November 22, 2017, 07:04:16 PM

Author Topic: ISO, Sensors, and Mark III, and !DX  (Read 5163 times)

sanyasi

  • Guest
ISO, Sensors, and Mark III, and !DX
« on: March 09, 2012, 11:42:19 AM »
I have seen others write that the ISO on the new 5D, Mark III is 2 stops improved and that the 1DX ISO reflects three stops of improvement.  If so, for me, the question is whether one more stop of improvement is worth close double the price (yes, I realize there are other differences, the this is the one that matters the most to me).

Is the way we measure improvement to the sensor just in terms of ISO, or are there other improvements that go beyond ISO that either increase image quality at any given ISO number or that further improve the ISO.

Bottom line:  If it is just image quality, on the 5D, Mark III, I assume now I can set the ISO to 3200 and get the same level of noise that I got at 800 on my 5D Mark II.  What other improvements should I expect?  I realize much of this will only become clear once people have worked with the cameras, but for those of you who know more about how the specs translate into image quality, I would appreciate your thoughts.

Thanks

Jack Siegel

canon rumors FORUM

ISO, Sensors, and Mark III, and !DX
« on: March 09, 2012, 11:42:19 AM »

neuroanatomist

  • CR GEEK
  • ************
  • Posts: 21774
Re: ISO, Sensors, and Mark III, and !DX
« Reply #1 on: March 09, 2012, 12:47:33 PM »
I have seen others write that the ISO on the new 5D, Mark III is 2 stops improved and that the 1DX ISO reflects three stops of improvement.

I've seen people write that the world will end on December 21st of this year. 

I honestly think it's unwise to base this comparison on what one person writes about seeing images in many cases viewed on the camera LCD only.  Do keep in mind that there are still no 1D X RAW files in the wild (at least, not that I'm aware of), and the 5DIII RAW files that are available were shot with pre-production cameras.

So, I'd say the jury is still way, way out on this issue...and will remain so until both cameras are released through retail channels and tested by competent individuals/groups.

I do believe that the 1D X will have the edge in ISO performance, and Canon has explicitly stated that.  How much of an edge is an open question...comparing the 1D X and 5DIII to the 5DII, maybe it's 3-stops vs. 2-stops, or maybe it's 2/3-stop vs. 1/2-stop.  Time will tell.
« Last Edit: March 09, 2012, 12:50:04 PM by neuroanatomist »
EOS 1D X, EOS M2, lots of lenses
______________________________
Flickr | TDP Profile/Gear List

YellowJersey

  • Guest
Re: ISO, Sensors, and Mark III, and !DX
« Reply #2 on: March 09, 2012, 10:06:39 PM »
Remember, you're also paying for the 12 fps with the 1DX, which is going to add to the cost significantly.

hhelmbold

  • EOS Rebel T7i
  • ****
  • Posts: 119
Re: ISO, Sensors, and Mark III, and !DX
« Reply #3 on: March 10, 2012, 12:31:40 AM »
I've seen people write that the world will end on December 21st of this year. 

Isn't this reason enough to go all or nothing?  ;D

LetTheRightLensIn

  • Canon EF 400mm f/2.8L IS II
  • *********
  • Posts: 4755
Re: ISO, Sensors, and Mark III, and !DX
« Reply #4 on: March 10, 2012, 12:36:28 AM »
I have seen others write that the ISO on the new 5D, Mark III is 2 stops improved and that the 1DX ISO reflects three stops of improvement.  If so, for me, the question is whether one more stop of improvement is worth close double the price (yes, I realize there are other differences, the this is the one that matters the most to me).

Is the way we measure improvement to the sensor just in terms of ISO, or are there other improvements that go beyond ISO that either increase image quality at any given ISO number or that further improve the ISO.

Bottom line:  If it is just image quality, on the 5D, Mark III, I assume now I can set the ISO to 3200 and get the same level of noise that I got at 800 on my 5D Mark II.  What other improvements should I expect?  I realize much of this will only become clear once people have worked with the cameras, but for those of you who know more about how the specs translate into image quality, I would appreciate your thoughts.

Thanks

Jack Siegel

The 5D3 won't be 2 stops better. Canon was talking about in cam jpg NR. Nobody really has a solid figure yet on it. The best educated guess points to the 5D3 being two-thirds of a stop better than the 5D2 in the RAW files, but that could easily change. It will be interesting to see what DxO comes up with. Two-thirds of a stop at this point now that they are nearing possible limits would actually be a solid improvement.

