Just when I think I'm going to get to take back all those nasty things I've said about Canon and their megapixel war, they release the 7D. The 7D is the newest crop sensor DSLR from Canon. Unlike the G11, the Canon marketing dope-wads got their grubby little meat hooks full into this camera...
It's sporting the Canon APS-C (22.3 x 14.9 mm) sensor with a "how do you like me now?" 18 megapixels. For all of you busting out your calculators, I'll save you some time - the pixel size is 4.3um. To put this in terms of a full frame DSLR (24mm x 36mm), that would translate to silly 47 megapixels.
First, the camera incorporates some new and potentially exciting features. The viewfinder, the auto-focus system, and metering are all features I'm glad to see refreshed. It needed to happen long ago and it looks like they put their best foot forward on many parts of this camera. These features really should have been incorporated into the 5DmkII. Lets hope that a 5DmkII "n" version get released where the pixel counts stays the same and the rest of the system gets these new features.
Getting back to the 7D - The pixel count from a perspective of need and lens diffraction is a huge disappointment (noise could certainly end up on this list as well). I know a large format printer who will tell say that you can take a 10MP image and as long as it's sharp, you can print it as large as you could ever possibly want. From a lens diffraction perspective, there were many out there who felt like the 50D with it's APS-C 15MP sensor was at the limit of usable aperture range before diffraction became a problem. What does 18MP mean for anything less than the top of the line "L" lenses?
These things in mind, I don't understand the massive pixel count numbers in the DSLRs these days. These large numbers offer very little in terms of additional capability, can possibly hurt image quality (depending on the lens), and then add cost to systems to store and process the files.
So release the hounds on Canon for being the winner of the race (or war) that no one wanted to even enter... Mind you, had it not been for the features they at least tried to get right, it would have been the hounds with bees in their mouths, so when they barked they would also shoot bees...
Update: from a hands-on over at Gizmodo:
"At ISO1600, it's not so bad, even if it's not 5D level, but it starts getting kinda gruesome at ISO3200"
It hurts to read that when you know they are comparing camera processed jpeg files where the in-camera noise reduction is most likely set to" stun." I'm just not sure why they added all those tiny pixels...
Read more!
Monday, August 31, 2009
Wednesday, August 19, 2009
End of the Megapixel War?
Every August(ish) Canon releases it's newest point and shoot cameras, so today would be like any other Canon release day with the exception of the release of the G11, the new version of the G10. So, yes, it's a new year and it came with a new camera, what's the big deal? The big deal is this: Canon took the G series line of point and shoot cameras from 14.7MP to 10.1 MP with the new release.
This is the first time, to my knowledge, that any manufacture has backtracked on the pixel count in a camera line. Yes, there have been cameras where count has stayed the same or a new model has come out with lower count, but this was a true reduction (albeit a very positive one) in megapixels.
What does this mean to the Megapixel war? I don't know. It could be that the cold dead hand of Canon's marketing department has finally been lopped off, the spell lifted, and consumer demand can and has actually change the world (outside of Tropicana). Or it could mean that we are experiencing a temporary low in the ebb and flow of the war (this could have simply slipped through the cracks). No matter, I see this as an extremely positive move by Canon. I hope to see it continue as they release cameras.
Now if they would release a sub $4500 full frame DSLR with a real working auto focus system I would have to retract all the bad thing I've said about them over the years...
Read more!
This is the first time, to my knowledge, that any manufacture has backtracked on the pixel count in a camera line. Yes, there have been cameras where count has stayed the same or a new model has come out with lower count, but this was a true reduction (albeit a very positive one) in megapixels.
What does this mean to the Megapixel war? I don't know. It could be that the cold dead hand of Canon's marketing department has finally been lopped off, the spell lifted, and consumer demand can and has actually change the world (outside of Tropicana). Or it could mean that we are experiencing a temporary low in the ebb and flow of the war (this could have simply slipped through the cracks). No matter, I see this as an extremely positive move by Canon. I hope to see it continue as they release cameras.
Now if they would release a sub $4500 full frame DSLR with a real working auto focus system I would have to retract all the bad thing I've said about them over the years...
