Have you ever thought about running for public office? While I do know your trying to answer my question, you sound just like a politician trying not to answer a question or be perceived as having a position on the subject.
God bless your little heart.
Divamum,
Thanks for your subjective and imperial based opinion.
After looking at the various sites and images available, I have concluded that the real life out of camera image noise is very similar in both cameras. The additional detail, or other technical differences seems to allow the 7D image to clean up better, so while the actual noise produced by the 7D does not seem improved, if you have the software and knowledge you can process an image to display / print with less visible noise.
Short answer: The 7D isn't any better (from a noise standpoint) at high ISO than a 30D, but can clean up better.
Having owned the 30D, 40D and now 7D, there is no question in my mind that the 7D image quality is the best of the bunch. I'm finding pixel-level noise of the 7D about the same as my 40D was up to about 1600 - at 1600 and above the 7D, even at the pixel level, seems to do better. I thought my 40D was an improvement over the 30D so there you go... I'll echo what has also been said about the nature of the noise -there seems to be much finer-grained noise with perhaps more luminosity noise and less chroma noise. I'm finding clean-up of even very high ISO images to result in usable shots above and beyond what the noise-level would indicate. That is, given equally noisy 40D and 7D images, my 7D images end up much better after processing.
For me, I'm also seeing better images because I'm getting more keepers. The focus system is much improved, especially the peripheral focus points. Tracking moving subjects has improved as well - I'm getting more keepers of my 2 1/2 year old twins who never sit still!
I wince a bit when I see people only using things like DXO Mark to compare cameras - the RAW image quality is only one piece of the picture. Besides, don't forget you're comparing RAW image quality as seen by their RAW converter - your real-life results may differ. Ability to acquire a better image in the first place thanks to other features cannot be ignored. Things like DXO Mark test very specific aspects of image quality and results should be interpreted in that context only - there is no "which camera is better" test...
That said, I have used my father-in-law's 5DMKII extensively and there's no doubt it has about a 1-stop advantage in overall noise performance over the 7D. Would I trade him cameras if he offered? Maybe, maybe not. I think the 7D is a better all-around camera, while the 5DMKII shines when you absolutely need the best possible high-ISO performance and per-pixel resolution. At the end of the day the shots I'm not happy with will likely be due to my own lack of skill rather than whether I was holding the 7D or 5DMKII...
Let's not forget that almost ANY of the mainstream cameras today far exceed the abilities of the top-of-the-line pro bodies of 5 years ago -- at that time people thought those were perfectly adequate to get amazing images. I doubt few people are truly gear-limited and would see more improvement by working on photography skills rather than buying the latest and greatest tech.. Of course, I'm a gear-head as well and love new toys!
I believe I have a good feel for the pluses and minuses of the 7D.
As stated before I am contemplating trading in my 5D mark I on a 7D. cry:cry
I really like the 5D mark I and it's a great backup camera. If I ever need it I know it will perform admirably. My major concern was the low light high ISO performance as a second camera.
But then again the 7D opens up other photographic doors.
... However I must call fowl on the print conversion as far as dynamic range is concerned. The print conversion DXO does is effectively reducing the resolution to 8MP.
The 30D is already 8MP, so no change is noted. The 7D and 5DII both gain 1EV in dynamic range. However I do not see how any decrease in resolution can create something new. (In this case more dynamic range.)
John, in rereading my response it does not appear that I answered the question you appear to pose.
I agree that at first glance dynamic range (DR) should not vary with pixel count. The answer lies in the definition of dynamic range. Imager Dynamic Range is a measurement of the brightest details in an image compared to the darkest details in an image, and typically the image measured is a capture of a standardized "grayscale step chart" with 1 EV steps. When you can no longer accurately measure detail, either high/white or low/black, that's what defines the range of tones.
A problem occurs in measuring the dark tones and when the noise levels equal the measurable detail, that is when the the bottom end is determined. So the dark tone details are capped when the SNR equals 1:1. That's just the standard used for the DR test.
When you downscale (in this case normalize a higher resolution to 8MP) with appropriate interpolation, the SNR is reduced due to intrinsic statistical sampling and the DR is extended in the shadow/dark tones.
That's why the DR varies with normalizing from an imager of higher than 8 MP and comparing to an 8 MP imager. An actual 8" x 10" print should bear out the basic premise and that's why the normalization is important in comparing imagers of disparate net pixel count.
"... Dynamic range is therefore limited by the darkest tone where texture can no longer be discerned; we call this the black level. The black level is limited by how accurately each photosite can be measured, and is therefore limited in darkness by image noise. ..."
John, in rereading my response it does not appear that I answered the question you appear to pose.
