Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.

OreoCookie

macrumors 68030
Apr 14, 2001
2,727
90
Sendai, Japan
Basically, the same pixel pitch as the old 5D (which produces very usable images at ISO3200), but with the 7D's AF system. Or, put another way, a Nikon D700.
I absolutely agree. Canon's decisions seem to be driven by marchitecture rather than putting the best cameras out there it can deliver. It'd be easily possible to use the 7D's AF system and put it into the 5D Mark II. Don't get me wrong, it is a very good camera, but it could have been a better camera at little extra cost. I mean, the Canon EOS 3 which debuted in 1998 had 45 (!) AF sensors and was similarly priced.

IMO the 7D would have been much more exciting if they would have kept the pixel count lower, e. g. by reusing the 1.3x crop sensor found in the 1D Mark III. That would have allowed Canon to offer a juicy upgrade compared to the 40D/50D and improving image quality as the photo sites are larger.

BTW, I've just printed a 3 MP (= cropped 4 MP) picture I've taken in late 2003 and printed it on canvas. Looked great. I don't think I need 21 MP, at least it's not an incitement for me to upgrade.
 

Edge100

macrumors 68000
May 14, 2002
1,562
13
Where am I???
IMO the 7D would have been much more exciting if they would have kept the pixel count lower, e. g. by reusing the 1.3x crop sensor found in the 1D Mark III. That would have allowed Canon to offer a juicy upgrade compared to the 40D/50D and improving image quality as the photo sites are larger.

This is basically what the 1DmkIV is; albeit at a MUCH higher price. From what I've seen, the noise performance of the 1DmkIV, at both the pixel and image levels, is as good as I've ever seen from a Canon DLSR.
 

Chris7

macrumors 6502
Original poster
Aug 8, 2008
396
0
Lost in Thought
...I went from a Rebel (XSi) to a 7D, and would agree that there is a difference of about two stops starting at ISO 800 (maybe one stop at most at ISO 400).
Thanks. Still learning photo jargon... Are you saying here that your 7D at ISO 800 produces about the same noise as your Rebel XSi does at ISO 400; and that your 7D at ISO 3200 produces about the same amount of noise as your Rebel XSi at ISO 800? Please correct me if I'm not understanding you here..
...If you resize a 5DmkII image and a Rebel XS image both to 800 pixels, you will notice far less noise based purely on binning twice as many pixels; those cameras could have precisely the same pixel-level noise and you'd still see that difference.

This is why looking at 100% crops is meaningless for real-world use.
This is a good point, and not one I had considered. I would seldom crop the picture down more 1/2 to 1/4 of its original size for real world use. So when I mean "noise" or "grain" I should specify that I am referring at most a 1/4 crop of the photo rather a 100% crop of the pixels.
As far as my personal subjective testing goes, this is it:
xxxD series: ISO 400 is acceptable
xxD series: ISO 800 is acceptable
xD series: ISO 1600 is acceptable...
Thanks. Could you maybe put some model numbers in so I can better understand?:)
there is a difference, but it is not "x" stops, and it is not a constant between camera series.
there is a slight difference from Rebels to 40/50D
1-2 stop difference from 50D to 7D
1-2 stop difference from 7D to 5DII...
Thanks. So, is it closer to 1 or 2 stops difference?
 

Edge100

macrumors 68000
May 14, 2002
1,562
13
Where am I???
This is a good point, and not one I had considered. I would seldom crop the picture down 1/2 to 1/4 of its original size for real world use. So when I mean "noise" or "grain" I should specify that I am referring to this.

In fact, you'll probably seldom use the full-sized image output. A 250 ppi print of a 21MP image works out to be 22.5 x 15" print (5,616 × 3,744 pixels). How often do you print that large? Moreover, how often do you look at those prints from 1-2 inches away? That is the exact equivalent of what you are doing when you look at actual pixels (i.e. 100% crops) on screen.

This is the key issue to remember. Pixel-level noise is only relevant when it starts to cause problems at the image level. And as you can see from my posted samples, at normal image level, even a 6-year-old camera produces acceptable results (i.e. someone paid money for them).
 

Edge100

macrumors 68000
May 14, 2002
1,562
13
Where am I???
Thanks. Still learning photo jargon... Are you saying here that your 7D at ISO 800 produces about the same noise as your Rebel XSi does at ISO 400; and that your 7D at ISO 3200 produces about the same amount of noise as your Rebel XSi at ISO 800? Please correct me if I'm not understanding you here..

