Ad
Ad
Ad
Pages: « 1 [2]   Bottom of Page
Print
Author Topic: Canon 5DII review  (Read 25341 times)
digitaldog
Sr. Member
****
Offline Offline

Posts: 9191



WWW
« Reply #20 on: December 07, 2008, 11:10:32 AM »
ReplyReply

Quote from: jjj
I understand that aspect, but it also reduces the overall quality and so seems like a false comparison.

Yes and no. Yes if you need all the pixels. No if you're trying an apples to apples comparison (upsizing the other capture would be more unfair).

We don't always use all the pixels the capture device produces. When we sample down, at least in terms of noise, there's some reduction.

All things being equal, well in this case we can't make such a comparison.
Logged

Andrew Rodney
Author “Color Management for Photographers”
http://digitaldog.net/
Panopeeper
Sr. Member
****
Offline Offline

Posts: 1805


« Reply #21 on: December 07, 2008, 11:47:08 AM »
ReplyReply

Quote from: digitaldog
No if you're trying an apples to apples comparison (upsizing the other capture would be more unfair
This is plain nonsense. This is not trying to compare apples with apples but trying to make sauer cherry from an orange and compare that to a water melone.

If you plan to purchase a 21 Mpix camera in order to dowres it to 12 Mpix, you are better off by buying a 12 Mpix camera; or buy a 6 Mpix and upres it, you sure will be happy with it.
Logged

Gabor
digitaldog
Sr. Member
****
Offline Offline

Posts: 9191



WWW
« Reply #22 on: December 07, 2008, 03:47:56 PM »
ReplyReply

Quote from: Panopeeper
This is plain nonsense. This is not trying to compare apples with apples but trying to make sauer cherry from an orange and compare that to a water melone.

If you plan to purchase a 21 Mpix camera in order to dowres it to 12 Mpix, you are better off by buying a 12 Mpix camera; or buy a 6 Mpix and upres it, you sure will be happy with it.

People capture 21 MP all the time and use less for all kinds of needs, there's nothing nonsensical about it. People don't buy cameras solely based on their native capture resolution either, although it is a very important consideration.

If the idea is to compare the quality of the capture of two camera systems and their respective noise, one is 12MP the other 21MP, its never going to be a totally complete apples to apples comparison as I said. But sampling the 21MP down to 12MP at least tells you want the potential output of both, with respect to noise would be ASSUMING the output at that 12MP is the max. Are you proposing that instead, the 12MP should be sampled UP to 21MP? That would be far more nonsensical IMHO.
Logged

Andrew Rodney
Author “Color Management for Photographers”
http://digitaldog.net/
Ray
Sr. Member
****
Offline Offline

Posts: 8908


« Reply #23 on: December 07, 2008, 05:15:24 PM »
ReplyReply

Quote from: Panopeeper
This is plain nonsense. This is not trying to compare apples with apples but trying to make sauer cherry from an orange and compare that to a water melone.

If you plan to purchase a 21 Mpix camera in order to dowres it to 12 Mpix, you are better off by buying a 12 Mpix camera; or buy a 6 Mpix and upres it, you sure will be happy with it.

Gabor,
No-one plans to downres an an image before it's taken or even conceived. Some shots lend themselves to large prints, which may involve some degree of interpolation, and some shots might be so noisy that 8x12" prints are the largest you might consider making.

The question you should be asking is, are there any image quality advantages of the D3 compared with the 1Ds3 (or 5D2) in respect of the total scene, total image and total composition? We know there are pixel quality advantages, but no-one makes a print of a single pixel, and even if you were to (that might be an idea for a minimalist photograph  ) noise would not be apparent in a single pixel as you know very well.
Logged
jjj
Sr. Member
****
Offline Offline

Posts: 3539



WWW
« Reply #24 on: December 07, 2008, 07:51:12 PM »
ReplyReply

Quote from: Panopeeper
This is plain nonsense. This is not trying to compare apples with apples but trying to make sauer cherry from an orange and compare that to a water melone.

If you plan to purchase a 21 Mpix camera in order to dowres it to 12 Mpix, you are better off by buying a 12 Mpix camera; or buy a 6 Mpix and upres it, you sure will be happy with it.
What he said.

