now then Fred.. are you falling for the mono sensor propaganda!?
it's only 15% after some fairly pointless geometric scaling factors used in this highly referenced article: http://www.astrosurf.com/buil/50d/test.htm (which otherwise has some nice test data).
real QE is not so bad at all compared to other (CMOS) colour sensors.
Phi
I'm talking pinnacle, not relative goodness. Maybe in the future they will be able measure the frequency of a photon as it bounces around in a well when it's received or something, without filters. Then res and qe will rocket.
I'm talking pinnacle, not relative goodness. Maybe in the future they will be able measure the frequency of a photon as it bounces around in a well when it's received or something, without filters. Then res and qe will rocket.
What is geometric QE? Never heard of anyone else using that measure.
I think the 30% QE figure is more correct for a 5D mark iii. That would be in line with colour CCDs using Bayer matrix. 15% would indeed be woeful but as we know that is not the case. I see about 1/2 the light sensitivity
60% QE ML8300 or PL16803 in my Canon 40D.
Are there low light high ISO samples of 5D mark iii done in RAW anywhere or are the only examples just jpegs?
I saw some samples showing low noise in jpeg form but obvious heavy noise reduction reducing detail quite a lot. A bit of a scary thought.
I hope the claims of low light high ISO performance isn't only in camera noise reduction performance. Posted RAW images are what are needed to confirm this aspect of the camera.
I do know that Canon is the only manufacturer that does not fiddle with the RAW data like some others to reduce noise. This inevitably results in faint detail being lost such as faint stars that look like noise.
Canon also makes sure that the Silicon is ultra pure so minimising hot and warm pixels. They control all stages of the fabrication of their chips.
Mainly Sony. Except Sigma Foveon (which seems to be more or less dead anyway). Same Sony chips get into quite respectable astro OSC CCDs (QHY, StarlightExpress, Atik and even heavy hitters like FLI and Apogee).
Shame Canon won't give their CMOS' to anyone. Both Canon and Sony have superior noise characteristics to Kodak chips. If they would just strip the Bayer filter and cool them, THAT would be a significant step forward!!!
There are some RAWs floating around the web, but, you can't open them with Digital Photo Professional, yet. This means, that any attempts at viewing the files in third party viewers will not show you the true output from the camera.
It won't be long until Canon releases an update to DPP, and, then, we can see what the thing can really do.
There are too many variables comparing these output images posted on the net.
There are jpegs with their incamera noise reduction routines.
Then there are the vagaries of which RAW converter and the differences they can create. I was reading an article about different RAW converters after you mentioned you only use the Canon one. The example photos showed quite a large variation in quality of images just from which RAW converter you used.
So until the cameras are out in the field it remains vague as to the performance of these new cameras. I think though both 5D mark iii and D800 will both perform better than their predecessors in the area of low light high ISO judging by sample images.
Nikon D800 is a Sony sensor designed by Nikon and per that article Bert mentioned Nikon obviously likes to design the sensor and get someone else to make it. Or at least have some input into the sensor process. Sony are good at making chips as all the planetary,QHY, Starlight Express imagers have proven over the years.
Again it seems to me both cameras are sensational. But I also see the mark ii as being a real bargain in that some of the new features of the iii, if you are not a professional, you could easily live without.
A new feature called Digital Lens Optimizer processes RAW images to achieve ideal optical characteristics for all types of optical aberration or diffraction, effects of a low-pass filter in front of a CMOS sensor, etc. This function improves image quality particularly in the image periphery in addition to the image center. This function is made possible because the entire design-through-manufacture process, for camera, CMOS sensor, EF lens, and DPP, is carried out entirely at Canon. Images are processed optimally using lens information in the image files (focal length, subject distance, and aperture) and lens data specially for the Digital Lens Optimizer. (However, the size of a .CR2 file will be two to three times larger after applying the Digital Lens Optimizer.) Adjustments are made for such aspects as spherical aberration, chromatic aberration, astigmatism, curvature of field, sagittal halo, chromatic aberration of magnification, axial chromatic aberration, diffraction, and the effects of a low-pass filter in front of the CMOS sensor. DPP’s Digital Lens Optimizer will be usable with any of 29 compatible lenses initially. It works with .CR2 files from EOS models released since 2006 (EOS 30D and forward).
But I know what you mean. I have a 20D and a 40D and I shot a 5D mark ii and the 40D seemed to have same or better noise performance as the 5D mark ii or at minimum it was similar.
A new feature called Digital Lens Optimizer processes RAW images to achieve ideal optical characteristics for all types of optical aberration or diffraction, effects of a low-pass filter in front of a CMOS sensor, etc. This function improves image quality particularly in the image periphery in addition to the image center. This function is made possible because the entire design-through-manufacture process, for camera, CMOS sensor, EF lens, and DPP, is carried out entirely at Canon. Images are processed optimally using lens information in the image files (focal length, subject distance, and aperture) and lens data specially for the Digital Lens Optimizer. (However, the size of a .CR2 file will be two to three times larger after applying the Digital Lens Optimizer.) Adjustments are made for such aspects as spherical aberration, chromatic aberration, astigmatism, curvature of field, sagittal halo, chromatic aberration of magnification, axial chromatic aberration, diffraction, and the effects of a low-pass filter in front of the CMOS sensor. DPP’s Digital Lens Optimizer will be usable with any of 29 compatible lenses initially. It works with .CR2 files from EOS models released since 2006 (EOS 30D and forward).
w00t w00t.