This is a good point, and not one I had considered. I would seldom crop the picture down more 1/2 to 1/4 of its original size for real world use. So when I mean "noise" or "grain" I should specify that I am referring at most a 1/4 crop of the photo rather a 100% crop of the pixels.

Thanks. Could you maybe put some model numbers in so I can better understand?:)

Thanks. So, is it closer to 1 or 2 stops difference?

Based on what I've seen reported, I'd say that at identical image sizes, the 5DmkII has a one stop advantage over the 7D, which, again identical image sizes, has less than a one-stop advtange over the 40D. The 40D, to my eyes, has about the same amount of image-level noise as the 20D had. The Rebel T1i shows a bit less image-level noise than the 50D, which is about the same as the 40D (again, at image level). The original 5D has lower image-level noise than probably any of these DSLRs, with the exception of the 5DmkII, which it trails by probably 1 stop.

The 1DmkIV is the best high ISO noise performer of any Canon DSLR I've seen, with probably 1/2 to 1 stop over the 5DmkII.

So I would say, again based on what I've seen around, it goes (at image-level):

40D=50D<Rebel T1i<7D<5D<5DmkII<1DmkIV

I havent seen a lot of Rebel XS images, but my guess is that it's similar to the 40D (same 10.1MP sensor).

Note that even this list is misleading. The 5D may be 1/2 to 1 stop better than the 7D, but the 5D tops out at ISO3200, so if you need ISO6400, the 5D wont help you (you could push the RAW, of course, but that's just amplifying noise). Plus, you may get a lot more detail in the 7D files, because it's 18MP vs. 13MP, so with a little noise reduction, the 7D might produce more detailed final images.

See how complicated it is? You can't just boil down "noise performance" to a single number.
 

Phrasikleia

macrumors 601
Feb 24, 2008
4,082
403
Over there------->
Thanks. Still learning photo jargon... Are you saying here that your 7D at ISO 800 produces about the same noise as your Rebel XSi does at ISO 400; and that your 7D at ISO 3200 produces about the same amount of noise as your Rebel XSi at ISO 800? Please correct me if I'm not understanding you here..

I think that's more definitive than I want to be, but in the most extreme cases (lots of even, dark areas), yes that's about right. You have to keep in mind that resolution really matters for me and is a factor in the equation. I have to interpolate my XSi images to 50MB for submission to my stock agency to meet their minimum. That magnifies any noise or loss of detail in the images. The 7D produces files that are already large enough; so the comparisons I'm making are at the 100% crop level, but after the images have been normalized to the native size of the 7D.

Like I said, the big picture is very complicated and ultimately different for different photographers.
 

OreoCookie

macrumors 68030
Apr 14, 2001
2,727
90
Sendai, Japan
This is basically what the 1DmkIV is; albeit at a MUCH higher price. From what I've seen, the noise performance of the 1DmkIV, at both the pixel and image levels, is as good as I've ever seen from a Canon DLSR.
Bingo. IMO the 7D should be what the 1D Mark IV what the D700 is to the D3(s).
 

Phrasikleia

macrumors 601
Feb 24, 2008
4,082
403
Over there------->
Bingo. IMO the 7D should be what the 1D Mark IV what the D700 is to the D3(s).

Should be?? For whom? I suppose you also think that the D300s should be APS-H as well?

For the price, the 7D is highly competitive right now. Its sensor strikes just the right combination for people who need the resolution and/or the extra "reach." It may not be the perfect combination for you, but with all of the bodies on the market right now, surely there's one that is.
 

Edge100

macrumors 68000
May 14, 2002
1,562
13
Where am I???
Should be?? For whom? I suppose you also think that the D300s should be APS-H as well?

For the price, the 7D is highly competitive right now. Its sensor strikes just the right combination for people who need the resolution and/or the extra "reach." It may not be the perfect combination for you, but with all of the bodies on the market right now, surely there's one that is.

True. For me, the best body "on the market" isn't even on the market. It's the original 5D. Wonderful IQ, usable images up to ISO3200, and nice and small (relative to the 1D-series). It strikes the perfect balance between cost and performance. Sure the AF system isn't up to snuff in dodgy lighting, but I have my 1DmkII for that.