Read more!
Tuesday, March 3, 2009
Consumer Backlash Can Change the World
About a week ago the Times ran an article on how consumer backlash caused Tropicana (the orange juice maker) to revert back to their old packaging due to consumers telling them how much they hated the new package. The significance is that Tropicana dropped $35million on creating that new branding. Yep, Tropicana literally threw away $35million dollars because customers simply didn't like it.
You can see the full article here: Tropicana in the Times
Why do I feel like I should tell you this and how does this relate to digital cameras? I bring this to your attention to show you that e-mails and consumer feedback can make a difference. With rumors of the megapixel race continuing in future DSLR bodies and the current batch of point and shoot cameras that were released for PMA, it's clear that camera manufactures haven't gotten the message. Stop the madness. We don't need 12MP point and shoot cameras. We want 6MP point and shoot cameras with relatively large sensors. What is so hard to understand?
If you find some time in the next few days when you are near you computer, take a minute and provide some feedback to your camera manufacture of choice. Let them know you don't want more megapixels. It's a simple e-mail and every manufacture will have one. It just takes a minute to let a manufacture know you are disappointed in their senseless choices and want something different....
Read more!
You can see the full article here: Tropicana in the Times
Why do I feel like I should tell you this and how does this relate to digital cameras? I bring this to your attention to show you that e-mails and consumer feedback can make a difference. With rumors of the megapixel race continuing in future DSLR bodies and the current batch of point and shoot cameras that were released for PMA, it's clear that camera manufactures haven't gotten the message. Stop the madness. We don't need 12MP point and shoot cameras. We want 6MP point and shoot cameras with relatively large sensors. What is so hard to understand?
If you find some time in the next few days when you are near you computer, take a minute and provide some feedback to your camera manufacture of choice. Let them know you don't want more megapixels. It's a simple e-mail and every manufacture will have one. It just takes a minute to let a manufacture know you are disappointed in their senseless choices and want something different....
Read more!
Wednesday, January 14, 2009
We Don't Need More Stinking Megapixels and Film vs. Digital
Ok, so over on the "The Gadget Show" they did a head to head comparison of film prints to digital prints. The images captured were printed at a size that required the photos to be hung from the side of a building. You can see the segment here:
http://fwd.five.tv/videos/challenge-blow-up-part-3
I have to say that I find the results fascinating for multiple reasons. I'm going to make you watch the video to find out the outcome of the challenge, but I do want to comment on the pixel count of the camera they used for the digital side of the challenge. They used a D700 - a 12 megapixel camera. They enlarged the print from a D700 to where it was SO large they had to hang it from a building. That begs the question - how many megapixels are enough?
See how I brought that back around and I'm back to on my old soap box. If you can take 12 megapixels and print that large, how many pixels are truly necessary? The megapixel race and the reasons for needing more pixels is becoming a very old joke. Sure, printing and size all depends on viewing distance, but for 99% of DSLR applications this (and lots of other things which I've previously ranted about) tells me that 12 megapixels is really plenty.
Camera manufactures - yes all of you producing any type of digital camera - STOP IT! STOP THE MADNESS. I know you need to drive sales of cameras some how, but put your time and energy into high ISO quality, AF systems, quality assurance (Yes, I'm looking at you Canon), digital horizons (good job Nikon), weather sealing, etc. We don't need more stinking megapixels...
Read more!
http://fwd.five.tv/videos/challenge-blow-up-part-3
I have to say that I find the results fascinating for multiple reasons. I'm going to make you watch the video to find out the outcome of the challenge, but I do want to comment on the pixel count of the camera they used for the digital side of the challenge. They used a D700 - a 12 megapixel camera. They enlarged the print from a D700 to where it was SO large they had to hang it from a building. That begs the question - how many megapixels are enough?
See how I brought that back around and I'm back to on my old soap box. If you can take 12 megapixels and print that large, how many pixels are truly necessary? The megapixel race and the reasons for needing more pixels is becoming a very old joke. Sure, printing and size all depends on viewing distance, but for 99% of DSLR applications this (and lots of other things which I've previously ranted about) tells me that 12 megapixels is really plenty.
