December 22, 2014, 07:57:55 PM

Author Topic: The last Canon crop sensor - ever  (Read 40942 times)

CarlTN

  • Canon EF 300mm f/2.8L IS II
  • ********
  • Posts: 2227
    • View Profile
Re: The last Canon crop sensor - ever
« Reply #165 on: August 14, 2013, 05:49:45 PM »
no my friend
these pictures are matched and measured
and if you do not se any difference in resolution you have some problem

Again, if you want to compare...please do a comparison between the D7100 and the 7D, with the same Nikon lens and an adaptor...at the same focal length.  That would be more pertinent to this thread.  Nobody cares how a D800 compares to a 5D2, do they?  There are so many factors at play, it would be best to limit a comparison to very similar cameras, especially crop cameras...mounted to the exact same Nikon lens.  At least from where I sit...as my earlier point is that the D7100 has a better sensor than the 7D, and very likely will have better "image quality" than the 7D2 (no guarantees of course).

Certainly if the 7D2 winds up being better than it looks like it will be...to me, I will have no problem admitting I was wrong with my guessing.  It's been known to happen. 
« Last Edit: August 14, 2013, 05:51:27 PM by CarlTN »

canon rumors FORUM

Re: The last Canon crop sensor - ever
« Reply #165 on: August 14, 2013, 05:49:45 PM »

zim

  • 1D Mark IV
  • ******
  • Posts: 795
    • View Profile
Re: The last Canon crop sensor - ever
« Reply #166 on: August 14, 2013, 05:52:25 PM »
I understand that one lenses '200' won't always be the the same as anothers (of a different make) but the difference in sizes of those houses, do lenses marked focal lengths really vary that much? (sorry more than a bit OT)

rs

  • 5D Mark III
  • ******
  • Posts: 694
    • View Profile
Re: The last Canon crop sensor - ever
« Reply #167 on: August 14, 2013, 05:55:59 PM »
no my friend
these pictures are matched and measured from the whitest to the darkest and the same curve , the white point is the same,  but Im looking after resolution
and if you do not se any difference in resolution you have some problem
se the wall at the X

and yes you se some more noise because of 36Mp and 100%
Stronger, direct light, resulting in stronger, harsher shadows in the details of that wall. How can that form the basis of a comparison?
5D II | 24-70 II | 70-200 II | 100L | 40 | Sigma 50/1.4 | 40D | 10-22 | 17-55 | 580 EX II | 1.4x TC II

jrista

  • Canon EF 400mm f/2.8L IS II
  • *********
  • Posts: 4817
  • EOL
    • View Profile
    • Nature Photography
Re: The last Canon crop sensor - ever
« Reply #168 on: August 14, 2013, 05:58:34 PM »
no my friend
these pictures are matched and measured from the whitest to the darkest and the same curve , the white point is the same,  but Im looking after resolution
and if you do not se any difference in resolution you have some problem
se the wall at the X

and yes you se some more noise because of 36Mp and 100%

You are making an incomplete argument. Yes, 36mp is higher res., with smaller pixels, therefor more detail and more noise (which, ironically, is the same issue with the 7D). You made the argument that you could just downscale to the 5D III size, and that noise would be the same. You are failing to point out that if you downscale, you no longer have a resolution advantage. All that extra detail you claim to see (I only really see sharper wires in the fence...detail-wise, the rest looks the same to me)...well it would all be gone. Its the whole having a cake, and eating it too problem. You can have your pretty high resolution cake, but to eat all the noise...well you gotta eat all the resolution too. :P

BTW, these two shots were clearly taken at different times of day. The shading is quite different between the two, which changes which details can be clearly seen. The D800 shot is also darker by maybe two thirds of a stop than the 5D III shot. In the D800 shot, the right side of the building looks blown...with all that extra DR, I wouldn't have expected blown highlights...maybe some recovery would fix that. Only the trim looks like it might be blown on the 5D III shot...but it would be interesting to see if, pulled down by 2/3rd stops, if that remained true.

