Canon EOS-1D X Mark III technical white papers released

degos · Jan 7, 2020

Very disappointed not just in the reduction in effective MP but also that the viewfinder AF coverage is slightly reduced from the earlier 1DX models.

155 of the 191 points are cross-type, 65 are available at f/8. Which is a big step up but again why the restriction? They tout the wonderful new square-pixel AF array but then mumble that the coverage isn't extended 'for technical reasons'.

Too many compromises, I'll wait for the R1D...

LDS · Jan 7, 2020

degos said:
mumble that the coverage isn't extended 'for technical reasons'.

Look at where the AF sensor is placed, and how it gets enough light to work, and you may start to understand why there are some technical reason it can't work the same way when reading data directly out of the sensor.

YuengLinger · Jan 7, 2020

degos said:
Very disappointed...

Dramatized for ya!

richperson · Jan 7, 2020

This is my number one wish and looks like it was granted: That plus increased fps is enough to get me on board.

While Canon engineers are careful not to over-promise on results users should expect, overall preliminary comparisons of images show about a 1-stop improvement in general noise performance vs. the previous EOS-1D X Mark II camera.

Architect1776 · Jan 7, 2020

I find it interesting that it can ignore heat waves in the atmosphere and focus on the subject at a distance and not get confused.
Very good Canon.

unfocused · Jan 7, 2020

richperson said:
This is my number one wish and looks like it was granted: That plus increased fps is enough to get me on board.

That's encouraging, but I am a little concerned about this statement:

The larger pixels of a 20 million pixel, fullframe CMOS image sensor provide an excellent foundation for delivering minimal noise and high image qualities at ISOs such as 3200, 6400, and above.

I'm primarily interested in the "and above" range since the 1DxII already performs pretty well through ISO 6400. If it offers a one-stop improvement above 6400 I will be pleased. If not...well...

richperson · Jan 7, 2020

unfocused said:
That's encouraging, but I am a little concerned about this statement:

I'm primarily interested in the "and above" range since the 1DxII already performs pretty well through ISO 6400. If it offers a one-stop improvement above 6400 I will be pleased. If not...well...

I've seen a few high ISO samples and the different looks pretty significant. From what I have seen the almost one stop difference seems pretty close. That would get me up to 12000, which is all I need in most situations.

AlanF · Jan 7, 2020

richperson said:
I've seen a few high ISO samples and the different looks pretty significant. From what I have seen the almost one stop difference seems pretty close. That would get me up to 12000, which is all I need in most situations.

If it adds a 1 stop improvement above iso 6400 it will be a miracle. At high isos, the overall noise for modern sensors and circuits is generated by the noise in the photon flux and not by the sensor or electronics, and the counting efficiency of modern sensors is well over 50%, leaving room for only a fraction of stop improvement.

scyrene · Jan 7, 2020

degos said:
the reduction in effective MP

What does this mean?

flip314 · Jan 8, 2020

scyrene said:
What does this mean?

The Mark III has 20.1M useable pixels, the Mark II had 20.2M. I think it's a stretch to call that a reduction, it's all but identical.

raptor3x · Jan 8, 2020

scyrene said:
What does this mean?

That he has no clue what he's talking about. The 1DXii and 1DXiii have exactly the same resolution at 5472 x 3648.

Jack Douglas · Jan 8, 2020

OUCH! https://www.amazon.ca/SanDisk-128GB-Extreme-CFexpress-Card/dp/B07WVDV3FG/ref=sr_1_2?keywords=cfexpress+cards&qid=1578457567&sr=8-2

I presently have 3 - 256, 1-128, and a 64 in CFast and I thought those were costly.

The CF slot was used sparingly and was a dumb idea from Canon, IMHO.

Jack

Aussie shooter · Jan 8, 2020

scyrene said:
What does this mean?

That living under a bridge means one gets a poor education?

jolyonralph · Jan 8, 2020

flip314 said:
The Mark III has 20.1M useable pixels, the Mark II had 20.2M. I think it's a stretch to call that a reduction, it's all but identical.

It's a bit of a cheat. The resolution of the images you get out is 5472 x 3648 which is 19.96 megapixels, and is identical for both the mark II and mark III. For the photographer this is the only figure that actually counts. the 20.1 megapixels is a marketing figure, and all the change does is make it slightly less deceptive than the 20.2 figure from the older camera. There is no reduction in resolution.