It's a pointless, meaningless comparison. You compare best possible outputs against each other, to determine which is best at outputting.
Would you test a Mini versus a Ferrari, but with the Ferrari's engine limited, so it couldn't go any faster than the Mini?
Logged

Tradition is the Backbone of the Spineless.   Futt Futt Futt Photography
Panopeeper
Sr. Member
****
Offline Offline

Posts: 1805


« Reply #25 on: December 07, 2008, 08:49:10 PM »
ReplyReply

The unit of digital photography is the pixel. Everything else is derived from that. There is no other generally accepted, measurable basis of comparison.

Pixels can be converted in print or in monitor display or whatever, but not the other way around. Therefor the results of pixel level comparisons can be translated in anything else; but it is ridiculous to transform something in an imaginary print size using an imaginary printer with an imaginary paper, viewed from an imaginary distance by an imaginary viewer equipped with a pair of imaginary glasses. The result has a value for sure, an imaginary one.
Logged

Gabor
Ray
Sr. Member
****
Offline Offline

Posts: 8908


« Reply #26 on: December 07, 2008, 08:56:06 PM »
ReplyReply

Quote from: jjj
What he said.

It's a pointless, meaningless comparison. You compare best possible outputs against each other, to determine which is best at outputting.
Would you test a Mini versus a Ferrari, but with the Ferrari's engine limited, so it couldn't go any faster than the Mini?

Yes. If you were comparing fuel consumption. If someone were to make the claim that the Mini is more fuel efficient than the Ferrari, then it would be reasonable to compare both cars at the same speed and similar accelerations. The Mini might still win, but not by such a great amount as would be the case if you were to drive both cars flat out.

Driving both cameras flat out, gives the resolution edge to the 1Ds3 but the noise edge to the D3. Slowing down a bit brings noise levels to a parity, approximately.


Logged
Ray
Sr. Member
****
Offline Offline

Posts: 8908


« Reply #27 on: December 07, 2008, 09:04:32 PM »
ReplyReply

Quote from: Panopeeper
The unit of digital photography is the pixel. Everything else is derived from that. There is no other generally accepted, measurable basis of comparison.

.. but it is ridiculous to transform something in an imaginary print size using an imaginary printer with an imaginary paper, viewed from an imaginary distance by an imaginary viewer equipped with a pair of imaginary glasses. The result has a value for sure, an imaginary one.

Okay! So I'm ridiculous. I can't help it. That's how I work. I imagine a print size, I imagine the printer, I imagine the paper, I imagine the viewing distance, I imagine the viewer equipped with an imaginary pair of glasses that enables him/her to appreciate the print from that imaginary viewing distance, then I try to create a reality from such imaginings. Sometimes I succeed. Sometimes I don't.
Logged
madmanchan
Sr. Member
****
Offline Offline

Posts: 2110


« Reply #28 on: December 07, 2008, 09:12:03 PM »
ReplyReply

The base capture unit is the pixel, yes, but the limiting or driving factor is often the output.

If you are regularly printing 24" x 30" prints like Charlie Cramer, then yes, every pixel counts. He's basically upsampling all the time.

On the other hand, Michael Kenna prints all his images at about 8" x 8". He's basically downsampling all the time (or would be, if he was shooting digital). You do not lose image quality with downsampling (compared to an original capture at the output resolution), as long as you are careful about doing the downsampling and apply appropriate sharpening.

The vast majority of images captured (even with DSLRs) are resampled down to approximately 800 pixels on the longer dimension and viewed on a ~90-dpi LCD screen with a sRGB-ish gamut, and maybe posted to web.

Most printed images go to glossy or luster surfaces with up to 8 stops of dynamic range, at sizes of 4" x 6" to 8" x 10" (think wedding, studio prints). Somewhat bigger images are printed into magazines and newsprint, but not that much bigger.