Chuck Alaimo

  • 1D Mark IV
  • ******
  • Posts: 988
    • View Profile
    • Chuck Alaimo Photography
Re: The last Canon crop sensor - ever
« Reply #169 on: August 14, 2013, 05:58:43 PM »
ok, so at least this topic has gone back to a more reasoned conversation, but, how id it become lets compare FF bodies/lenses thing?  Yes there's noise in the d800 image.  Yes there's less detail in the 100% crop for the mk2....   Isn't this topic supposed to be about APS-C sensors, more specifically the idea that this round of body updates represents the end of the APS-C sensor in general?   
Owns 5Dmkiii, 6D, 16-35mm, 24mm 1.4, 70-200mm 2.8, 50mm 1.4, 85 mm 1.8, 100mm 2.8 macro, 1-600RT, 2 430 EX's, 1 video light

jrista

  • Canon EF 400mm f/2.8L IS II
  • *********
  • Posts: 4817
  • EOL
    • View Profile
    • Nature Photography
Re: The last Canon crop sensor - ever
« Reply #170 on: August 14, 2013, 06:06:30 PM »
no my friend
these pictures are matched and measured from the whitest to the darkest and the same curve , the white point is the same,  but Im looking after resolution
and if you do not se any difference in resolution you have some problem
se the wall at the X

and yes you se some more noise because of 36Mp and 100%
Stronger, direct light, resulting in stronger, harsher shadows in the details of that wall. How can that form the basis of a comparison?

there are difference between the cameras and the lenses even if I try to make it even, BUT in this case and resolutiowise and the wall with a X the light are not the problem.

The angle of lighting does matter. Different phase angle changes which details are visible. The lighting in the D800 shot looks like it is mostly coming from off-scene right, where as in the 5D III shot it is more direct, above. The light in the D800 offers a greater phase angle, which brings out surface detail better (phase angle was an interesting debate on birdphotographers.net forums a while ago, and a fairly important factor in getting highly detailed bird photographs.)

To provide a proper comparison, you would really need to take the shot with both cameras from exactly the same spot, with exactly the same focal length, with the same aperture, shutter, and ISO, at exactly the same time of day (a few minutes difference won't matter, but more than that, especially hours difference, does). One would then be able to make a proper comparison, either at native size or to normalized image dimensions.

neuroanatomist

  • CR GEEK
  • ************
  • Posts: 15238
    • View Profile
Re: The last Canon crop sensor - ever
« Reply #171 on: August 14, 2013, 06:12:04 PM »
Nobody cares how a D800 compares to a 5D2, do they?

Sorry, Carl but you are completely incorrect about that.  Mikael/ankorwatt obviously cares about such a comparison, so much so that he posts similar sets of pictures in thread after thread after thread.  Wait, the thread has nothing to do with DR or sensor IQ?  That's ok, Mikael can fix that.   Ask him the time in Stockholm, he'll answer, "The D800 has better DR than any Canon sensor."

ok, so at least this topic has gone back to a more reasoned conversation, but, how id it become lets compare FF bodies/lenses thing?

Because Mikael started posting in it.  He DRips and DRibbles DRibs and DRabs of his DRool into every thread he can, DRoning on endlessly. Just add this one to the list of threads DRagged down into the dolDRums.
EOS 1D X, EOS M, and lots of lenses
______________________________
Flickr | TDP Profile/Gear List

canon rumors FORUM

Re: The last Canon crop sensor - ever
« Reply #171 on: August 14, 2013, 06:12:04 PM »

jrista

  • Canon EF 400mm f/2.8L IS II
  • *********
  • Posts: 4817
  • EOL
    • View Profile
    • Nature Photography
Re: The last Canon crop sensor - ever
« Reply #172 on: August 14, 2013, 06:12:42 PM »
no my friend
these pictures are matched and measured from the whitest to the darkest and the same curve , the white point is the same,  but Im looking after resolution
and if you do not se any difference in resolution you have some problem
se the wall at the X

and yes you se some more noise because of 36Mp and 100%

You are making an incomplete argument. Yes, 36mp is higher res., with smaller pixels, therefor more detail and more noise (which, ironically, is the same issue with the 7D). You made the argument that you could just downscale to the 5D III size, and that noise would be the same. You are failing to point out that if you downscale, you no longer have a resolution advantage. All that extra detail you claim to see (I only really see sharper wires in the fence...detail-wise, the rest looks the same to me)...well it would all be gone. Its the whole having a cake, and eating it too problem. You can have your pretty high resolution cake, but to eat all the noise...well you gotta eat all the resolution too. :P

