Canon 1DX dynamic range and banding measurements

Started Jul 8, 2012 | Discussions
Shop cameras & lenses ▾
Thierry Legault
Thierry Legault Junior Member • Posts: 48
Canon 1DX dynamic range and banding measurements

hello, here is a chart showing the dynamic range of the 1DX compared to 5D mark II and III (it's consistent with the DxO Labs dynamic range charts in "screen" mode).

Thanks to a special processing, I also have isolated the banding pattern and compared its level beween 1DX, 5DII, 5DIII and 7D.

All charts and comments in this page: http://legault.perso.sfr.fr/1DX.html

The 1DX looks like a good continuation of its predecessors, without big surprise (good or bad) or technological break.

thierry Legault
http://www.astrophoto.fr

Canon EOS 5D Mark II Canon EOS 5D Mark III Canon EOS 7D Canon EOS-1D
If you believe there are incorrect tags, please send us this post using our feedback form.
jffielde Regular Member • Posts: 371
Re: Canon 1DX dynamic range and banding measurements

Thanks. I'm pretty impressed with the dynamic range if it holds up as well as it appears. That would make the 1Dx king of the hill above ISO1000 or so (looking at the sensor database of the best sensors at DxO's website), would it not? I admittedly know nothing about this, so perhaps I misunderstand.

Thanks for doing the work.

AC1 Senior Member • Posts: 1,171
Re: Canon 1DX dynamic range and banding measurements

So is this showing the 5D2 has better DR through much of the range than 5D3?

 AC1's gear list:AC1's gear list
Canon PowerShot S95 Canon EOS 7D Canon EF 50mm f/1.4 USM Canon EF-S 10-22mm f/3.5-4.5 USM Canon EF-S 15-85mm f/3.5-5.6 IS USM +2 more
jeminijoseph Veteran Member • Posts: 5,976
Do you have the chart for noise?

Thanks for the effort. How's the high ISO noise compared to 5d III?

-- hide signature --

Thanks
Jemini Joseph

http://www.wildbirdimages.com

 jeminijoseph's gear list:jeminijoseph's gear list
Nikon D500
KewlEugene Regular Member • Posts: 306
Re: Canon 1DX dynamic range and banding measurements

at ISO 25k 52k 104k 208k ?

Thierry Legault
Thierry Legault OP Junior Member • Posts: 48
Re: Canon 1DX dynamic range and banding measurements

jffielde wrote:

Thanks. I'm pretty impressed with the dynamic range if it holds up as well as it appears. That would make the 1Dx king of the hill above ISO1000 or so (looking at the sensor database of the best sensors at DxO's website), would it not?

I think so. It seems equal to D4 at 1600 iso and higher above 3200

-- hide signature --

Thierry Legault - http://www.astrophoto.fr

Thierry Legault
Thierry Legault OP Junior Member • Posts: 48
Re: Canon 1DX dynamic range and banding measurements

AC1 wrote:

So is this showing the 5D2 has better DR through much of the range than 5D3?

Yes, this is confirmed by Dxo Labs charts. This is due to the higher readout noise of the mark III (34 electrons vs 25 electrons at 100 iso), a very unusual evolution!

However, banding must be taken into account because even if its amplitude is much lower than readout noise (5 to 10 times), human vision is extremely efficient in lines detection and in practice, banding often limits dynamic range more than background noise. And the mark III shows comparable vertical banding but lower horizontal banding than mark II.

-- hide signature --

Thierry Legault - http://www.astrophoto.fr

Thierry Legault
Thierry Legault OP Junior Member • Posts: 48
Re: Canon 1DX dynamic range and banding measurements

KewlEugene wrote:

at ISO 25k 52k 104k 208k ?

The rule of thumb for most DSLRs (including Canon and Nikon ones) is to decrease the DR by 1 Ev for each iso step at high iso. This is nicely shown by DxO Labs charts which generally become linear above 1600 or 3200 iso (the D800 even shows a linear chart on the overall iso range).

For the 1DX, this leads to a bit less than 8 Ev at 25k, 7 Ev at 52k, 6 Ev at 104k and 5 Ev at 208k (at this level, is it still photography? ;-))

-- hide signature --

Thierry Legault - http://www.astrophoto.fr

Earthlight Veteran Member • Posts: 3,180
Re: Canon 1DX dynamic range and banding measurements

Thank you so much for sharing this Thierry!

Pattern noise is and has been for years my biggest complaint as a Canon shooter. Topaz denoise helps some though. I wish they could get rid of it but as I have been waiting for that to happen ever since the 20D I do not have my hopes up.

Earthlight

Ps. eagerly awaiting for your shots of the ring of the atmosphere of Venus and shots after the transit.

-- hide signature --
photorama Junior Member • Posts: 49
Re: Canon 1DX dynamic range and banding measurements

Bonjour Thierry,

Thanks a lot for sharing these interesting results.

I read that the Nikon D800 has a much wider dynamic range than the 5D III and I wander why because if you compare pixel size it should be the contrary; Is this a matter of read noise or something else?

Jacques Joffre

 photorama's gear list:photorama's gear list
Canon EOS 5D Mark III Nikon 1 V2 Canon EOS-1D X Mark II Canon EF 300mm f/2.8L IS USM Canon EF 24-105mm f/4L IS USM +2 more
Keyser Sze Regular Member • Posts: 478
Re: Canon 1DX dynamic range and banding measurements

Thanks for the analysis.

-- hide signature --

Who's Keyser Soze?

Thierry Legault
Thierry Legault OP Junior Member • Posts: 48
Re: Canon 1DX dynamic range and banding measurements

photorama wrote:

Bonjour Thierry,

Thanks a lot for sharing these interesting results.

