Large crop factor advantage of the new 18mp Canon EOS 7D

[robsphotography]

A while ago on this forum, there was a very interesting discussion about the small crop factor advantage that the Sony A700 has when compared with the Sony A900. Recently, I wrote some mathematical notes about the crop factor advantage, and I put them on my web site here:

http://www.robsphotography.co.nz/crop-factor-advantage.html

This page has been looked at by several people on the DPR forum, and it seems that the mathematics are OK, but I would also appreciate your views on this.

A recent development is the introduction of the 18 megapixel Canon EOS 7D, see here for details:

http://www.dpreview.com/previews/canoneos7d/

So, I am interested in your views on whether Sony is likely to match this by also marketing an 18mp APS-C camera? Do you think that sensor technology has improved recently so that it’s feasible to pack 18mp on to a 22.3mm wide sensor and still maintain excellent image quality?

The pixel density (in pixels per linear centimetre) of the Canon 7D is 2325 (5184 / 2.23), and this is a huge 38% greater than that of the Sony A900, which is 1685 (6048 / 3.59). This means that the crop factor advantage in this instance is also 38%. Now wouldn’t this be great for sports and wildlife photographers who often use telephoto lenses?

Regards
Rob
http://www.robsphotography.co.nz/Sony-A900.html

[bfitzgerald]

I think moving to 18mp is going to stretch a fair few lenses.

And at the price of the Canon 7d, it's really asking if you might be better off going FF for that sort of outlay (unless you do sports etc etc)

I think IQ will suffer with such high pixel densities.

[Winston]

Just to sumarize...

Pixel pitch:
A900 = 5.94µ
A700 = 160 5.49µ
   7D = 4.30µ

A900 / A700 = 1.08
A900 / 7D    = 1.38
A700 / 7D    = 1.28

While the A700 has only an 8% reach advantage over the A900, the 7D does indeed have a 38% advantage.
More interesting to me is the reach advantage that the 7D has over the A700, 28%.

As an A700 user (2) and wildlife photographer, I might be interested in a 7d-like model if I didn't have to give upanything in noise performance. I would be very interested if I could get an improvement in noise performance.

This based in my belief that my 70400G can handle the 4.3µ pixel pitch.

Should the density increase be accompanied by an improvement in noise performance, expect a 46.5 MPx version of the A900

[Dusty]

Should the density increase be accompanied by an improvement in noise performance, expect a 46.5 MPx version of the A900

And the megapixel war continues.......

Dusty

[robsphotography]

Just to sumarize...

Pixel pitch:
A900 = 5.94µ
A700 = 160 5.49µ
   7D = 4.30µ

A900 / A700 = 1.08
A900 / 7D    = 1.38
A700 / 7D    = 1.28

While the A700 has only an 8% reach advantage over the A900, the 7D does indeed have a 38% advantage.
More interesting to me is the reach advantage that the 7D has over the A700, 28%.

As an A700 user (2) and wildlife photographer, I might be interested in a 7d-like model if I didn't have to give upanything in noise performance. I would be very interested if I could get an improvement in noise performance.

This based in my belief that my 70400G can handle the 4.3µ pixel pitch.

Should the density increase be accompanied by an improvement in noise performance, expect a 46.5 MPx version of the A900

Thanks Winston, I guess we could add to the above list the Canon 5D Mark II, because this would be of particular interest to Canon users! The 5D II has image dimensions of 5616 pixels x 3744 pixels, a 36mm x 24mm sensor, and 21 megapixels. Its pixel density (in pixels per linear inch) is 1560, and its pixel pitch is 6.41 microns per pixel (36.0 x 1000 / 5616).

So, the pixel density (and pixel pitch) of the new Canon 7D is 49% greater than that of the Canon 5D Mk II. This means the crop factor advantage is also 49% in this example. If the 5D MkII had the same pixel density as the 7D, it would be a 47 megapixel camera, instead of a 21 megapixel camera!!

In terms of image width, an uncropped 7D image is 5184 pixels wide, and a cropped 5DII image with the same field of view as the 7D is 3479 pixels wide (5616 x 22.3 / 36.0). So, if printed at 200 pixels per inch, the print width of the full sized 7D image measures 25.9 inches, and the cropped image of the 5D II produces a print width of 17.4 inches. Therefore, the 7D print is 49% larger than the comparable 5DII print.

But when you compare the Sony A900 with the A700, the crop factor gain is only 7.9%. If you compare the A900 with the A380, the crop factor gain increases to 16%. So the gain to Canon users of 49% (7D vs 5DII) is very significant, provided that the image quality of the 7D is not degraded by the increased pixel density.