BTW, these two shots were clearly taken at different times of day. The shading is quite different between the two, which changes which details can be clearly seen. The D800 shot is also darker by maybe two thirds of a stop than the 5D III shot. In the D800 shot, the right side of the building looks blown...with all that extra DR, I wouldn't have expected blown highlights...maybe some recovery would fix that. Only the trim looks like it might be blown on the 5D III shot...but it would be interesting to see if, pulled down by 2/3rd stops, if that remained true.

if I down scale I do not have the resolution advantage?
My best Jrista, if you look at details it is exactly what I have even if I down scale the d800 to a 5dmk2 5dmk3 file size
I have more details and less noise

By nature, downscaling reduces resolution. I'm not sure what's confusing about that. Downscaling will average neighboring pixels together. That does have the effect of reducing noise...but it is also averaging separate elements of fine detail into fewer elements of fine detail.

You cannot both have a resolution advantage, and reduce noise, when downscaling. You are normalizing resolution. That implies that any resolution advantage disappears. You gain less noise, at the cost of detail. You cannot simultaneously compare a 7360x4912 pixel full-size and more noisy D800 image while also comparing a 5760x3840 pixel downscaled and less noisy copy. You either compare at full size, and have your noisier resolution advantage, or you compare at the same size, and have less noisy detail that is about the same as the 5D III. Some things on the D800 would remain a bit sharper, but you could really only tell that if you took two proper, honestly identical shots as I mentioned in my last reply.

schill

  • Rebel T5i
  • ****
  • Posts: 115
    • View Profile
Re: The last Canon crop sensor - ever
« Reply #173 on: August 14, 2013, 06:28:54 PM »
I'm going to lose interest if that house doesn't start running. :)

As mentioned above, this discussion doesn't seem to have anything to do with this thread.  I figured this "running" comment couldn't take it more off topic so it is probably safe.

Don Haines

  • Canon EF 300mm f/2.8L IS II
  • ********
  • Posts: 3602
  • also on superhero vacation
    • View Profile
Re: The last Canon crop sensor - ever
« Reply #174 on: August 14, 2013, 06:37:54 PM »
To try and get back on topic, I believe that Canon will continue to make crop sensors as long as they make digital cameras. FF sensors are only for a small selection of cameras, the vast bulk are tiny cameras like point/shoots that will never have FF sensors.....

For example, the following picture is with a crop sensor.... Devices like an iPad will never have the real estate for FF and the required lenses. Crop sensors, in all their many formats, are here to stay.

Match point, game over. I'm going to take a nap.... Bye...
The best camera is the one in your hands

neuroanatomist

  • CR GEEK
  • ************
  • Posts: 15238
    • View Profile
Re: The last Canon crop sensor - ever
« Reply #175 on: August 14, 2013, 07:14:31 PM »
Neuro:The 5dmk3 and 6D are verry similar to 5dmk2

No, they are very different cameras, that happen to have fairly similar sensor performance.  I'll keep saying that, and maybe someday a miracle will occur, and you'll actually understand the distinction.
EOS 1D X, EOS M, and lots of lenses
______________________________
Flickr | TDP Profile/Gear List

Pi

  • 1D Mark IV
  • ******
  • Posts: 937
    • View Profile
    • Math and Photography
Re: The last Canon crop sensor - ever
« Reply #176 on: August 14, 2013, 09:08:11 PM »
By nature, downscaling reduces resolution. I'm not sure what's confusing about that. Downscaling will average neighboring pixels together. That does have the effect of reducing noise...but it is also averaging separate elements of fine detail into fewer elements of fine detail.

You cannot both have a resolution advantage, and reduce noise, when downscaling. You are normalizing resolution. That implies that any resolution advantage disappears.