Camera manufactures - yes all of you producing any type of digital camera - STOP IT! STOP THE MADNESS. I know you need to drive sales of cameras some how, but put your time and energy into high ISO quality, AF systems, quality assurance (Yes, I'm looking at you Canon), digital horizons (good job Nikon), weather sealing, etc. We don't need more stinking megapixels...
Read more!
Tuesday, January 6, 2009
Pixel Count and Noise
I keep coming back to the Canon 5D mark II and its 21 megapixels and what this means to me as a Canon shooter. Yes, I call myself a Canon shooter as the cost of lenses keeps me shooting Canon. There has been much debate all over the internet about the noise performance between the Canon large 21 megapixel sensors (5DmkII and 1DsmkIII) and the ~12 megapixel Nikon image sensors (D700 and D3).
Noise and how you measure it is debatable. There are some that say when comparing noise between cameras or image sensors you have to down sample to the smaller image sensor's resolution and then compare noise. If we think about this, it means you have to "merge" or "compress" those precious pixels down to the smaller sensor size in order to compare performance between lower count image sensors and higher count image sensors.
What's that? Huh? On one hand the "megapixel war mongers" are telling me that I have to "throw out" pixels to evaluate performance, but in the same breath they are telling me that by increasing pixel count I will have more latitude for cropping my image. Hold on, I have to "throw out" or merge pixels in order to evaluate noise performance, but I now have more latitude to crop? Does that make sense? The short answer - Absolutely Not! If I crop my image because I have more megapixels, I can't compress the native noise and get comparable image quality. It's simple, pixels can't be merged if they are not there...
When evaluating noise, a 100% crop at full resolution must be used. Nothing else makes sense. It's that simple. Merging pixels to measure noise is marketing hype that should be ignored. Throwing away megapixels when evaluating image quality means that those pixels are not of value to you in regards to your final image. Well, it's that, or you'll have to merge pixels to get the quoted image quality in the end anyway.
Products should be tested at their limits so we we need to use them at their limits we know what to expect from them.
That said, the big question is do I start with a smaller, better performing sensor, or merge pixels down to get image quality from a larger sensor? Sounds like a half a dozen of one or 6 of the other. Only you can decide which solution is better, but don't let the marketing department convince you that pixel count can come without penalty.
Read more!
Noise and how you measure it is debatable. There are some that say when comparing noise between cameras or image sensors you have to down sample to the smaller image sensor's resolution and then compare noise. If we think about this, it means you have to "merge" or "compress" those precious pixels down to the smaller sensor size in order to compare performance between lower count image sensors and higher count image sensors.
What's that? Huh? On one hand the "megapixel war mongers" are telling me that I have to "throw out" pixels to evaluate performance, but in the same breath they are telling me that by increasing pixel count I will have more latitude for cropping my image. Hold on, I have to "throw out" or merge pixels in order to evaluate noise performance, but I now have more latitude to crop? Does that make sense? The short answer - Absolutely Not! If I crop my image because I have more megapixels, I can't compress the native noise and get comparable image quality. It's simple, pixels can't be merged if they are not there...
When evaluating noise, a 100% crop at full resolution must be used. Nothing else makes sense. It's that simple. Merging pixels to measure noise is marketing hype that should be ignored. Throwing away megapixels when evaluating image quality means that those pixels are not of value to you in regards to your final image. Well, it's that, or you'll have to merge pixels to get the quoted image quality in the end anyway.
Products should be tested at their limits so we we need to use them at their limits we know what to expect from them.
That said, the big question is do I start with a smaller, better performing sensor, or merge pixels down to get image quality from a larger sensor? Sounds like a half a dozen of one or 6 of the other. Only you can decide which solution is better, but don't let the marketing department convince you that pixel count can come without penalty.
Read more!
Tuesday, December 16, 2008
Back by Popular Demand - Image Sensor Sizes
I think this must be the blog post that has gotten more views than any other. It's for that reason that I've updated the table with additional information. Something to note, I've added the sensor size for the Panasonic DMC-LX3 (my current top pick for a point and shoot camera) and the Micro FourThirds Sensors.