H
Also from Chuck Westfall is
"We won't disclose that information"
When asked about the noise improvement for RAWs. For JPGs they are talking big time but the decision to avoid discussing RAWs for me is bad news.
I guess the 5DIII is very optimized for JPGs but will show little improvement over the 5DII for RAWs and the price is heavy.
It doesn't seem like a good camera for astro. It would be great instead for sports, journalism and weddings.
Also from Chuck Westfall is
When asked about the noise improvement for RAWs. For JPGs they are talking big time but the decision to avoid discussing RAWs for me is bad news.
I guess the 5DIII is very optimized for JPGs but will show little improvement over the 5DII for RAWs and the price is heavy.
It doesn't seem like a good camera for astro. It would be great instead for sports, journalism and weddings.
I think this is true for the cheaper EOS xxx series. But the 5 is a professional camera so I don't know why not optimized for RAWs.
I think only a very few cameras have less noise at high ISO than the 5Mk3, but I doubt whether the 5Mk3 is much better than the 5Mk2, but I think as an SLR the 5Mk2 or 5Mk3 are one of the best astro SLRs due to the full frame and low light performance.
I'm still happy, even if it is just one stop of performance. It's one stop that wasn't there before.
Your Mark II will still take beautiful landscapes, until Canon makes a high megapixel camera. Perhaps, even a new product line. But, I won't speculate as that will make me a rumour monger!
Full Well depth is usually simply a function of the size of the pixel (there may be other considerations but that is what I know). 5D iii pixels are slightly smaller than Mark ii. (http://www.ptgrey.com/support/kb/index.asp?a=4&q=192 unless Canon lowered the voltage).
So assuming full wells are smaller then the only way noise is lower is if dynamic range was also larger than the mark ii - right?
Unless they managed to increase full well depth by reducing voltages.
Well per one poster on DPReview, dynamic range for the mark iii is slightly lower than for the mark ii.
So how could ISO performance in the mark iii be improved over the mark ii without some trickery (like RAWs are being smoothed first - mind you Nikon does that).
I think the answer to all these questions will be answered in about 6 months time, after the cameras (both 5D MkIII & D800) have been in the hands of non-factory photographers. Until then, most of these claims/postulations are based on an information flow controlled by the manufacturers.
Probably different exposures, lenses so for comparisons assume it is the best the respective cameras are capable of considering its a kind of marketing release of sorts.
Price at our local store here in Japan. Must admit not cheap but as a holder of a foreign passport, you can get an additional 5% off the price + member discount of up to 10%.
The Nikon section is always full of people here so the Japanese do love their Nikons.
BTW, they are beginning to clear out 7Ds. Don't know why but not very popular here. I asked the guys about 4 weeks ago when they started reducing prices for both 7D and 5D II and then just last week, the new 5D came out. "I know nothing - sorry" says the guy at the Canon counter.
It puts to rest the notion the 5D mark iii will outperform the D800 in low light high ISO RAW images. Canon has a more aggressive Digic 5 noise reduction that seems to work very well in jpegs at the cost of some miniature details. So jpeg performance seems to favour Canon. If you shoot jpeg a lot of the time that may be a worthwhile consideration.
The D800 has higher dynamic range by 2 stops, higher sensor QE so ISO's don't quite match up in comparisons and also the ability to be downsized to 22mp from 36mp to further reduce noise in jpegs so the 5D mark iii jpeg noise advantage will not be by as much as you would think for those reasons.
So you can have your high megapixel cake and eat it too!
Either way both cameras seem to be sensational.
The real super high performance low light cameras seem to be the 1DX or the Nikon D4. Those 2 cameras are much more evenly matched and if it were my choice I'd pick the 1DX.
The choice may be more determined by how much money you have invested in lenses.
Greg.
Last edited by gregbradley; 15-03-2012 at 10:10 AM.
I'll /still/ be happy with my purchase. The Mark II makes fantasticpictures, and, the Mark III will make just as good an image.
Re: that thread: once again, using non-Canon RAW converter on a CanonRAW file means nothing to me. I will wait and see what the results looklike when converted using DPP. I once used dcraw to convert a 5DMark II RAW on my Mac and it looked like garbage. I tried bothportraits and landscape images to see what would happen. Neveragain.
You'd have to ask why Canon hasn't released the RAW converter at the same time as the camera? If you ask me its deliberate to prevent comparisons so they can claim things that only an educated public would see as being misled. Like posting only jpeg examples of low light performance and not RAW. I find that very off-putting.
I am not neccessarily a Nikon fanboy either as their smoothing of RAW data (I assume they are still doing that) is not good either.
Its funny in a way because Nikon has been the one for years arguing the benefits of low light performance and low megapixels, now the roles are reversed not that 22mp is low megapixels.
One thing I think anyone can agree with here is the new Sony Exmor sensor is an amazing engineering feat and it would appear Sony are ahead of the rest at this point. 5 years from now they may be the main DSLR everyone compares against at this rate.