I continue to encourage everyone I know who is in the market for anything except an entry-level DLSR to strongly consider a gently used 5D.
 

OreoCookie

macrumors 68030
Apr 14, 2001
2,727
90
Sendai, Japan
Should be?? For whom? I suppose you also think that the D300s should be APS-H as well?
For competition: I would like to see FF bodies become more affordable. And yes, I would like to see Nikon drop the price of the D700 (or similar) to where the D300s is now.

Although I don't lust for a full frame body because of noise, but because of the smaller depth of field (if equivalent focal lengths are compared).
 

Chris7

macrumors 6502
Original poster
Aug 8, 2008
396
0
Lost in Thought
I think that's more definitive than I want to be, but in the most extreme cases (lots of even, dark areas), yes that's about right...
Many thanks.
Based on what I've seen reported, I'd say that at identical image sizes, the 5DmkII has a one stop advantage over the 7D, which, again identical image sizes, has less than a one-stop advtange over the 40D...
I havent seen a lot of Rebel XS images, but my guess is that it's similar to the 40D (same 10.1MP sensor)...
Thank you for your detailed responses. Also, when you say that the 7D "has less than a one-stop advtange over the 40D," would you guestimate this to be about a 2/3 stop advantage?

Would the people here agree with Edge's assessment?

As an aside, does the Nikon D700 have any low light advantage over the 5D Mk II?
 

funkboy

macrumors regular
Apr 25, 2008
179
11
elsewhere
The best way to compare noise is to take the 18MP image and the 10MP image and make a JPEG or print of the same size from each (i.e. ignore 100% crops). Only then can you make an accurate comparison.

I contend that unless you're making very large prints or have very, very picky clients, it's not really a big deal. There are far more important things to concern yourself with when it comes to taking a photograph.

Thank you Thank you Thank you

The signal-to-noise ratio (by definition) is always lower in the shadows of an image. The fact that there's any useful shadow detail at all in high-resolution high-pixel-density cameras like the 7D at ISO6400 (meaning that you're 6x 6 stops more sensitive to light than at ISO100) continues to amaze me.

Pretty much every Canon DSLR made since (& including) the 20D really has image quality good enough for the kind of professional work previously done with the equivalent ISO pro-grade 35mm format film. The 1DsII of that era was good enough for landscape photographers (& others making larger prints) to start switching away from their medium-format film systems that allow them to make high-quality A3 & larger prints.

Anything better than that is a bonus, and has opened the door for photographers around the world to reach beyond what was possible in the film days. It's also created a storm of noise from people that spend more time in photography forums than they do taking & working on pictures.

High ISO performance is but one piece of the puzzle. If you're really going to be shooting available-light stuff most of the time (concerts, weddings, etc) or doing it for money then something with a bigger sensor like the 5DII is probably the way to go. Honestly if you're not making prints bigger than A3 or you're not working for paying clients then a good-handling solid camera like the 40D will likely suit your needs very very well. It has for me for a very long time; I traded up to the 40D after 4 years traveling Europe with my 10D, which is now with my brother in Africa & still making great images almost 7 years after I bought it.

Get a good camera. Then spend your time learning how to take & process good pictures, not worrying about noise.
 

toxic

macrumors 68000
Nov 9, 2008
1,664
1
Would the people here agree with Edge's assessment?

pretty much.

As an aside, does the Nikon D700 have any low light advantage over the 5D Mk II?

no, or at least not substantial enough to matter.

another thing to consider is whether it's even worth the money to be chasing after the "best" high-ISO performance. if you only upload images online...it isn't. noise is an issue for those who give/sell high-res digital images (e.g. wedding/event photographers) and for those who print. everything looks good on the internet since everything is tiny.

what is worth chasing after (for RAW shooters, anyway) is dynamic range, and Canon isn't anywhere near pushing the boundaries in this area.
 

funkboy

macrumors regular
Apr 25, 2008
179
11
elsewhere
As an aside, does the Nikon D700 have any low light advantage over the 5D Mk II?

It depends on how big your prints are. The 5DII has almost double the resolution of the D700. Scaled to the same size, the 5DII would probably look better as the D700 has slightly better per-pixel low-light performance but the 5DII has a whole lot more pixels.