The quality of the noise might look a bit nicer on the 5D3 compared to the 5D2, at high ISO at least, it's hard to say exactly to what degree yet, especially for the low ISO case. If this difference is big enough it might make the apparent SNR improvement seem larger than the actual measured values.

It appears that the maxium dynamic range of the 5D3 will most likely be the same as that with the 5D2. Some suggest that the 1DX may use a new fab that allows for a bit better dynamic range and that they used and old process for the 5D3 again, that remains to be seen. (One easy way to get more low ISO dynamic range is through digital CDS and column-parallel ADC on chip but Sony has patents all over that it seems. Some suggest that super high quality fab and quality control might be a way for Canon to get closer to that without that tech but they say it is a much more expensive way to get there and it alone will only so far. We will see. One person claimed Canon had been offered some sort of dynamic range at base ISO enhancement years ago but turned it down, who knows the details even if so. Some say that without the Sony patents that the Canon engineers will have a heck of a time figuring out some creative way to get past the point they have reached since the 1Ds3.)

If the 5D3 uses some old sensor fab and the 1DX uses a new fab perhaps it could have a stop or so better max DR??? Maybe another 1/2 stop in SNR??? Or maybe it will be more or less the same. We have no data at all from the 1DX. Some varying claims from various Canon reps that don't all fit together and a claim by someone with some connections that didn't have any numbers attached to it. You may need to wait a month or two to find the answers to this part.

Maybe the color filter array will be less color blind, maybe not. Such like that will have to await DxO.

In all honesty you probably need to wait a few weeks and come back and ask again. It probably has no max DR improvement (I hope I am wrong though but it is much more likely than not it won't, but still possible), the other stuff is a lot more speculative at this point or simply not even known at all.

I guess you said the other stuff doesn't matter much for your needs, but to mention a little bit of it:

The AF should be light years better assuming they didn't make some royal disaster of the new system, but that seems hard to believe. On paper it would be very much better for both still and AI Servo shooting in the center frame and worlds much better away from center frame for both.

Shutter response time is shorter and fps are up to 6.

The movie mode appears to have a lot less moire and aliasing, the SNR may be noticeably better too. Hopefully the sharpness will be much better, although none of the demos shown demonstrate this yet, but most appear to have used heavy NR or been compressed for the web.

AutoISO still isn't 100% perfect but may have finally reached the first usable stage outside of 1 series camera. Insane that something so trivial has been dribbled out over a decade.

Some nice UI and general functionality improvements have been made.

Supports dual cards.

Sadly there are no replacement view screens now although the VF is improved to 100% view and has built in electronic levels and rule of thirds lines and such.

etc.
« Last Edit: March 10, 2012, 12:40:34 AM by LetTheRightLensIn »

jrista

  • Canon EF 600mm f/4L IS II
  • **********
  • Posts: 5319
  • EOL
    • Nature Photography
Re: ISO, Sensors, and Mark III, and !DX
« Reply #5 on: March 10, 2012, 02:57:35 PM »
If the 5D3 uses some old sensor fab and the 1DX uses a new fab perhaps it could have a stop or so better max DR??? Maybe another 1/2 stop in SNR??? Or maybe it will be more or less the same. We have no data at all from the 1DX. Some varying claims from various Canon reps that don't all fit together and a claim by someone with some connections that didn't have any numbers attached to it. You may need to wait a month or two to find the answers to this part.

Maybe the color filter array will be less color blind, maybe not. Such like that will have to await DxO.

Just so you have some solid information from Canon on that point. From http://cpn.canon-europe.com/content/education/technical/inside_canon_eos_5d_mark_iii.do (Image Quality, Processing & Playback):

Quote from: Canon
Full-frame CMOS sensor

Canon has a long history of designing and building its own CMOS sensors and is unique in the world of photography in that it also designs and produces the machines that make the sensors, providing complete control from the start of the production process. The full-frame sensor found in the EOS 5D Mark III is an evolution of the sensor in the EOS 5D Mark II and it also includes some of the sensor technology that’s in Canon’s flagship EOS-1D X DSLR.

CMOS sensors make use of microlenses to direct the light into each pixel well. Canon’s first full-frame sensor with a gap-less microlens design is found in the EOS-1D X and the sensor in the EOS 5D Mark III features the same technology for improved light gathering ability. Gap-less microlenses mean that no matter what angle the light arrives at the sensor from it is directed into a pixel well where it can be used most effectively. In essence it means that no light that gets to the sensor is wasted by not making it into a pixel.

