Pages: [1]   Go Down

Author Topic: The 50D versus the 5D versus the D700  (Read 6972 times)

Ray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 10365
The 50D versus the 5D versus the D700
« on: January 21, 2009, 03:08:59 am »

The 5D2 would be the logical choice for me, but there are no wide-angle Canon zooms on a par with the Nikkor 14-24/2.8. I'll probaly eventually get a 5D2, or maybe I'll wait for its successor with improved video flexibility.

In the meantime, and on my next photographic trip to exotic locations, I'll probably carry the Nikon D700 with 12-24/2.8 lens and possibly a Nikkor 50mm prime; the Canon 50D with 17-55/2.8 IS,  24-105/F4 IS, 100-400/F5.6 IS, and leave the 5D at home.

You'll notice that all the Canon lenses mentioned have IS, which is an advantage when travelling with a lightweight and probably slightly inadequate tripod which one rarely uses.

What I want to determine is, if there is any disadvantage to taking a 50D instead of buying more lenses for the D700, or indeed taking along the 5D in place of the 50D.

In order to answer this question, I've resorted to examining DXOmark graphs, which is a lot less time-consuming than making my own tests and prints.

As always with the cropped format, shallow DoF is more difficult to achieve. The Canon EF-S 17-55/2.8 is roughly equivalent to the Nikkor 24-70/2.8, except it's slightly longer at 27-88mm FF equivalent, which is actually better for me.

What I'm more concerned about is the sharpness of a lens at its maximum aperture rather than its shallowness of DoF at that maximum aperture. The EF-S 17-55/2.8 really is surprisingly sharp at full aperture. I doubt that the Nikkor 24-70 is sharper. Once again we could do with some thorough and real MTF tests at various apertures and resolutions, but Photozone can give us a rough idea as to the comparative resolution of these two lenses.

Photozone advise against comparing results across different brands of cameras. I presume this is primarily because the pixel count of the different brands of cameras used may vary. In fact, most Canon lenses are tested with an 8mp cropped format (350D), and the Nikkor lenses are tested, I believe, with the 10mp D200.

However, a few Canon lenses have been retested by Photozone with the 15mp 50D. One such lens is the EF-S 10-22mm. One can compare results from a 350D with results from the 50D with this lens. The 50D results show approximately a 20% increase in resolution at 50% MTF.

If one applies this 20% figure to the EF-S 17-55/2.8 results with the 8mp 350D, one gets a centre resolution of 2490 LW/PH at 17mm and F2.8. (2075 x 20%).

If we look at the Nikkor 24-70 at F2.8 tested with the D200, we get an LW/PH figure of approximately 2210, significantly less than the Canon EF-S 17-55 with the 20% adjustment for the higher resolution of the 50D. But let's not quibble. A full frame sensor is less demanding of lens resolution, so let's call it a draw. I'm feeling magnanimous.

Let's now move to the noise and DR implications of using the 50D in place of the D700. Surely the D700 has lower noise and higher DR than the 50D. Could it be otherwise?

Well, maybe that's not such a big deal when you factor in DoF and shutter speed considerations, and especially when you factor in the IS advantage of the EF-S 17-55.

Typically at fairly close distances, there's a 2 stop difference between the cropped format and full frame with respect to DoF, and there's also a two stop advantage if the lens has IS or VR.

Comparing the 50D with 17-55 lens, with the D700 with 24-70 lens, the former has a 4 stop advantage with hand-held shots, at equal DoF. If DoF is not an issue and the fastest aperture is used with both cameras, the 50D has just a 2 stop advantage. It's sufficient to wipe out any noise and DR advantage of the D700.

Below is the  DR comparison of the 3 cameras, 5D, 50D and D700, from DXOmark, at the pixel level rather than the reduced print size of 8x12".

The horizontal axis represents the real, tested ISO, and the blobs on the graph indicate that the 3 cameras' ISOs are all overstated (exaggerated).