It is true that one does not always know at the time of capture exactly how (or even if) that image will be printed. But most photographers can draw conclusions based on their own printing habits over time, e.g., how often you make 12" prints versus 18" prints versus 30" prints.
« Last Edit: December 07, 2008, 09:14:06 PM by madmanchan » Logged

Panopeeper
Sr. Member
****
Offline Offline

Posts: 1805


« Reply #29 on: December 07, 2008, 10:23:01 PM »
ReplyReply

Quote from: madmanchan
The vast majority of images captured (even with DSLRs) are resampled down to approximately 800 pixels on the longer dimension and viewed on a ~90-dpi LCD screen with a sRGB-ish gamut, and maybe posted to web
While many of them get cropped, or displayed 600x400 or larger.

Quote
Most printed images go to glossy or luster surfaces with up to 8 stops of dynamic range, at sizes of 4" x 6" to 8" x 10"
Right. However, 4" x 6" to 8" x 10" is not a specific size, and the density can hugely vary.

Sticking to pixels, one can convert the result to any size in any presentation form.

This is pretty much the same issue as the measuring the noise in an image.

As the noise will be a composite of the noises of the raw channels, the result depends on several factors:

- the scenery
- the source of illumination
- the intensity of illumination
- the spectral characteristics of the sensor
- your expectation/tolerance of the noise level

Now, try to give a figure for a camera based on some scenery with some illumination. Then a photog goes out and makes a shot in bright sunlight with a polar filter and complains that the sky is noisy, though all tests show that this camera "is not noisy". The other photog shoots indoor with a low temp incandescent, the blue is down in the cellar, and the image is noise. (These are examples from the cases I have analyzed for complaining photographers.)

So, I measure the noise purely in the raw channels. This does not depend on *anything but the sensor*, and the result is comparable accross all cameras for all situations. If someone find an image noisy, one needs only to take a look at the raw image: "this channel is in the ninth stop @ ISO 1600, it would be a miracle if the result were not noisy".
Logged

Gabor
jjj
Sr. Member
****
Offline Offline

Posts: 3539



WWW
« Reply #30 on: December 08, 2008, 05:18:14 AM »
ReplyReply

Quote from: Ray
Yes. If you were comparing fuel consumption. If someone were to make the claim that the Mini is more fuel efficient than the Ferrari, then it would be reasonable to compare both cars at the same speed and similar accelerations. The Mini might still win, but not by such a great amount as would be the case if you were to drive both cars flat out.

Driving both cameras flat out, gives the resolution edge to the 1Ds3 but the noise edge to the D3. Slowing down a bit brings noise levels to a parity, approximately.
If you are concerned with fuel consumption you would not buy the Ferrari in the first place and wouldn't care if the mini gave a few more mpg. So again a pointless, irrelevant comparison.
« Last Edit: December 08, 2008, 05:19:31 AM by jjj » Logged

Tradition is the Backbone of the Spineless.   Futt Futt Futt Photography
Ray
Sr. Member
****
Offline Offline

Posts: 8908


« Reply #31 on: December 08, 2008, 05:51:21 AM »
ReplyReply

Quote from: jjj
If you are concerned with fuel consumption you would not buy the Ferrari in the first place and wouldn't care if the mini gave a few more mpg. So again a pointless, irrelevant comparison.

You would if someone designed a Ferrari which had the fuel consuption of a Mini when driven like a Mini, but the performance of a racing car when you put your foot down. This is basically the situation when comparing the D3 with the 1Ds3, except the differences are not so extreme. If you want the image size you think reasonable from a 12mp camera, you get the same image quality from the 1Ds3 as the D3 provides, regarding noise. If you want the resolution that only a 21mp camera can provide, then the D3 doesn't make the grade. That's not difficult to comprehend, is it?
Logged
madmanchan
Sr. Member
****
Offline Offline

Posts: 2110


« Reply #32 on: December 08, 2008, 07:33:47 AM »
ReplyReply

Quote from: Panopeeper
Right. However, 4" x 6" to 8" x 10" is not a specific size, and the density can hugely vary.

Sticking to pixels, one can convert the result to any size in any presentation form.

Gabor, I agree with you fully about the value of understanding and analyzing noise at the raw level. It explains why, for instance, if you shoot an image of a ColorChecker, the purple patch usually looks the noisiest (unless one uses a magenta filter ...)