I read that the Nikon D800 has a much wider dynamic range than the 5D III and I wander why because if you compare pixel size it should be the contrary; Is this a matter of read noise or something else?

Jacques Joffre

Hello Jacques!

Indeed, the DxO chart shows a very high dynamic range for the D800 at low iso (at high iso, other DSLRs catch it up). Although I can't measure easily the same parameters as I did with Canon (because the Nikon firmware sets the offset level to zero before recording the raw file, leading to the loss of small negative variations of noise), I think that it is due to a very low readout noise at all iso settings, contrary to other Canon and Nikon DSLRs whose readout noise starts at a high level and decreases with iso setting.

For example, the 1DX starts at 34 electrons at 100 iso and ends at less than 2 electrons at high iso. If we suppose that the D800 electronics offer 2 electrons of noise regardless of iso seeting and that the pixel capacity at 100 iso is about 25000 electrons (considering their smaller size), this leads to a DR of 13,6 Ev (DxO chart shows 13.2 Ev).

This important difference in the charts is the evidence that the technology of the sensor and/or electronics of the D800 is significantly different.

-- hide signature --

Thierry Legault - http://www.astrophoto.fr

Mikael Risedal
Mikael Risedal Veteran Member • Posts: 4,294
Re: Canon 1DX dynamic range and banding measurements

Same old story, pattern noise and high read out noise

-- hide signature --

Life is a battle wishes aunt Titti

David Hull
David Hull Veteran Member • Posts: 6,000
Re: Canon 1DX dynamic range and banding measurements

Earthlight wrote:

Thank you so much for sharing this Thierry!

Pattern noise is and has been for years my biggest complaint as a Canon shooter. Topaz denoise helps some though. I wish they could get rid of it but as I have been waiting for that to happen ever since the 20D I do not have my hopes up.

Earthlight

Nik Define 2.0 also does a nice job on the 5DIII vertical noise so there is another option as well.

Ps. eagerly awaiting for your shots of the ring of the atmosphere of Venus and shots after the transit.

-- hide signature --
 David Hull's gear list:David Hull's gear list
Canon EOS 50D Canon EOS 5D Mark III
victor china
victor china Forum Member • Posts: 87
great work ! is it normalized for the same pixel size?

great work

interesting findings

did you normalize for the same pixel size or is it the full rez results?

Mikael Risedal
Mikael Risedal Veteran Member • Posts: 4,294
Re: Canon 1DX dynamic range and banding measurements

if the figures are right above 1600isoit means great QE
hmm

Thierry Legault wrote:

photorama wrote:

Bonjour Thierry,

Thanks a lot for sharing these interesting results.

I read that the Nikon D800 has a much wider dynamic range than the 5D III and I wander why because if you compare pixel size it should be the contrary; Is this a matter of read noise or something else?

Jacques Joffre

Hello Jacques!

Indeed, the DxO chart shows a very high dynamic range for the D800 at low iso (at high iso, other DSLRs catch it up). Although I can't measure easily the same parameters as I did with Canon (because the Nikon firmware sets the offset level to zero before recording the raw file, leading to the loss of small negative variations of noise), I think that it is due to a very low readout noise at all iso settings, contrary to other Canon and Nikon DSLRs whose readout noise starts at a high level and decreases with iso setting.

For example, the 1DX starts at 34 electrons at 100 iso and ends at less than 2 electrons at high iso. If we suppose that the D800 electronics offer 2 electrons of noise regardless of iso seeting and that the pixel capacity at 100 iso is about 25000 electrons (considering their smaller size), this leads to a DR of 13,6 Ev (DxO chart shows 13.2 Ev).

This important difference in the charts is the evidence that the technology of the sensor and/or electronics of the D800 is significantly different.

-- hide signature --

Thierry Legault - http://www.astrophoto.fr

-- hide signature --

Life is a battle wishes aunt Titti

(unknown member) Senior Member • Posts: 2,261
Looks like it's going to kick the Nikon D4's a$$ to the curb! nt

no text

Mikael Risedal
Mikael Risedal Veteran Member • Posts: 4,294
Re: Looks like it's going to kick the Nikon D4's a$$ to the curb! nt

Not so sure about that.

Apewithacamera wrote:

no text

-- hide signature --

Life is a battle wishes aunt Titti

Thierry Legault
Thierry Legault OP Junior Member • Posts: 48
Re: great work ! is it normalized for the same pixel size?

No normalization, everything is full size (anyway pixel size is very close between 1DX, mark II and III)

victor china wrote:

great work

interesting findings

did you normalize for the same pixel size or is it the full rez results?

-- hide signature --

Thierry Legault - http://www.astrophoto.fr

rwbaron Veteran Member • Posts: 9,505
Re: Looks like it's going to kick the Nikon D4's a$$ to the curb! nt

My guess is the D4 will be noticably better in the shadows at ISO 100 to 400, about equal at 800 to 1600 and the 1DX will have a slight noise and DR advantage at 3200 and above. The higher ISO's is where this type of camera is most often used but I doubt either will be a "system changer" for any pro heavily invested in one system or the other.

Both are great tools and in the right hands will produce identical results.

Bob
--
http://www.pbase.com/rwbaron

 rwbaron's gear list:rwbaron's gear list
Canon EOS 7D Fujifilm X-T1 Canon EF 400mm f/5.6L USM Fujifilm XF 18-55mm F2.8-4 R LM OIS Fujifilm XF 55-200mm F3.5-4.8 R LM OIS +11 more
Keyboard shortcuts:
FForum MMy threads