[attachment=11032:DXOmark_...R_screen.jpg]
Logged

stever

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1250
The 50D versus the 5D versus the D700
« Reply #1 on: January 21, 2009, 12:43:58 pm »

i would take the Photozone resolution numbers with several grains of salt, i've used Imatest on a few lenses with 20D, 40D, and 50D and generally can't get very close to his numbers.  I use RAW and convert in LR with the default 25% sharpening (if you use 50% sharpening, the MTF numbers are 25% higher).  I suspect Photozone is using a higher level of sharpening.  This is also a reason for difficulty in comparing Canon to Nikon.

Logged

NikosR

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 622
    • http://
The 50D versus the 5D versus the D700
« Reply #2 on: January 21, 2009, 01:56:00 pm »

Quote from: Ray
Photozone advise against comparing results across different brands of cameras. I presume this is primarily because the pixel count of the different brands of cameras used may vary. In fact, most Canon lenses are tested with an 8mp cropped format (350D), and the Nikkor lenses are tested, I believe, with the 10mp D200.

However, a few Canon lenses have been retested by Photozone with the 15mp 50D. One such lens is the EF-S 10-22mm. One can compare results from a 350D with results from the 50D with this lens. The 50D results show approximately a 20% increase in resolution at 50% MTF.

If one applies this 20% figure to the EF-S 17-55/2.8 results with the 8mp 350D, one gets a centre resolution of 2490 LW/PH at 17mm and F2.8. (2075 x 20%).

If we look at the Nikkor 24-70 at F2.8 tested with the D200, we get an LW/PH figure of approximately 2210, significantly less than the Canon EF-S 17-55 with the 20% adjustment for the higher resolution of the 50D. But let's not quibble. A full frame sensor is less demanding of lens resolution, so let's call it a draw. I'm feeling magnanimous.

Utter nonsense. You cannot make these kind of extrapolations. Take some time to understand the assumptions and limitations behind Imatest testing. Also understand what system resolution is with regards to digital camera systems.


Quote from: Ray
Let's now move to the noise and DR implications of using the 50D in place of the D700. Surely the D700 has lower noise and higher DR than the 50D. Could it be otherwise?

Well, maybe that's not such a big deal when you factor in DoF and shutter speed considerations, and especially when you factor in the IS advantage of the EF-S 17-55.

Typically at fairly close distances, there's a 2 stop difference between the cropped format and full frame with respect to DoF, and there's also a two stop advantage if the lens has IS or VR.

Comparing the 50D with 17-55 lens, with the D700 with 24-70 lens, the former has a 4 stop advantage with hand-held shots, at equal DoF. If DoF is not an issue and the fastest aperture is used with both cameras, the 50D has just a 2 stop advantage. It's sufficient to wipe out any noise and DR advantage of the D700.


Too many assumptions, unsubstantiated statements and arbitrary conclusions to make this train of thought worthwhile to anybody other than one using exactly the same logic as you. Even if all your statements and conclusions were correct your logic would worth to someone only to the degree that one's criteria and shooting requirements exactly matched yours.
« Last Edit: January 21, 2009, 02:11:59 pm by NikosR »
Logged
Nikos

Ray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 10365
The 50D versus the 5D versus the D700
« Reply #3 on: January 21, 2009, 03:10:47 pm »

Quote from: NikosR
Utter nonsense. You cannot make these kind of extrapolations. Take some time to understand the assumptions and limitations behind Imatest testing. Also understand what system resolution is with regards to digital camera systems.

It's because I do understand system resolution that I am able to make some guesses here. Since I own both full frame and cropped format cameras I'm very much aware that a full frame camera of similar pixel count can deliver marginally better results with the same lens, except in the corners, because it accesses a slightly lower lens resolution at a higher MTF.

Quote
Too many assumptions and unsubstantiated statements to make this train of thought worthwhile to anybody other than one using exactly the same logic as you.

It's all I've got to go by. I see no assumptions but merely deductions from a few facts. There's only one logic.

Here are the facts:

(1) The EF-S 17-55/2.8 is a susrprisingly sharp lens that is shown both by Photozone and my own testing to be as sharp as the Canon 50/1.8 prime at certain focal lengths.