But, really, they're both great cameras, and one would be silly making the choice between them based solely on noise performance. Do you need the 5DII's extra resolution and video features? Do you already own Nikon glass or just like the feel of the D700 better?

To slip into the oft-used car analogy, your seem to be asking someone to choose between a Lamborghini and an Aston Martin based on the 0-60 time of the car alone. It's certainly a factor, but only a very small piece of the much bigger pie. They're both fabulous cars & mind-blowingly fast, so you're much better off just choosing the car you feel happier with that better suits your needs all-round.
 

funkboy

macrumors regular
Apr 25, 2008
179
11
elsewhere
I continue to encourage everyone I know who is in the market for anything except an entry-level DLSR to strongly consider a gently used 5D.

Great suggestion. In France anyway, a 5D in good shape will run you about 1300 euros, and something a bit more beat up will be about 1100. If you don't need hot-shot autofocus & high frame rates for sports or fast-moving wildlife then it's really hard to do better for the same money.

A friend of mine just had a gallery opening with ~40 prints she made after spending a few weeks in Argentina with her 5D. Some of them are A3 size, and they all look fabulous.
 

OreoCookie

macrumors 68030
Apr 14, 2001
2,727
90
Sendai, Japan
As an aside, does the Nikon D700 have any low light advantage over the 5D Mk II?
Yes, it does. The question is: does it matter to you? In all likelihood no. Sort of like arguing on a school ground whether your dad's car is faster on a track than your neighbors. Digital cameras are like that, too, it's not easy to find a metric to compare them to.

One attempt is the Dxomark of these two cameras. As with all synthetic benchmarks, they have to be supplemented by real-world benchmarks.

If you compare the numbers and figures, you will see that the D700 is consistently better. Take the noise chart, for instance. Here, the signal-to-noise ratio is plotted in decibel (as is customary). An increase by 10 decibels corresponds to a 20-fold increase (at first I used the usual convention, but upon research, there is a different convention when it comes to image sensors). The difference in noise is consistently between about 2.5 and 2.9. This translates to 10^0.125-10^0.145 = 1.33-1.40, i. e. approximately 1 EV. The 5D Mark II has 75 % more pixels allowing the prints to be 32 % larger on either side (1.32 x 1.32 = 1.75). So the increase in noise is comparable to the increase in print size. That's hardly surprising, given the fact that both cameras use state-of-the-art sensors which are limited by the same laws of physics.

So noise-wise, you could say that both are more or less equal. But not only noise increases if the pixel size shrinks, other things such as dynamic range and tonal range are affected. This is something that cannot be compensated for by scaling. Here, the D700 also has an edge. Again, if it is significant, that's a different question.

There are other factors that haven't been taken into consideration at all, e. g. the AF system, how does that work in low light. Arguably, the D700 features a much more modern and sophisticated AF system. (Do you need it?)

Again, I digress that I've used only one synthetic benchmark (for simplicity) to make this argument. You shouldn't compare cameras you won't buy too much anyway, unless you expect some of the technology to trickle down. Nikon and Canon have simply designed the two cameras for different things. I don't doubt that if Canon had attempted to design the 5D Mark II for the same niche as the D700, they would have made a very competitive camera (lower resolution, better AF system).

If I were you, I wouldn't pixel-peep too much and stay away from generalizations such as A has better lenses, B has a better body. In 95 % of the cases they are plain-out wrong. In all likelihood, the camera you use is no longer the limiting factor, today's dslrs are in many things way better than their analog predecessors. 4.5 fps is something `normal' for advanced consumer cameras. A little more than 10 years ago, this was something only pro and semi-pro bodies could do (e. g. the Nikon F100 was capable of 4.5 fps ditto for the Canon EOS3 and EOS1V if you don't use battery grips).

On a side note: I've just had a 3 MP pixel pic (a crop from a 4 MP image) printed on canvas (40x75 cm^2). Turned out great, no complaints in terms of image quality. I've even added noise ;) You don't need the latest and greatest to take good photographs.
 

Edge100

macrumors 68000
May 14, 2002
1,562
13
Where am I???
The fact that there's any useful shadow detail at all in high-resolution high-pixel-density cameras like the 7D at ISO6400 (meaning that you're 6x more sensitive to light than at ISO100) continues to amaze me.