Just like the EOS-1D X, the sensor in the EOS 5D Mark III has been developed to excel at shooting stills and also at capturing 1080p Full HD movie footage. Compared to the sensor in the EOS 5D Mark II both the sensor and the associated image processing have been developed to show reduced moiré patterning and false colour; offering greater detail and improved image quality for shooting movies.

And from a related caption:

Quote from: Canon
Like the EOS-1D X’s sensor the full-frame CMOS sensor of the EOS 5D Mark III offers increased light sensitivity, image noise reduction at higher ISO speeds, and a wider dynamic range.

Certainly not definitive...but at least from the horses mouth. Given how CMOS devices are designed, it can't be "old fab" if it includes 1D X improvements. The 5D III sensor is a new design requiring fabrication of new sensors, its just not something that was started fresh from the ground up, completely scrapping all prior knowledge Canon had about sensor design (why would they do something like that?)

The same "evolution" rather than "revolution" of sensor design is true for the 1D X as well. Taken from http://cpn.canon-europe.com/content/education/technical/eos_1d_x_explained.do:

Quote from: Canon
Full-frame CMOS sensor

Canon has a long history in designing and building its own CMOS sensors, so the full-frame sensor found in the EOS-1D X is an evolution of the sensor found in the EOS-1D Mark IV that incorporates the latest advances in photodiode construction technologies.

The new photodiode construction has resulted in an improved photoelectric conversion rate that gives increased light sensitivity. Combined with improved transistors inside the pixels, the signal-to-noise ratio has been made even better, allowing access to the higher ISO speeds without increasing image noise.

CMOS sensors make use of microlenses to direct the light into each pixel well. As seen on other Canon CMOS sensors, the sensor in the EOS-1D X uses gapless microlenses located above each photodiode to maximise the light gathering capability, no matter what angle the light arrives at the sensor from. It is the first time that gapless microlenses have been employed on a Canon full-frame sensor and they are one key piece of the puzzle that opens up the high ISO capabilities of the camera. The newly designed CMOS circuits have also been designed to capture a wider dynamic range, meaning more detail is retained in both shadows and highlights in comparison to previous cameras.

Unlike previous CMOS sensors found in other EOS DSLRs, the sensor in the EOS-1D X has been developed to excel not only at stills shooting but also when capturing Full 1080p HD Movie footage. Compared to the sensor found in the EOS 5D Mark II both the sensor in the EOS-1D X and the associated image processing have been developed to show reduced moiré patterning and false colour; offering greater detail and improved image quality.

A full-frame, 18 Megapixel sensor shooting at 14 frames per second produces a large signal stream that needs to be removed from the sensor rapidly. To enable this, the EOS-1D X uses a 16-channel high-speed output with two-vertical-pixel simultaneous readout. It is around 1.4 times faster than the readout system found in the EOS-1D Mark IV and it's this that allows the 14 frames per second shooting speed — a first for a camera with a 35mm full-frame digital sensor. Note that at ISO 32,000 or higher the frame rate will be reduced to 10fps (which still equals the fastest shooting rate of the EOS-1D Mark IV).

New sensors, requiring new fabrication. No "reuse" of existing sensors with improvements extrasensor.

It appears that the maxium dynamic range of the 5D3 will most likely be the same as that with the 5D2. Some suggest that the 1DX may use a new fab that allows for a bit better dynamic range and that they used and old process for the 5D3 again, that remains to be seen. (One easy way to get more low ISO dynamic range is through digital CDS and column-parallel ADC on chip but Sony has patents all over that it seems. Some suggest that super high quality fab and quality control might be a way for Canon to get closer to that without that tech but they say it is a much more expensive way to get there and it alone will only so far. We will see. One person claimed Canon had been offered some sort of dynamic range at base ISO enhancement years ago but turned it down, who knows the details even if so. Some say that without the Sony patents that the Canon engineers will have a heck of a time figuring out some creative way to get past the point they have reached since the 1Ds3.)

Dynamic range has been improved to be better, in both the shadows and the highlights, over previous cameras...in both the 1D X and 5D III sensors. BOLD claim if it isn't true by even a 10th of a stop. BOLD CLAIM.
« Last Edit: March 10, 2012, 03:11:34 PM by jrista »

bycostello

  • EOS 5DS R
  • ******
  • Posts: 904
    • London Weddings
Re: ISO, Sensors, and Mark III, and !DX
« Reply #6 on: March 12, 2012, 05:49:30 AM »
don't think i've ever shot over 1000 iso, so couldn't give too much of a monkeys.... 

canon rumors FORUM

Re: ISO, Sensors, and Mark III, and !DX
« Reply #6 on: March 12, 2012, 05:49:30 AM »