The G10 is great for keeping in my pocket and providing a decent shot when it's my only choice. I really hope the new trend for Canon is to make a simple large sensor pocket cameras instead of these overly complex prosumer models.

Thanks again, Michael, for the insight into this new camera.

[Slobodan Blagojevic]

... The exposures were identical. The processing was identical...

Hate to be a pest, but based on your explanations, the exposure is not necessarily identical:

Both cameras were set to f/4 in Aperture Priority mode, ...allowing the camera to compensate for exposure by varying the shutter speed.

Indeed, measuring the same wall area in Photoshop, there appears to be a difference of 1/2 stop (i.e. G10 is underexposed relative to G11). The exposure difference is obvious even to the naked eye. As underexposed or darker areas tend to be noisier, hence my original question if the comparison for noise is accurate enough.

[michael]

What I meant is that each camera was set identically. If one underexposes compared to the other then that's the way it is.

There are all sorts of ways to conduct such a test. If I'd equalized the exposures by analyzing the images then how realistic would that have been? Clearly either the cameras meter differently or have different sensitivities at the same ISO.

I really hate publishing these comparisons because there's always someone that second guess the methodology.  

Michael

[DigitalSteve]

I did a quick comparision using your 1600 ISO crops.  Resing the G10 down to 10MP (about a 82% width and height reduction) looks like the G10 noise and sharpness are better than the G11.

Resing the G11 up to the G10 size...well yuck.

Seems to me taking G10 full size image and reducing it to a 10MP image will look better than the G11 at the same ISO.

I think I'll keep my G10, thanks for saving me $500 bucks.

What say you?

Steve Snyder
Aromas, CA

[fike]

You folks are lunatics.....


slobodan, a half a stop difference has you upset.  Give me a break.  This is not DP Review.  Go there if you want look at those kinds of exhaustive analyses.

DigitalSteve, You resized and analyzed a 45-50KB compressed JPG versions and are drawing conclusions about that.  Mein Gott!!

This is lots simpler than that.

The G11 has slightly less noise, but with a lower resolution.  The difference isn't enough to consider it revolutionary in any way.  DONE!  

Now I can see why Michael is reluctant to say anything for fear of being pixel-peeped to death.


I appreciate a more holistic analysis that Michael offers. My take on it....

To my eye, the grain in the G11 seems a bit finer.  That may be a bit better than the blotchier noise that can be difficult to remove.  

But it is all for naught when you consider the Olympus and Panasonic offerings that are in the same ballpark for compactness yet have far superior quality.

[Slobodan Blagojevic]

... there's always someone that second guess the methodology.  

Did not mean to make you angry or second-guess ... my assumption is that we are all on the same side of the table, trying to contribute to a better understanding of the issue at hand. You certainly provide a valuable community service with your articles, and it is appreciated.

As someone (myself) who does have a G10 and has been looking forward to the IQ improvements G11 is supposed to bring, I have no reason to steer the discussion for or against either camera. With that in mind, I would dare to suggest that the five inch difference in camera positions, coupled with the inherently contrasty scene, could have thrown a curve to the metering and white balance systems in both cameras, resulting in (slightly) different exposure and white balance. Just a thought.

[Slobodan Blagojevic]

You folks are lunatics..... slobodan, a half a stop difference has you upset.  Give me a break.  This is not DP Review.  Go there ...

Nice and classy!

DigitalSteve, You resized and analyzed a 45-50KB compressed JPG versions and are drawing conclusions about that.  Mein Gott!!

OK... fair enough... and yet, based on the same compressed JPG versions, you apparently have no problem drawing your own pixel-peeping conclusions... Mein Gott indeed!

The G11 has slightly less noise... To my eye, the grain in the G11 seems a bit finer...

[eronald]

Michael,

 You did ok with your subjective impressions. Let the pixel peepers do their own tests and post the results here.

 Comparison with the Panasonic was a nice idea. May actually sell some panasonics.

Edmund

What I meant is that each camera was set identically. If one underexposes compared to the other then that's the way it is.

There are all sorts of ways to conduct such a test. If I'd equalized the exposures by analyzing the images then how realistic would that have been? Clearly either the cameras meter differently or have different sensitivities at the same ISO.

I really hate publishing these comparisons because there's always someone that second guess the methodology.  

Michael

[Chris Pollock]

With that said, I too was a bit disappointed to see that there is only about a stop of added low-light performance.  I find myself hoping that the RAW converter IS a problem.  I really expected more from dropping back to 10MP.

Reducing the number of pixels by roughly 1/3 only increases the sensor area per pixel by about 50%, so a one stop improvement isn't bad.

I think the most important lesson from the comparison was how superior the GF1 is to either Canon camera. May the micro four-thirds platform live long and prosper. All it needs now is a wider selection of lenses.

[Daniel Browning]

Reducing the number of pixels by roughly 1/3 only increases the sensor area per pixel by about 50%, so a one stop improvement isn't bad.

One must also consider that the total sensor area itself is 1.5X greater.

[woof75]

What I meant is that each camera was set identically. If one underexposes compared to the other then that's the way it is.

There are all sorts of ways to conduct such a test. If I'd equalized the exposures by analyzing the images then how realistic would that have been? Clearly either the cameras meter differently or have different sensitivities at the same ISO.

I really hate publishing these comparisons because there's always someone that second guess the methodology.  

Michael

In writers circles there's a code of conduct of never answering to your critics, I see the point now. I wouldn't bother Michael, you'll go round and round in circles and never make everyone happy. I find your reviews to be useful and informative, there's a good reason your site is so popular.

[Paulo Bizarro]

First of all, I would like to congratulate Michael for an informative comparison. The setting in which the G11 has to compete is completely diffeent from when the G10 appeared. Micro-four thirds cameras are clearly pushing the envelope in terms of the combination of "larger sensor/small camera" style of cameras.

However, I think there is still a place for cameras like the G11 (or Panasonic/Leica LX3/D-Lux 4, or the little Ricohs), if you want a travel camera with compact form factor, and do not have to use high ISO's. The advantage of the G11 is the flexibility of the zoom; once you attach a zoom lens to the Panasonic GF1, you end up with not so much a compact camera anymore.

In the end, there is no contention that cameras with larger sensors have better image quality; however, at lower ISO's, the difference is less notorious. I am happy there is a large choice for us photographers, it means we can pick the right tools for our endeavours.

[Chris Pollock]

One must also consider that the total sensor area itself is 1.5X greater.

Where did you get your information? Everything I've read indicates that the G11's sensor is the same size as the G10's. A larger sensor would require a new lens, and other major changes, which doesn't appear to be the case.

[fike]

Nice and classy!

Haha.  I was speaking facetiously, of course.