SecureGSM · Jan 8, 2020

AlanF said:
If it adds a 1 stop improvement above iso 6400 it will be a miracle. At high isos, the overall noise for modern sensors and circuits is generated by the noise in the photon flux and not by the sensor or electronics, and the counting efficiency of modern sensors is well over 50%, leaving room for only a fraction of stop improvement.

according to photons to photos, 1Dx II low light performance was: ISO 5189
Nikon D5 - around ISO 7000-ish
It seems that 1Dx III low light performance may end up being around ISO 10,000.
Now, this is amazing if true.... Simply amazing. This will afford me shooting with 300/4 lens where previously I had to resort to a heavy and expensive 300/2.8 indoors. I hope this tech will be used in 5DV. Exciting times indeed.

AlanF · Jan 8, 2020

SecureGSM said:
according to photons to photos, 1Dx II low light performance was: ISO 5189
Nikon D5 - around ISO 7000-ish
It seems that 1Dx III low light performance may end up being around ISO 10,000.
Now, this is amazing if true.... Simply amazing. This will afford me shooting with 300/4 lens where previously I had to resort to a heavy and expensive 300/2.8 indoors. I hope this tech will be used in 5DV. Exciting times indeed.

According to DxOmark, it's the other way round. D5 is 2434 and the 1DXII is 3207! https://www.dxomark.com/Cameras/Compare/Side-by-side/Nikon-D5-versus-Canon--EOS-1D-X-Mark-II___1062_1071
Why the difference? The first is that photonstophotos uses the iso stated by the manufacturer, and they are rarely accurate, but DxO actually measures them However, the fundamental laws of physics apply irrespective of these numbers, and it would be a miracle to get a stop extra.

unfocused · Jan 8, 2020

richperson said:
I've seen a few high ISO samples and the different looks pretty significant. From what I have seen the almost one stop difference seems pretty close. That would get me up to 12000, which is all I need in most situations.

The high ISO examples seem to be from jpgs with noise reduction applied in-camera. They look good from a noise standpoint, but I wasn't impressed with the smearing of details, especially since Canon is touting the low-pass filter as providing more detail. I think we will have to wait until we have raw files to really see what if any high ISO improvements there are. I'd like to believe Canon and I'd love a one stop improvement, but I'm not getting my hopes up.

@AlanF understands this stuff a lot better than I do and I tend to trust his opinion on this.

richperson · Jan 8, 2020

unfocused said:
The high ISO examples seem to be from jpgs with noise reduction applied in-camera. They look good from a noise standpoint, but I wasn't impressed with the smearing of details, especially since Canon is touting the low-pass filter as providing more detail. I think we will have to wait until we have raw files to really see what if any high ISO improvements there are. I'd like to believe Canon and I'd love a one stop improvement, but I'm not getting my hopes up.

@AlanF understands this stuff a lot better than I do and I tend to trust his opinion on this.

You are correct and it will be interesting to see what I can do in post when I get mine. But it does give me reason to be optimistic as there are several test photos in low light and ISO >10k that look much better than what I have been able to achieve at lower ISO values.

privatebydesign · Jan 8, 2020

Jack Douglas said:
OUCH! https://www.amazon.ca/SanDisk-128GB-Extreme-CFexpress-Card/dp/B07WVDV3FG/ref=sr_1_2?keywords=cfexpress+cards&qid=1578457567&sr=8-2

I presently have 3 - 256, 1-128, and a 64 in CFast and I thought those were costly. The CF slot was used sparingly and was a dumb idea from Canon, IMHO.

Jack

SanDisk 128GB Extreme PRO CFexpress Card Type B

Buy SanDisk 128GB Extreme PRO CFexpress Card Type B featuring 128GB Storage Capacity, PCIe 3.0 Bus, Max Read Speed: 1700 MB/s, Max Write Speed: 1200 MB/s, Records Raw 4K Video, Compatible with Select XQD Devices, Downloadable Photo Recovery Software. Review SanDisk null

www.bhphotovideo.com

Amazon Canada, or that seller, are trying to rip somebody off...

EOS RP

EOR R

EOS 5D MK IV

EOS 80D

Defining the poetics of space through Architecture

EOS 1D MK II

EOS 80D

Canon 5DSR II

EOS 5D MK IV

EOS RP

EOS 7D MK II

CR for the Humour

@brett.guy.photography

EOS R Mark II

2 x 5D IV

Canon 5DSR II

EOS 1D MK II

EOS 80D

Would you take advice from a cartoons stuffed toy?

Related posts

Sign up for our newsletter