Once again, just as a refresher, I've created a number of images that shows the relative active pixel array size (the important part of the imager - where the pixels are) to a piece of 35mm film. The sensor size is indicated by the red box (with black outline) and the larger black box is the size of a piece of 35mm film. Now there is no way for me to make these show up on your screen in an actual size, so keep in mind that 35mm film is 36mm x 24mm. In the images size from sensor to 35mm film is relative and accurate in comparison.
| vs. 35mm Film | Sensor Type | Aspect Ratio | Diagonal (mm) | Width (mm) | Length (mm) |
|---|---|---|---|---|---|
 | 1/3.6" | 4:3 | 5 | 4 | 3 |
 | 1/3.2" | 4:3 | 5.68 | 4.536 | 3.416 |
 | 1/3" | 4:3 | 6 | 4.8 | 3.6 |
 | 1/2.7" | 4:3 | 6.721 | 5.371 | 4.035 |
 | 1/2.5" | 4:3 | 7.182 | 5.76 | 4.29 |
 | 1/3.2" | 4:3 | 8 | 6.4 | 4.8 |
 | 1/1.8" | 4:3 | 8.933 | 7.176 | 5.319 |
 | 1/1.7" | 4:3 | 9.5 | 7.6 | 5.7 |
 | 1/1.63" | 4:3 | 10.3 | 8.2 | 6.2 |
 | 2/3" | 4:3 | 11 | 8.8 | 6.6 |
 | 4/3" | 4:3 | 22.5 | 18 | 13.5 |
 | ~1.6 Crop | 3:2 | 28.4 | 25.1 | 16.7 |
 | ~1.3 Crop | 3:2 | 33.8 | 28.1 | 18.7 |
 | Full Frame | 3:2 | 43.3 | 36 | 24 |
Again, good visual reference... Some of those sensors are super small...
Read more!
Wednesday, September 24, 2008
Video in DSLR Cameras
There's been an awful lot of hub-bub about video in DSLRs around the net lately with the release of the 5DmkII. Vincent Laforet was able to sweet talk Canon out of a pre-production 5DmkII for 48hours. In that time he managed to spend $5000.00 ($2K on a helicopter ride) and put together an incredibly cheesy, but impressive looking (considering it was shot with a DSLR) Short. Of course the video has set the Net on fire and now it seems everyone wants to become videographers (including Vincent).
By no means do I want anyone to think that I'm down on Vincent's video project. I'm am a little shocked as photography is Vincent's bread and butter and he seems pretty distracted by video right now. He says he took only a handful of stills and that was during his filming - like actually while filming (not as in stopped filming, but as momentary pause. The truth of it is that I enjoy Vincent's photography and I sincerely hope that video does not mean his photography will suffer.
In regards to video in DSLRs, I think that Rob Haggart over at "A Photo Editor" put it best in a recent blog entry called "What If Your Still Camera Also Shot Great Video?"
"I’m reminded of people who are writer/photographer and how at a certain level in this industry you do one or the other better and the result is always better when you put your efforts into one or the other not both. I think the same will ring true for photographer/videographer." - Rob Haggart
Read more!
By no means do I want anyone to think that I'm down on Vincent's video project. I'm am a little shocked as photography is Vincent's bread and butter and he seems pretty distracted by video right now. He says he took only a handful of stills and that was during his filming - like actually while filming (not as in stopped filming, but as momentary pause. The truth of it is that I enjoy Vincent's photography and I sincerely hope that video does not mean his photography will suffer.
In regards to video in DSLRs, I think that Rob Haggart over at "A Photo Editor" put it best in a recent blog entry called "What If Your Still Camera Also Shot Great Video?"
"I’m reminded of people who are writer/photographer and how at a certain level in this industry you do one or the other better and the result is always better when you put your efforts into one or the other not both. I think the same will ring true for photographer/videographer." - Rob Haggart
Read more!
Subscribe to:
Posts (Atom)