On the other hand, for a given output process (e.g., a specific lab or printer), the output density generally does not very much, if at all. For example, if one is printing to an online lab like WHCC, the output density is always fixed (at 300 ppi), so for an 8" x 10" print it becomes highly predictable the total number of pixels. For a Lightjet it's also around 300 ppi. Same goes with a modern inkjet printer, which usually falls within the 300 ppi to 360 ppi range, although there are some driver options allowing 600 ppi to 720 ppi.
Logged

jjj
Sr. Member
****
Offline Offline

Posts: 3539



WWW
« Reply #33 on: December 08, 2008, 08:36:31 AM »
ReplyReply

Quote from: digitaldog
People capture 21 MP all the time and use less for all kinds of needs, there's nothing nonsensical about it.
Nonsense refers to foolish 'rational' of comparing 2 products using one at a reduced quality. Obviously cameras will often have their results viewed at varying and much lower resolutions than capture, such as on the web, but that is not relevant if trying to compare and contrast the best performance of the two items.

Quote
People don't buy cameras solely based on their native capture resolution either, although it is a very important consideration.
Actually that apparently is exactly what they do and why MP are used to sell cameras.
Show two lots of people 2 sets of prints. One set a better quality than the other and most unsurprisingly will choose the better images. But if you tell the people that the poorer quality shots were taken on a camera with more MP  [before deciding] and suddenly they will tend to prefer the inferior images.
I was going to start a topic on this rather interesting finding, but it seems germane to mention here.


Quote
If the idea is to compare the quality of the capture of two camera systems and their respective noise, one is 12MP the other 21MP, its never going to be a totally complete apples to apples comparison as I said. But sampling the 21MP down to 12MP at least tells you want the potential output of both, with respect to noise would be ASSUMING the output at that 12MP is the max. Are you proposing that instead, the 12MP should be sampled UP to 21MP? That would be far more nonsensical IMHO.
If comparing two products on their best performance, you test in a way that is challenging, not one that is limiting. Unless you are actively trying to favour one product, which by limiting the higher MP camera to that of the lower MP camera is what you are doing. If only ever outputting to a max of 12 MP, why even buy/use the 21mp camera.

So producing large prints  to the max ability of the 21mp camera, which is rationally why you would tend to buy a higher res camera over a lower res one [assuming it's not for bragging rights], then most people will upsample the lower res one to be able to match output resolution. So it's not so nonsensical, more representative of how large prints will actually be done. If you don't upres, then the noise/grain of the lower res file will be larger compared to the higher res file, plus detail willl be lower.


Surprisingly - upsampling does not always reduce quality. I tested the Genuine Fractals versus upresing in ACR some years back with the 20D [ACR was much  better and free].  But what was surprising was that the ACR file that was upressed, looked better than the native sized file.


Logged

Tradition is the Backbone of the Spineless.   Futt Futt Futt Photography
jjj
Sr. Member
****
Offline Offline

Posts: 3539



WWW
« Reply #34 on: December 08, 2008, 08:38:41 AM »
ReplyReply

Quote from: Ray
You would if someone designed a Ferrari which had the fuel consuption of a Mini when driven like a Mini, but the performance of a racing car when you put your foot down. This is basically the situation when comparing the D3 with the 1Ds3, except the differences are not so extreme. If you want the image size you think reasonable from a 12mp camera, you get the same image quality from the 1Ds3 as the D3 provides, regarding noise. If you want the resolution that only a 21mp camera can provide, then the D3 doesn't make the grade. That's not difficult to comprehend, is it?
And which was my point entirely. Duh!
Logged

Tradition is the Backbone of the Spineless.   Futt Futt Futt Photography
clearing
Newbie
*
Offline Offline

Posts: 2


« Reply #35 on: December 08, 2008, 11:30:46 AM »
ReplyReply

Quote from: Ray
You would if someone designed a Ferrari which had the fuel consuption of a Mini when driven like a Mini, but the performance of a racing car when you put your foot down. This is basically the situation when comparing the D3 with the 1Ds3, except the differences are not so extreme. If you want the image size you think reasonable from a 12mp camera, you get the same image quality from the 1Ds3 as the D3 provides, regarding noise. If you want the resolution that only a 21mp camera can provide, then the D3 doesn't make the grade. That's not difficult to comprehend, is it?