I agree that at first glance dynamic range (DR) should not vary with pixel count. The answer lies in the definition of dynamic range. Imager Dynamic Range is a measurement of the brightest details in an image compared to the darkest details in an image, and typically the image measured is a capture of a standardized "grayscale step chart" with 1 EV steps. When you can no longer accurately measure detail, either high/white or low/black, that's what defines the range of tones.
A problem occurs in measuring the dark tones and when the noise levels equal the measurable detail, that is when the the bottom end is determined. So the dark tone details are capped when the SNR equals 1:1. That's just the standard used for the DR test.
When you downscale (in this case normalize a higher resolution to 8MP) with appropriate interpolation, the SNR is reduced due to intrinsic statistical sampling and the DR is extended in the shadow/dark tones.
That's why the DR varies with normalizing from an imager of higher than 8 MP and comparing to an 8 MP imager. An actual 8" x 10" print should bear out the basic premise and that's why the normalization is important in comparing imagers of disparate net pixel count.
"... Dynamic range is therefore limited by the darkest tone where texture can no longer be discerned; we call this the black level. The black level is limited by how accurately each photosite can be measured, and is therefore limited in darkness by image noise. ..."
Thank you for the excellent reply. So in essence, the tools used to compute dynamic range 'give up' measuring at certain noise levels. So when the image is down sampled it actually allows the tools to measure more because of a reduction of noise.
For all we know, we could run a 30D image through noise reduction and the dynamic range will magically increase. The only real way to compare dynamic range using these tools is to do so at native image size.
Canon Gear: 5D MkII, 30D, 85 1.2 L, 70-200 2.8 IS L, 17-40mm f4 L, 50 1.4, 580EX, 2x 580EXII, Canon 1.4x TC, 300 f4 IS L, 100mm 2.8 Macro, 100-400 IS L Other Gear: Olympus E-PL1, Pan 20 1.7, Fuji 3D Camera, Lensbaby 2.0, Tamron 28-75 2.8, Alien Bees lighting, CyberSyncs, Domke, HONL, FlipIt.
~ Gear Pictures
Thank you for the excellent reply. So in essence, the tools used to compute dynamic range 'give up' measuring at certain noise levels. So when the image is down sampled it actually allows the tools to measure more because of a reduction of noise. ...
I think that DXOMark uses the same criteria as "ImaTest" software with regard to clamping the shadow tones when the SNR reaches 1:1. Beyond this point noise is more significant than detail. I believe this is also an industry standard for imager DR measurement.
... For all we know, we could run a 30D image through noise reduction and the dynamic range will magically increase. The only real way to compare dynamic range using these tools is to do so at native image size.
The reason that "normalization" to a particular megapixel standard is important is because that's the same thing that happens when you print or scale for the Internet, for example. In particular normalizing should give a more accurate comparative depiction of printed results. I appreciate that DXOMark provides both measurements, but the normalized results give me the best concept of what will happen when I print, and that's the most important consideration for my work.
Noise reduction will impact DR measurements and it is significant that the DXOMark measurements are only against RAW images "and" DXO is the RAW converter used for the tests. Adobe ACR RAW conversion would yield slightly different results.
I don't use DXOMark as the only criteria for evaluating cameras and I don't recommend anyone else do that either. It is significant that it confirms other sites' evaluations of the Nikon D3S and D3X in terms of amazingly low noise footprints, dynamic range and color rendition. It also gives Canon and Sony full-frame owners similar cause to rejoice as it shows a distinct improvement over crop cameras by the same manufacturers. Even entry level crop cameras are much better than middle line crop cameras of just a few years ago. (But older cameras also benefit from improved current software.)
The truth is that we have an amazing selection of different photographic tools and systems available to satisfy almost any need and budget, and the image quality potential of current dSLRs is just spectacular at almost every level.
Comments
Have you ever thought about running for public office? While I do know your trying to answer my question, you sound just like a politician trying not to answer a question or be perceived as having a position on the subject.
God bless your little heart.
Divamum,
Thanks for your subjective and imperial based opinion.
After looking at the various sites and images available, I have concluded that the real life out of camera image noise is very similar in both cameras. The additional detail, or other technical differences seems to allow the 7D image to clean up better, so while the actual noise produced by the 7D does not seem improved, if you have the software and knowledge you can process an image to display / print with less visible noise.
Short answer: The 7D isn't any better (from a noise standpoint) at high ISO than a 30D, but can clean up better.
Sam
For me, I'm also seeing better images because I'm getting more keepers. The focus system is much improved, especially the peripheral focus points. Tracking moving subjects has improved as well - I'm getting more keepers of my 2 1/2 year old twins who never sit still!
I wince a bit when I see people only using things like DXO Mark to compare cameras - the RAW image quality is only one piece of the picture. Besides, don't forget you're comparing RAW image quality as seen by their RAW converter - your real-life results may differ. Ability to acquire a better image in the first place thanks to other features cannot be ignored. Things like DXO Mark test very specific aspects of image quality and results should be interpreted in that context only - there is no "which camera is better" test...