I agree, but just so we're all being completely accurate here, ISO6400 is 2^6 times (or 64x) more sensitive to light than ISO100, not 6x.

Pretty much every Canon DSLR made since (& including) the 20D really has image quality good enough for the kind of professional work previously done with the equivalent ISO pro-grade 35mm format film. The 1DsII of that era was good enough for landscape photographers (& others making larger prints) to start switching away from their medium-format film systems that allow them to make high-quality A3 & larger prints.

I've seen excellent poster-sized prints from images shot on a original Canon 1D (a 4MP camera released in 2001), so it's certainly possible. That said, there is a great deal of benefit to be realized from higher MP counts, particularly in the ability to resolve small details. A 50MP PhaseOne is going to out-resolve even the best 35mm DLSRs; sure, it poops out at ISO400, but if you need the absolute highest degree of resolution currently possible, you gotta pay the big money! Likewise, medium format film, shot with a high quality body/lens system, and then drum scanned, will outresolve a 5DmkII or D3x. But that's not practical for most people, and the differences are unlikely to matter to all but the most finicky of clients.

Anything better than that is a bonus, and has opened the door for photographers around the world to reach beyond what was possible in the film days. It's also created a storm of noise from people that spend more time in photography forums than they do taking & working on pictures.

Couldn't agree more. What is possible today on a sub-$1000 entry level DSLR coupled with a decent lens is absolutely astounding. We're arguing about ISO3200, for goodness sake. Go over to the Flickr group for Ilford Delta 3200, and see what kind of "noise" you get with ISO3200 film. Or check out Ektachrome P1600 color chromes. The noise/grain on these films is huge. Granted, digital noise often doesn't look quite as nice as film grain, but the point is that what we can achieve in low light was unheard of in the film era.

High ISO performance is but one piece of the puzzle. If you're really going to be shooting available-light stuff most of the time (concerts, weddings, etc) or doing it for money then something with a bigger sensor like the 5DII is probably the way to go. Honestly if you're not making prints bigger than A3 or you're not working for paying clients then a good-handling solid camera like the 40D will likely suit your needs very very well. It has for me for a very long time; I traded up to the 40D after 4 years traveling Europe with my 10D, which is now with my brother in Africa & still making great images almost 7 years after I bought it.

I learned on a 10D. 6MP of sweet camera. Slow as hell, tiny RAW buffer, but produced excellent quality images at ISO400 or below.
 

Edge100

macrumors 68000
May 14, 2002
1,562
13
Where am I???
Yes, it does. The question is: does it matter to you? In all likelihood no. Sort of like arguing on a school ground whether your dad's car is faster on a track than your neighbors. Digital cameras are like that, too, it's not easy to find a metric to compare them to.

One attempt is the Dxomark of these two cameras. As with all synthetic benchmarks, they have to be supplemented by real-world benchmarks.

If you compare the numbers and figures, you will see that the D700 is consistently better. Take the noise chart, for instance. Here, the signal-to-noise ratio is plotted in decibel (as is customary). An increase by 10 decibels corresponds to a 20-fold increase (at first I used the usual convention, but upon research, there is a different convention when it comes to image sensors). The difference in noise is consistently between about 2.5 and 2.9. This translates to 10^0.125-10^0.145 = 1.33-1.40, i. e. approximately 1 EV. The 5D Mark II has 75 % more pixels allowing the prints to be 32 % larger on either side (1.32 x 1.32 = 1.75). So the increase in noise is comparable to the increase in print size. That's hardly surprising, given the fact that both cameras use state-of-the-art sensors which are limited by the same laws of physics.

So noise-wise, you could say that both are more or less equal. But not only noise increases if the pixel size shrinks, other things such as dynamic range and tonal range are affected. This is something that cannot be compensated for by scaling. Here, the D700 also has an edge. Again, if it is significant, that's a different question.

All true. Of course, a 21MP FF sensor will outresolve a 12MP FF sensor (as you can see here), so there are trade offs with lower pixel counts too.

There are other factors that haven't been taken into consideration at all, e. g. the AF system, how does that work in low light. Arguably, the D700 features a much more modern and sophisticated AF system. (Do you need it?)

It really is amazing that Canon did not upgrade the AF system in the 5DmkII. The 5D AF system isn't bad, especially for studio and landscape work, but it's really not suitable for fast action.

