I've not been able to reproduce the low ISO issue that you are experiencing. However, I use Adobe Camera Raw, you may already have dealt with this already, but do you get the same issue with different RAW converters?
Due to weather and lunar phase considerations I haven't had a proper chance to put my recently aquired 5D2 through its paces for my night photography (high ISO!), but I may be able to get out in a week or so. My preliminary tests with my 24L, only from the back garden, are making me really excited about the 5D2's night performance, this could be the camera I've been waiting for all decade. Satisfying 24L shots of star fields look distinctly possible at f2, 10 seconds and 3200 ISO with NO BANDING. Wow, the Holy Grail of photography for me!
The trend with DSLR's is to make denser sensors fo increased MP while at the same time acheiving better and better high ISO capabilities ( 2 seemingly conradictory things). There must be some trade off to other image qualities if you have to put the resulting sensor data though the camera's internal processing 'mangle' to achive this compromise and this is before it gets to RAW.
The fact that both cams share a 21MP sensor, doesn't mean both cams share the same IQ. The 1Ds3 is just a few steps beyond the 5D2 IMHO. Of course the 1Ds3 costs two times more than the 5D2...
I'm not going to contest at this time that the 5D2 image isn't as good as the 1Ds3 image, but your reasoning is invalid. That's like saying my current $1500 laptop is an inferior data processing machine compared to my $5,000 1985 IBM PC.
I've not been able to reproduce the low ISO issue that you are experiencing. However, I use Adobe Camera Raw, you may already have dealt with this already, but do you get the same issue with different RAW converters?
Hi David
I've not tried any other raw convertors yet, and I may not, since that will cost money, and I hope that Canon will do something to reassert DPP as one of the best convertors for my camera. However, I can see that samples on this thread processed through ACR seem to be a lot better than what I am getting.
The only issue I have about all this is that Canon should not have claimed that the 5D2 has superior image quality to a 1Ds3 if it isn't really true, at least below iso 800. Canon's statement influenced my decision to move from 1Ds3+450D to a pair of 5D2's. In fact, on balance, the 5D2's do suit me better, but low iso performance has been a bit of a fly in the ointment.
brainiac wrote:
The only issue I have about all this is that Canon should not have claimed that the 5D2 has superior image quality to a 1Ds3 if it isn't really true, at least below iso 800.
I think you probably hit the nail on the head right there. You can barely read a camera review published in the last two-three years, especially since the D3 came out, that doesn't talk extensively about high ISO performance. They've always been about sharpness too. So Canon says their camera has the best image quality ever. They probably meant at the high resolution, high ISO end of things, because that's what the market wants right now.
Question - does low noise at high ISO come at the expense of higher noise at low ISO? Can chips/cameras optimized for one at the expense of the other?
RDKirk wrote:
I'm not going to contest at this time that the 5D2 image isn't as good as the 1Ds3 image, but your reasoning is invalid. That's like saying my current $1500 laptop is an inferior data processing machine compared to my $5,000 1985 IBM PC.
Agreed... Money shouldn't have anything to do with it (it didn't for me actually). One camera is just better than the other IMHO.
I think you probably hit the nail on the head right there. You can barely read a camera review published in the last two-three years, especially since the D3 came out, that doesn't talk extensively about high ISO performance. They've always been about sharpness too. So Canon says their camera has the best image quality ever. They probably meant at the high resolution, high ISO end of things, because that's what the market wants right now.
Chuck Westfall is the one who said the 5D2 is the best current imager; Chuck's statements often have to be parsed with care, but I haven't discovered him ever lying.
I think he probably meant that given equal and optimum exposure and processing, the 5D2 image is slightly better than the 1Ds3 image.
And, indeed, that's the way it was with film cameras, too. It's when the situation is not optimum that the price differential between cameras has always become manifest.
I'm not going to contest at this time that the 5D2 image isn't as good as the 1Ds3 image, but your reasoning is invalid. That's like saying my current $1500 laptop is an inferior data processing machine compared to my $5,000 1985 IBM PC.
That's not a fair comparison ( and thus an invalid criticism IMO) - you are talking about two items seperated by a huge advance in technology. I'm talking about an advance of only one generation. If Canon have improved every area of perfromance without compromise to another and maintained the price (actually lowered it allowing for 3 years inflation) then we all have nothing to worry about with the 5D vs 5DII. If they have replicated the image quality of the 1DsIII in a smaller package at less than 1/2 the price ( something they didn't do with the 5D vs 1DsII BTW) it would be astonishing.Lets wait and see some more tests.
Shirozina, thank you. Very well articulated! Agree 100%.
I don't find this hard to believe at all. Many people said the image quality ( not resolution, not noise performance) i.e the sophistication and realism of the tonal reproduction of the scene was superior on the original 1ds than on the improved 1dsII. All the conventional tests would say other wise though. I find the sophistication of the tonal rendering of most DSLR's to be quite crude compared to MF digital. They simply cannot render subtle tonal and color variations in a scene with the realism a good MF digital back can - maybe this has many factors - CCD vs CMOS, 12/14 vs16bit, lack of an AA filter, A/D conversion? but it does show that these things cannot be easily measured with test charts. The trend with DSLR's is to make denser sensors fo increased MP while at the same time acheiving better and better high ISO capabilities ( 2 seemingly conradictory things). There must be some trade off to other image qualities if you have to put the resulting sensor data though the camera's internal processing 'mangle' to achive this compromise and this is before it gets to RAW. Years ago there was a big thread with many claiming that the 5D was just as good at the 1DsII in regards to resolution ( and therefore image quality) and many side by side crops were produced to prove this claim. Again after many year s working with the image files from these 2 cameras it is clear that in other areas of image quality the 1DsII is superior such as less noise in the shadows at low ISO means the 1dsII image can be 'pushed' much more , less 'jaggies' on angled lines and edges ( less sopisticated i.e cheaper AA filter on th 5D?) means the 1DsII image can be upressed more. I'm sure the 5DII is a very good camera for the money but I'm also sure Canon don't just charge conciderably extra for the 1DsIII for a bigger body and tighter manufacturing tollerances.