There is a fairly detailed review of the Canon 7D here:

http://www.imaging-resource.com/PRODS/E7D/E7DA.HTM

This article concludes that:

“All in all, the Canon 7D delivers exceptional print quality, with phenomenal detail (especially when working from RAW) at low ISOs, and a very graceful trade-off between noise and subject detail as you go up the ISO scale.”

Regards
Rob
http://www.robsphotography.co.nz/Sony-A900.html

[KevinBarrett]

Sony could go to something like 16.5 MP without getting as dense as the EOS 50D, and somewhere close to 20 MP in the case of the EOS 7D--the advantage of a larger sensor. Canon sure is getting ballsy with their pixel densities--enough to make Sony look rather reserved.

[peterottaway]

With the Sony A850, Canon EOS 7D and Nikon D300s all now doing the rounds, I had a look at the prices being quoted by a Hong Kong dealer on these rather different cameras.

Sony A850 is GBP 1460

Canon 7D pre-order is GBP 1440

Nikon D300s is GBP 1380

[David Kilpatrick]

Full frame 24 megapixels is, for many, a compelling spec versus any crop factor camera.

David

[robsphotography]

Full frame 24 megapixels is, for many, a compelling spec versus any crop factor camera.

David

I guess to people who have a long history of taking photographs with 35mm SLRs (such as myself) it’s a logical progression to choose a full frame camera. If the digital technology could have coped with full frame cameras some 10 years ago when digital cameras were first introduced, I wonder whether many people would have been calling for cameras with 24mm x 16mm sized sensors?

Certainly there are distinct advantages and disadvantages of both FF and “crop” cameras, but looking into the future, I wonder whether in 5 or 10 years time, the crop cameras will continue to be manufactured if FF cameras can be purchased for similar prices?

The prices quoted by Peter are very interesting, 1460 GBP for a full frame Sony A850, and 1440 GBP and 1380 GBP for two crop cameras. For people who are entering the DSLR market for the first time, this makes the decision quite difficult, and a knowledge of the key differences between FF and APS-C cameras is even more important. I think a lot of people (not all newbies) struggle to understand the concepts underlying the crop factor advantage, hence the continuing (and at times confused) discussions about this topic on photography forums.

Regards
Rob
http://www.robsphotography.co.nz/Sony-A900.html

[Greg Beetham]

Here is an interesting article on various formats and pixel density vs lens resolution...if you take what they have to say literally in the conclusion then 18MP in an APS-C sensor is a waste of time....
Greg

http://luminous-landscape.com/tutorials ... tion.shtml

[Dusty]

So my 350 is about perfect, the new Canon is over MPed and a FF camera should max out at 35 MP, according to that article.

Guess I'll get a 550 until I can afford an 840 or A1000.

Dusty

[robsphotography]

Here is an interesting article on various formats and pixel density vs lens resolution...if you take what they have to say literally in the conclusion then 18MP in an APS-C sensor is a waste of time....
Greg

Thanks, I agree it’s a good article. In a recent DPReview of the Canon 7D, it says this:

“The EOS 7D sports a new 18 MP APS-C CMOS sensor which is an in-house development. Canon claims the sensor delivers an improved signal to noise ratio which is achieved through a new photodiode and microlens design. The EOS 7D's sensor features the gapless microlenses that we first saw on the EOS 50D but on the new model the distance between microlenses and photodiodes has been reduced which results in the light being more easily focused onto the photodiode.”

http://www.dpreview.com/previews/canoneos7d/page3.asp

So I guess this could be the reason why 18mp is now possible? Another really good article that helps with making valid comparisons of cameras can be seen here:

http://www.josephjamesphotography.com/equivalence/

Regards
Rob
http://www.robsphotography.co.nz/Sony-A900.html

[Greg Beetham]

Hmm yes Rob point taken, I haven't had time to read it all yet, but it seems the new structure of the sensor that Canon has introduced will alter the playing field somewhat regarding density, but there is still a question mark over lens performance being adequate, especially possibly the midrange sector...might be a camera for the more well heeled in the wallet department...
Greg

ps. I don't quite follow the f2=f2=f2 argument totally in the Luminous Landscape discussion....wherin it's stated that a larger format has intinsically 'more' light transmission by virtue of the larger opening in the diaphragm. One would think that that's already a 'given' anyway because it has too to maintain the T-stop value, (more light yes, but spread over a much larger area). What I'm stuggling with is, does that necessarily equate to more photons per square mm, or per sensel, as their argument seems to attest?
Maybe it could I guess, under certain sensel size-spacing-density FF sensors, but we don't have one in FF yet, equivalent to the design of the new Canon APS-C one that is....