If I were you, I wouldn't pixel-peep too much and stay away from generalizations such as A has better lenses, B has a better body. In 95 % of the cases they are plain-out wrong. In all likelihood, the camera you use is no longer the limiting factor, today's dslrs are in many things way better than their analog predecessors. 4.5 fps is something `normal' for advanced consumer cameras. A little more than 10 years ago, this was something only pro and semi-pro bodies could do (e. g. the Nikon F100 was capable of 4.5 fps ditto for the Canon EOS3 and EOS1V if you don't use battery grips).

Yup. Again, modern DSLRs, even at the very low end, are doing things that were unheard of in the film era, and are capable of producing absolutely stellar results, when coupled to good glass.

I propose that everyone who concerns themselves primarily with whether camera A is X stops better at ISO bazillion than camera B should ask themselves this question: do I own at least one fast prime lens? Are all of my zooms capable of a constant f/2.8? If the answer is "no", then the sensor noise issue is not the limiting factor; stick with the DSLR you have and buy faster glass...high ISO noise problem solved. On the other hand, if you're shooting your 85/1.2L wide open and still having to use ISO6400, then, and only then, is it time to consider sensor noise as a key factor. Until then, your money should go into glass.
 

OreoCookie

macrumors 68030
Apr 14, 2001
2,727
90
Sendai, Japan
All true. Of course, a 21MP FF sensor will outresolve a 12MP FF sensor (as you can see here), so there are trade offs with lower pixel counts too.
Certainly, that was the trade-off Canon wanted to make when designing the 5D Mark II.
It really is amazing that Canon did not upgrade the AF system in the 5DmkII. The 5D AF system isn't bad, especially for studio and landscape work, but it's really not suitable for fast action.
That was a really, really stupid move that has angered a lot of Canonites. This was clearly marchitecture and while you're right that not everybody needs a better AF system, nobody would have complained if Canon had used a state-of-the-art AF system. That said, I think the 5D's AF system would be plenty for me.
I propose that everyone who concerns themselves primarily with whether camera A is X stops better at ISO bazillion than camera B should ask themselves this question: do I own at least one fast prime lens?
Not only that, we are used to limits being pushed. In the film days, it was ridiculous to think anyone could buy an ISO 12,800 film! (Perhaps for some very, very special scientific applications such films may have existed, the fastest films I've seen were rated ISO 3,200. To my knowledge, they were ~ISO 1,000 films that could be pushed to ISO 3,200 -- talk about film grain :))

Flaws of optics that most users in the film days did not see, because they weren't blowing everything up to ridiculous sizes remained undiscovered. Nowadays, it takes no patience and skills to peruse your digital `negatives' at 100 % magnification.

To be honest, I think the development of dslrs has peaked. What else can the manufacturers include? They're trying to push video, but I'm not really into that. I'm hoping that this pointless pixel/fps race has come to an end and they're trying to make good technology more affordable instead.
 

Edge100

macrumors 68000
May 14, 2002
1,562
13
Where am I???
To be honest, I think the development of dslrs has peaked. What else can the manufacturers include? They're trying to push video, but I'm not really into that. I'm hoping that this pointless pixel/fps race has come to an end and they're trying to make good technology more affordable instead.

Now there's a topic; where do we go from here with DSLRs? What will a pro-level DSLR look like in 5 years?

First off, it would seem that we've hit the limit of the resolution/noise battle; we can cram more pixels onto a FF sensor, but noise is going to start being a problem. There is this new CMOS technology developed by Sony (http://www.sony.net/SonyInfo/News/Press/200806/08-069E/index.html), which could help bring noise levels down by a stop or so. This would likely allow 30+ MP FF DSLRs to have the same noise performance as today's 21-24MP DSLRs. For APC-S sensors, I'm not sure where things can go. 18MP seems to be past the limit where noise begins to really creep in.

Alternatively, it's possible we could see a whole new lens system developed, such that MP counts could increase, thereby increasing resolution, but keeping pixel pitch the same or even lower. This seems unlikely though; are pro users going to pay the big bucks to move their lens collection over to a larger sensor format? Seems unlikely.

Will be very interesting to see what happens as we move forward.
 