RDKirk wrote:
I think he probably meant that given equal and optimum exposure and processing, the 5D2 image is slightly better than the 1Ds3 image.
He probably was referring to JPEG's... The 5D2 have a better JPEG engine than the 1Ds3 IMO. Less noise while still retaining very good detail. Not that the 1Ds3 JEPG engine is bad, the 5D2 is just a little better. In fact, if I still had a 5D2, I would probably shoot mostly JPEG, because I would get a better output than when converting RAW in LR/ACR. Just another opinion of course
Chuck Westfall specifically stated that the RAW image quality of the 5DII would be superior to the 1Ds MK III in the November 04, 2008 Q&A session on the PhotographyBLOG website...
"You’ll be able to see the difference in RAW image quality between the EOS 5D Mark II and the EOS-1Ds Mark III for yourself within the next month or so, but most of the differences in this comparison will derive from the improvements that Canon made to the CMOS image sensor of the 5D Mark II. These include a more transmissive color filter array to improve S/N ratios, a more advanced output amplifier design, and a new method of on-chip noise reduction that occurs before the signal is output to the DIGIC 4 processor."
gman1339 wrote:
Chuck Westfall specifically stated that the RAW image quality of the 5DII would be superior to the 1Ds MK III in the November 04, 2008 Q&A session on the PhotographyBLOG website...
"You’ll be able to see the difference in RAW image quality between the EOS 5D Mark II and the EOS-1Ds Mark III for yourself within the next month or so, but most of the differences in this comparison will derive from the improvements that Canon made to the CMOS image sensor of the 5D Mark II. These include a more transmissive color filter array to improve S/N ratios, a more advanced output amplifier design, and a new method of on-chip noise reduction that occurs before the signal is output to the DIGIC 4 processor."
That's not a fair comparison ( and thus an invalid criticism IMO) - you are talking about two items seperated by a huge advance in technology. I'm talking about an advance of only one generation. If Canon have improved every area of perfromance without compromise to another and maintained the price (actually lowered it allowing for 3 years inflation) then we all have nothing to worry about with the 5D vs 5DII.
No, you're being unfair. If Canon had increased performance in any area and yet reduced price in absolute dollars (not to mention accounting for inflation), that would be a true improvement of the 5DII over the 5D. Even if they had changed nothing but reduced the price, that would have been an improvement.
In fact, they have increased performance in a number of areas and yet reduced the price of the camera.
R. Eisenberg wrote:
Is now a good time to take the plunge for a 1Ds3
I already have the 1Ds3, I was looking at the 5D2 as a backup. In the end after looking at the files, I passed for now as I do not need a back up for another month or so...
Sample variation is a known issue with Canon lenses so is there any reason not to believe that a lot of what we are seeing here could just be sample variation with sensors?
Lots of hair splitting going on here. Having used or owned to some degree nearly all Canon bodies released in the last 5-6 yrs and being a former Nikon user...I'd just say that some people are focusing and trying way too hard to justify the cost difference here. I cannot think of a single instance (PERSONALLY) where the difference in the 1DsM3 and the 5D2 image would translate to something that would make me more money and/or all but the most retentive pixel peepers could tell the difference. Sure I understand that many will disagree...But I am just stating the same...I disagree that there is really much to talk about here.
I cannot think of a single instance (PERSONALLY) where the difference in the 1DsM3 and the 5D2 image would translate to something that would make me more money
Daan B wrote:
All I can say it doesn't correspond with my own findings... 1Ds3 RAW files are in a different league IMO... Sounds like the usual marketing BS to me
Will Patterson wrote:
A different league? That's a little drastic.
I used the phrase IMO
But to me it is... Not by looking at the individual parameters as sharpness or noise (which are pretty comparable anyway), but more by looking at the overall quality of the output. I guess this has more to do with colors/tone curves/contrast/metering/exposure. I have said it before... the 1Ds3 (RAW) files look almost "magical" to me. It is something that is hard to describe, but I see the difference with 5D2 (RAW) files that lack that "magical" quality.
Andrew Gough wrote:
I already have the 1Ds3, I was looking at the 5D2 as a backup. In the end after looking at the files, I passed for now as I do not need a back up for another month or so...
Same here... I am also curious if PMA will bring something worthwhile...
Daan B wrote:
I have both the 5D2 and 1Ds3... I will do some comparison tests when I have the time... but right now my feeling is that the 1Ds3 has the edge at lower ISO's and the 5D2 has the edge at higher ISO's (ISO1600/3200). Besides that, the 1Ds3 contains more shadow detail. Shadows in 5D2 files are easily clipped and contain nothing more than noise and other ugly pollution when you open them in PP. Even at low ISO's. Not so with the 1Ds3...
The 1Ds3 is a wonderful camera that produces excellent files with an almost "magical" quality. The 5D2 also produces nice files, but in a different league IMO. The fact that both cams share 21MP's doesn't mean they share the same IQ (appearantly). I prefer the 1Ds3 files over the 5D2 files any time ...Show more →
According to DPreview's review of the 1Ds Mark III, its RAW headroom has about 11 stops of exposure latitude where the 5DII has just over 10 stops of exposure headroom. I will be picking up a 1DsIII when its successor comes out. I can't wait to own a 1Ds body to compliment my classic 5D