(2) The EF-S 10-22 is shown by Photozone to produce about a 20% greater number of 'line widths per picture height' when used with a higher resolution camera, the 50D. Photozone haven't tested the EF-S 17-55 with the 50D yet.

(3) The EF-S 17-55 has image stabilisation.

Here are the deductions:

(1) The EF-S 17-55 on the 50D may also produce a 20% greater 'picture' resolution on the 50D, just as the EF-S 10-22 produces a 20% increase in picture resolution, according to Photozone. (That's quite reasonable, isn't it?)

(2) A 20% increase in LW/PH would place the EF-S 17-55 (with 50D) as having a higher system resolution than the 24-70 with D200.

(3) Since I would not be using the 24-70 with the D200, but with the D700, one might expect 'picture' resolution to be slightly greater with the D700, and therfore about equal to the EF-S 17-55 on the 50D.

(4) The combined benefit of image stabilisation and greater DoF at any at any F stop will allow the use of a lower ISO and will tend to cancel any noise advantage the D700/ 24-70 might have, especially when the higher pixel count of the 50D is also factored in.

(5) Little purpose would be served by getting a Nikkor 24-70 lens, considering I already own a 50D and EF-S 17-55/2.8 IS.
Logged

Ray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 10365
The 50D versus the 5D versus the D700
« Reply #4 on: January 21, 2009, 03:22:59 pm »

Quote from: stever
i would take the Photozone resolution numbers with several grains of salt, i've used Imatest on a few lenses with 20D, 40D, and 50D and generally can't get very close to his numbers.  I use RAW and convert in LR with the default 25% sharpening (if you use 50% sharpening, the MTF numbers are 25% higher).  I suspect Photozone is using a higher level of sharpening.  This is also a reason for difficulty in comparing Canon to Nikon.

Well, I'm only concerned with comparative results. I'd be surprised if Photozone uses sharpening on images before testing resolution. However, if they do, I'd be surprised if they were to apply different degrees of sharpening with different cameras and lenses. That would make the results very unreliable. The key to testing camera equipment is consistency of methodology.

It so happens after buying the Canon EF-S 17-55/2.8, a decision based upon Photozone's test of this lens, I compared the lens with my own copy of the Canon 50/1.8 (at F2.8), and sure enough the results were very similar to Photozones on a comparative bases. I didn't bother counting lines of course, so I've no idea what the LW/PH would have been.
Logged

stever

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1250
The 50D versus the 5D versus the D700
« Reply #5 on: January 21, 2009, 10:59:35 pm »

i agree that the Photozone results for Canon mount lenses are a basis for RELATIVE comparison (although subject to the quality of the particular lens tested, they have tested a few bad ones and a few good ones that have had repeated trips to Canon).  However, the numbers reported cannot be acheived without some sharpening - and i believe they use in-camera sharpening of JPEG images.  In-camera sharpening is different between manufacturers, hence the caveat on comparison.

my test of my Canon 50 1.4 is quite consistent with DP Review on both my 5D and 40D - they also shoot RAW images and convert in ACR

i cannot generally recommend Imatest because of the cost and learning curve (particularly for long lenses), but once you're up the curve, you can get useful objective data over the entire image field that i believe is much less time consuming and less subjective than pixel peeping ad-hoc images

Logged

Ray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 10365
The 50D versus the 5D versus the D700
« Reply #6 on: January 22, 2009, 12:05:55 am »

Quote from: stever
However, the numbers reported cannot be acheived without some sharpening - and i believe they use in-camera sharpening of JPEG images.  In-camera sharpening is different between manufacturers, hence the caveat on comparison.

I don't believe this is the case. It's clearly not the impression I get from the PZ FAQ at http://www.photozone.de/lens-test-faq

Here's the relevant quote dealing with these issues:

Quote
Q: Why are the quality ratings different from system to system ?