Considering the quote above, as well as the other valuable info on this thread, it seems there is an angle here I'd not understood or considered before. Since I'm now trying to decide between getting a 12mp D700 or a 21mp 5D2, the issue of superior low light performance, which I value highly, has shifted away from the D700 that I was leaning toward. I mean, heck, what is the IQ advantage to a D700 if the 5D2 offers equal high iso performance, when downsampled and printed at the same size as the D700 ?? I could then also have the increased resolution the 5D2 offers, when needed for larger prints, not downsampled...ie, "2 cameras in one"! I only have a minimal amount of aps-c stuff, and I'm brand agnostic. If I can believe what I've read on this thread, there is no reason at all for me to buy a 12mp full frame camera,
especially when I can have a 21mp camera for nearly the same price (when finally available, of course).

I think I WILL wait until Michael gives a full review of the 5D2, and maybe even the D700/D3 v.s. 5D2 v.s. Sony A900 shootout that he mentioned, before I make a decision. Of course, his shootout may not address the downsampling technique that has been discussed on this thread, but I really hope it does. It seems like an important issue.
Logged
Panopeeper
Sr. Member
****
Offline Offline

Posts: 1805


« Reply #36 on: December 08, 2008, 12:12:44 PM »
ReplyReply

Quote from: DesW
I have dealt with this gentleman for years and he has always delivered as he states

Des W


From HongKong Dec 5 2008

Hi Des,

Thanks for your reply,I had called Zeiss agent and was told
That the Zeiss Ikon 21mm/F:2.8 lens for Canon will be launched
Sometime in Jan.I will mail you when I have the price.
Cheers
James

I contacted Zeiss again to clarify this issue. The answer is:

There must be a misunderstanding around this question. My collague will contact the respective distributor in ordnung to clarify the issue.

The planned availability of the 2.8/21mm is:

- Distagon 2.8/21 ZF for Nikon: 1. quarter of 2009
- Distagon 2.8/21 ZK for Pentax: 1. half of 2009
- Distagon 2.8/21 ZE for Canon: end of 2009


I'm sorry for the bad news, but I thought the sooner you know it the better.
Logged

Gabor
clearing
Newbie
*
Offline Offline

Posts: 2


« Reply #37 on: December 08, 2008, 04:23:10 PM »
ReplyReply

Correction to my post above- Michael may have mentioned a shootout between the D3x (not D700/D3), 5D2 and A900.
Logged
Ray
Sr. Member
****
Offline Offline

Posts: 8908


« Reply #38 on: December 08, 2008, 11:52:20 PM »
ReplyReply

Quote from: clearing
I mean, heck, what is the IQ advantage to a D700 if the 5D2 offers equal high iso performance, when downsampled and printed at the same size as the D700 ?? I could then also have the increased resolution the 5D2 offers, when needed for larger prints, not downsampled...ie, "2 cameras in one"! I only have a minimal amount of aps-c stuff, and I'm brand agnostic. If I can believe what I've read on this thread, there is no reason at all for me to buy a 12mp full frame camera,
especially when I can have a 21mp camera for nearly the same price (when finally available, of course).

This is precisely the point that I and others have been trying to make in a thousand different ways. I just can't understand why some folks just don't seem to get the point. Could it be they just happen to own Nikon shares.

I believe there are some advantages to the D700, compared with the 5D2. Fundamental image quality is not the be all and end all. I believe the D700 has a faster frame rate than the 5D2 and possibly more flexible options with regard to auto-bracketing of ISO. There's also at least one Nikkor lens of a quality that has no Canon equivalent, the Nikkor 14-24/2.8 which is remarkably sharp even into the corners of full frame 35mm. It would be interesting to compare the corners of the Canon 14/2.8 prime on a 5D2 with the corners of the Nikkor 14-24 on a D700, either upsampling the D700 image to the 5D3 size, or downsampling the 5D2 image to 12mp.

Is anyone able and willing to do such a comparison?  
Logged
Pages: « 1 [2]   Top of Page
Print
Jump to:  

Ad
Ad
Ad