I disagree. Resolution is not measured by the number of pixels. Image captured by a higher resolution sensor is captured with more detail, and then resampled in a digital, "perfect" way usually with some smart algorithm which does involve some "AA-like" smoothing. Image captured at the final, lower resolution, is captured with an analogue AA filter, dependent on the available technology, and with a Bayer type of sampling. The AA filter is stronger what the pixel density requires because of the Bayer design. The result is less resolution. Then you have all those CA and other lens corrections which make things even worse for the lower resolution sensor.

Downsizing can be bad if the ratio is not far enough from 1. The bottom line is: we need not just more mp, we need many more.  :)

jrista

  • Canon EF 400mm f/2.8L IS II
  • *********
  • Posts: 4817
  • EOL
    • View Profile
    • Nature Photography
Re: The last Canon crop sensor - ever
« Reply #177 on: August 14, 2013, 10:19:39 PM »
By nature, downscaling reduces resolution. I'm not sure what's confusing about that. Downscaling will average neighboring pixels together. That does have the effect of reducing noise...but it is also averaging separate elements of fine detail into fewer elements of fine detail.

You cannot both have a resolution advantage, and reduce noise, when downscaling. You are normalizing resolution. That implies that any resolution advantage disappears.

I disagree. Resolution is not measured by the number of pixels. Image captured by a higher resolution sensor is captured with more detail, and then resampled in a digital, "perfect" way usually with some smart algorithm which does involve some "AA-like" smoothing. Image captured at the final, lower resolution, is captured with an analogue AA filter, dependent on the available technology, and with a Bayer type of sampling. The AA filter is stronger what the pixel density requires because of the Bayer design. The result is less resolution. Then you have all those CA and other lens corrections which make things even worse for the lower resolution sensor.

Downsizing can be bad if the ratio is not far enough from 1. The bottom line is: we need not just more mp, we need many more.  :)

I agree about needing way many more. I'll happily take a 50 or 60mp FF if that's what Canon comes out with. Give me all the megapixels I can get.

As for resolution, I was talking image resolution. Spatial resolution is certainly a different story. Downscale a D800 image that is identical to a 5D III image, and the total pixels (image resolution) will be the same. The D800 image would still be sharper than the 5D III. You could sharpen the 5D III image, and gain back some of what the optical AA filter took away, but you run the risk of haloing (I did mention that before.) The difference between the two images wouldn't be huge though, even if the edge went to the D800.

canon rumors FORUM

Re: The last Canon crop sensor - ever
« Reply #177 on: August 14, 2013, 10:19:39 PM »

GaryJ

  • Canon AE-1
  • ***
  • Posts: 76
    • View Profile
Re: The last Canon crop sensor - ever
« Reply #178 on: August 14, 2013, 11:30:38 PM »

Did anyone ever call 35mm film "full frame" before digital came along (except maybe in the context of those cameras that could shoot 2 or 4 pictures within a single 35mm frame)?  I think you might be able to argue that anytime some refers to FF in this context they are talking digital.

No, quite the opposite.  35mm was traditionally called the miniature or leica format.

All those folk with 6D's and 5D3's thinking that 'full frame' is just the best thing ever...  the medium format guys haven't even noticed you are in the room.
+1
7D,6D,100mmL IS f2.8 Macro,70-300mm L,MPE-65, 24-105 f4 L,50mm f1.8,70-200mm L f4,Siggy 150mmf2.8 Macro, Tokina 11-16mm Kenko Extenders and 1.4 TC [for 70-300] 40mm f2.8 STM

neuroanatomist

  • CR GEEK
  • ************
  • Posts: 15238
    • View Profile
Re: The last Canon crop sensor - ever
« Reply #179 on: August 14, 2013, 11:43:34 PM »
All those folk with 6D's and 5D3's thinking that 'full frame' is just the best thing ever...  the medium format guys haven't even noticed you are in the room.

That's ok, I haven't noticed them, either.  Personally, I prefer to measure my shooting speed in frames per second, not seconds per frame.
EOS 1D X, EOS M, and lots of lenses
______________________________
Flickr | TDP Profile/Gear List

canon rumors FORUM

Re: The last Canon crop sensor - ever
« Reply #179 on: August 14, 2013, 11:43:34 PM »