As mentioned above the lens quality is affected by the sensor "system". Every additional step in the pipeline decreases the output quality, specifically the low-pass filter in front of the sensor. Assuming you mount the same lens on different system its maximum resolution will vary according to the max. quality of the sensor system. There're also evolutions regarding the RAW converter quality so more recent system tests starts can benefit from this - e.g. Canon/Olympus RAWs are/were converted using ACR 3.2 whereas Pentax/Sony RAWs are/were converted via ACR 3.7 and there was an increase in converter quality with ACR 3.4). This must all be taken into account regarding the rating system.

Q: What is the meaning of LW/PH (MTF50) ?

A: LW/PH means "Line widths per picture height" - it's basically a unit for resolution (not sharpness which is a combination of contrast and resolution) limited by the max. resolution of the camera's image sensor. You shouldn't take the LW/PH value itself too seriously because it is dependent on a number of factors. The analyzing tool (Imatest) is quite vulnerable to the quality of the source material. All PZ sample images are taken as RAW files and converted via Photoshop ACR (default settings without automatic image correction and contrast set to 0). If you convert RAWs via other imaging applications the LW/PH figures will be lower or higher due to the different sharpening & contrast algorithms. In the future I will probably drop the LW/PHs in favor of a school mark system which is easier to read. Refer to the reference scale on the left of the MTF charts to classify the quality.

There's no implication here that sharpening is applied, other than what may be built into the ACR software and cannot be switched off. I get the impression that ACR does some basic noise reduction which cannot be circumvented and which is why some converters seem to produce marginally more detailed results than ACR, but I'm speculating here.

Logged

NikosR

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 622
    • http://
Logged
Nikos

Ray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 10365
The 50D versus the 5D versus the D700
« Reply #8 on: January 22, 2009, 01:41:45 am »

Quote from: NikosR
http://www.imatest.com/docs/sharpening.html

So what's your point, Nikos? I can find two statements in that document that appear to refer to different situations regarding sharpening, in-camera jpegs with various amount of sharpening already applied, and conversions from RAW images that have presumably not been sharpened.

(1)
Quote
To obtain a good measure of a camera's sharpness— to compare different cameras on a fair basis, the differences in sharpening must be removed from the analysis. The best way to accomplish this is to set the sharpening of all cameras to a standard amount. This means sharpening undersharpened images and de-sharpening (blurring) oversharpened images.

The above quote would presumably be relevant to comparison of images from P&S cameras without RAW capability.

(2)
Quote
Raw MTF50, without standardized sharpening, produces the most accurate results for comparing lenses and for comparing the sharpness at the center and edge of a single image.

I presume this second quote refers to the method used by Photozone, but of course I don't know for certain.
« Last Edit: January 22, 2009, 02:35:24 am by Ray »
Logged

JohnKoerner

  • Guest
The 50D versus the 5D versus the D700
« Reply #9 on: January 24, 2009, 12:54:46 pm »

Quote from: Ray
The 5D2 would be the logical choice for me, but there are no wide-angle Canon zooms on a par with the Nikkor 14-24/2.8. I'll probaly eventually get a 5D2, or maybe I'll wait for its successor with improved video flexibility.


Not so. I have read several revies showing the Canon 14mm f2.8 II is as good, and I for one would like to see this lens compared on a 5DMkII to a Nikkor 14-24/2.8 on a D700 ...
Logged

jani

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1624
    • Øyet
The 50D versus the 5D versus the D700
« Reply #10 on: January 24, 2009, 06:02:14 pm »

Quote from: JohnKoerner
Not so. I have read several revies showing the Canon 14mm f2.8 II is as good, and I for one would like to see this lens compared on a 5DMkII to a Nikkor 14-24/2.8 on a D700 ...
Yes, in some respects it's very good, but that 14mm f/2.8L MkII is not a zoom lens.
Logged
Jan

NikosR

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 622
    • http://
The 50D versus the 5D versus the D700
« Reply #11 on: January 25, 2009, 02:42:31 am »

Quote from: GBPhoto
16-9.net 14-24 vs. 14L II on 1Ds3

Judging by the crops in the above test, the particular sample of the 14L appears noticeably inferior to the sample of the 14-24. And no it is not a zoom lens.
Logged
Nikos
Pages: [1]   Go Up