Chris7

macrumors 6502
Original poster
Aug 8, 2008
396
0
Lost in Thought
Then spend your time learning how to take & process good pictures, not worrying about noise.
I thought if I just had a great camera and lenses I would automatically be a great photographer.;)
It depends on how big your prints are. The 5DII has almost double the resolution of the D700. Scaled to the same size, the 5DII would probably look as the D700 has slightly better per-pixel low-light performance but the 5DII has a whole lot more pixels.
Thanks.
what is worth chasing after (for RAW shooters, anyway) is dynamic range, and Canon isn't anywhere near pushing the boundaries in this area.
This is really interesting. Could you please say more or provide some links? Also just so I understand you correctly, I understand dynamic range to be the range between light and dark that in a photo (without masking and adjusting different parts of a RAW separately) and latitude to be the amount you can correct over or underexposure (e.g. a RAW file would generally have more latitude, but not necessarily more dynamic range than a TIFF). So with colloid film, a single print will have a much more dynamic range than digital photos, and would also have much more latitude in the negative (you could probably push or pull it two stops), but would not have more latitude in the print. But please correct me if my terms are off.
...So noise-wise, you could say that both are more or less equal. But not only noise increases if the pixel size shrinks, other things such as dynamic range and tonal range are affected. This is something that cannot be compensated for by scaling. Here, the D700 also has an edge... Again, if it is significant, that's a different question.

There are other factors that haven't been taken into consideration at all, e. g. the AF system, how does that work in low light...
I appreciate your time.

..To be honest, I think the development of dslrs has peaked. What else can the manufacturers include?...
I'm hoping the dynamic range catches up with celluloid film.



(BTW, I’m going to be away from the computer this weekend. Will check back later…)
 

Chris7

macrumors 6502
Original poster
Aug 8, 2008
396
0
Lost in Thought
Rebel T1i?

Based on what I've seen reported, I'd say that at identical image sizes, the 5DmkII has a one stop advantage over the 7D, which, again identical image sizes, has less than a one-stop advtange over the 40D...
I havent seen a lot of Rebel XS images, but my guess is that it's similar to the 40D (same 10.1MP sensor)...
Where would the people here say the T1i fits in here?

Thanks.
 

toxic

macrumors 68000
Nov 9, 2008
1,664
1
This is really interesting. Could you please say more or provide some links? Also just so I understand you correctly, I understand dynamic range to be the range between light and dark that in a photo (without masking and adjusting different parts of a RAW separately) and latitude to be the amount you can correct over or underexposure (e.g. a RAW file would generally have more latitude, but not necessarily more dynamic range than a TIFF). So with colloid film, a single print will have a much more dynamic range than digital photos, and would also have much more latitude in the negative (you could probably push or pull it two stops), but would not have more latitude in the print. But please correct me if my terms are off.

at least for digital (and in common use), dynamic range is the range of light levels over which the sensor can display useful information (I guess it can phrased that way). this would be the equivalent of the dynamic range as you defined it above, plus the exposure latitude.

there are two websites I know of that give DR numbers: DPReview and DXOMark. DXOMark goes purely by sensor testing, and DPReview goes by how much information they can squeeze out of ACR (something like that). I don't like DXOMark's way of rating cameras since it is completely theoretical, so I go by DPReview, even though their method isn't entirely dependable, either. one way or another, based on reviews and DPR's numbers, Canon SLRs have very ordinary dynamic range.

note that, based on what more knowledgeable people have told me, that DXO's raw numbers are correct. it's just that there can be a huge disconnect between their theoretical results and what actually happens.
 

AlaskaMoose

macrumors 68040
Apr 26, 2008
3,559
13,408
Alaska
As far as my personal subjective testing goes, this is it:

xxxD series: ISO 400 is acceptable
xxD series: ISO 800 is acceptable
xD series: ISO 1600 is acceptable

But I am quite a bitch about noise, so most people can live with a full stop faster; however, this is how I feel about it and because I "only" have a 40D I also must have f/2.8 capable glass.

I've been dreaming about a 2nd hand 1DSmIII but my wife would like the smaller 5DmII better, so I'm quite torn about my next upgrade ;P
But you can also switch to f/1.4, or just keep on using the same lenses and take some of the noise out with software.
---------
Even the dynamic range can be increased these days with software. In my view, manufacturers are a long way for reaching a climax relating to sensor technology. These will only become better and better with technological advances.
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.