That said, I have used my father-in-law's 5DMKII extensively and there's no doubt it has about a 1-stop advantage in overall noise performance over the 7D. Would I trade him cameras if he offered? Maybe, maybe not. I think the 7D is a better all-around camera, while the 5DMKII shines when you absolutely need the best possible high-ISO performance and per-pixel resolution. At the end of the day the shots I'm not happy with will likely be due to my own lack of skill rather than whether I was holding the 7D or 5DMKII...
Let's not forget that almost ANY of the mainstream cameras today far exceed the abilities of the top-of-the-line pro bodies of 5 years ago -- at that time people thought those were perfectly adequate to get amazing images. I doubt few people are truly gear-limited and would see more improvement by working on photography skills rather than buying the latest and greatest tech.. Of course, I'm a gear-head as well and love new toys!
Ziggy, Pathfinder, Divamom, Cswinton
I believe I have a good feel for the pluses and minuses of the 7D.
As stated before I am contemplating trading in my 5D mark I on a 7D.
I really like the 5D mark I and it's a great backup camera. If I ever need it I know it will perform admirably. My major concern was the low light high ISO performance as a second camera.
But then again the 7D opens up other photographic doors.
Sam
John, in rereading my response it does not appear that I answered the question you appear to pose.
I agree that at first glance dynamic range (DR) should not vary with pixel count. The answer lies in the definition of dynamic range. Imager Dynamic Range is a measurement of the brightest details in an image compared to the darkest details in an image, and typically the image measured is a capture of a standardized "grayscale step chart" with 1 EV steps. When you can no longer accurately measure detail, either high/white or low/black, that's what defines the range of tones.
A problem occurs in measuring the dark tones and when the noise levels equal the measurable detail, that is when the the bottom end is determined. So the dark tone details are capped when the SNR equals 1:1. That's just the standard used for the DR test.
When you downscale (in this case normalize a higher resolution to 8MP) with appropriate interpolation, the SNR is reduced due to intrinsic statistical sampling and the DR is extended in the shadow/dark tones.
That's why the DR varies with normalizing from an imager of higher than 8 MP and comparing to an 8 MP imager. An actual 8" x 10" print should bear out the basic premise and that's why the normalization is important in comparing imagers of disparate net pixel count.
More about dynamic range measurement here:
http://www.cambridgeincolour.com/tutorials/dynamic-range.htm
In particular this statement:
"... Dynamic range is therefore limited by the darkest tone where texture can no longer be discerned; we call this the black level. The black level is limited by how accurately each photosite can be measured, and is therefore limited in darkness by image noise. ..."
Moderator of the Cameras and Accessories forums
Thank you for the excellent reply. So in essence, the tools used to compute dynamic range 'give up' measuring at certain noise levels. So when the image is down sampled it actually allows the tools to measure more because of a reduction of noise.
For all we know, we could run a 30D image through noise reduction and the dynamic range will magically increase. The only real way to compare dynamic range using these tools is to do so at native image size.
Other Gear: Olympus E-PL1, Pan 20 1.7, Fuji 3D Camera, Lensbaby 2.0, Tamron 28-75 2.8, Alien Bees lighting, CyberSyncs, Domke, HONL, FlipIt.
~ Gear Pictures
I think that DXOMark uses the same criteria as "ImaTest" software with regard to clamping the shadow tones when the SNR reaches 1:1. Beyond this point noise is more significant than detail. I believe this is also an industry standard for imager DR measurement.
The reason that "normalization" to a particular megapixel standard is important is because that's the same thing that happens when you print or scale for the Internet, for example. In particular normalizing should give a more accurate comparative depiction of printed results. I appreciate that DXOMark provides both measurements, but the normalized results give me the best concept of what will happen when I print, and that's the most important consideration for my work.
Noise reduction will impact DR measurements and it is significant that the DXOMark measurements are only against RAW images "and" DXO is the RAW converter used for the tests. Adobe ACR RAW conversion would yield slightly different results.
I don't use DXOMark as the only criteria for evaluating cameras and I don't recommend anyone else do that either. It is significant that it confirms other sites' evaluations of the Nikon D3S and D3X in terms of amazingly low noise footprints, dynamic range and color rendition. It also gives Canon and Sony full-frame owners similar cause to rejoice as it shows a distinct improvement over crop cameras by the same manufacturers. Even entry level crop cameras are much better than middle line crop cameras of just a few years ago. (But older cameras also benefit from improved current software.)
The truth is that we have an amazing selection of different photographic tools and systems available to satisfy almost any need and budget, and the image quality potential of current dSLRs is just spectacular at almost every level.
Moderator of the Cameras and Accessories forums