Canon 5dmk2 cheats with ISO

Started Dec 13, 2008 | Discussions
pforsell
pforsell Senior Member • Posts: 2,267
Canon 5dmk2 cheats with ISO

The new 5dmk2 image quality and especially noise performance has been tested extensively. What most website/magazine tests have failed to take into account is the forgery Canon uses with the ISO ratings to make their camera look better in tests.

When set to ISO 1600 the 5dmk2 is actually ISO 1093, which means it is 2/3rds stops below ISO 1600. To actually get ISO 1600 out of 5dmk2 you'll have to crank up the ISO reading to 2500.

What users don't realize is that at setting ISO 1600 the 5dmk2 is actually closer to 800 than 1600. It makes the camera look okay in tests but in real world this doesn't help. You'll have to crank up the ISO to get the shutter speed you need... and suffer the loss of dynamic range, details, color purity and increase in noise.

The original 5D has almost 2 stops better noise and high ISO (actual, not forged) performance than 5dmk2. Without the forgery the 5dmk2 would have been immediate epic failure in image quality tests.

Click ISO SENSITIVITY on this page:
http://www.dxomark.com/index.php/eng/Image-Quality-Database/Canon/EOS-5D-Mark-II

 pforsell's gear list:pforsell's gear list
Nikon D1X Nikon D2X Nikon D3X Nikon D3S Nikon D4S +22 more
leniin New Member • Posts: 7
Re: Canon 5dmk2 cheats with ISO
-- hide signature --

Do ? You own Canon 5D Mark II or ? you tray to be smart boy , get life start doing some picture ...... and go do NIKON FORMUM ........

wildlife1212 Senior Member • Posts: 1,109
5dmk2 & 1Dsmk3 & D700 Iso Sensitivity almost the same

pforsell wrote:

The new 5dmk2 image quality and especially noise performance has been
tested extensively. What most website/magazine tests have failed to
take into account is the forgery Canon uses with the ISO ratings to
make their camera look better in tests.

When set to ISO 1600 the 5dmk2 is actually ISO 1093, which means it
is 2/3rds stops below ISO 1600. To actually get ISO 1600 out of 5dmk2
you'll have to crank up the ISO reading to 2500.

What users don't realize is that at setting ISO 1600 the 5dmk2 is
actually closer to 800 than 1600. It makes the camera look okay in
tests but in real world this doesn't help. You'll have to crank up
the ISO to get the shutter speed you need... and suffer the loss of
dynamic range, details, color purity and increase in noise.

The original 5D has almost 2 stops better noise and high ISO (actual,
not forged) performance than 5dmk2. Without the forgery the 5dmk2
would have been immediate epic failure in image quality tests.

Click ISO SENSITIVITY on this page:
http://www.dxomark.com/index.php/eng/Image-Quality-Database/Canon/EOS-5D-Mark-II

I compared 5D2 vs 1Ds3 vs D700 here
http://forums.dpreview.com/forums/read.asp?forum=1032&message=30314001

.

didd Regular Member • Posts: 261
Re: Canon 5dmk2 cheats with ISO

read that
http://www.dxomark.com/index.php/eng/Technologies/Measurement-definitions

The way dxo measures the sensitivity does not correspond to the usual ISO sensitivity adjustment.
dxo definition refers to saturation (vsat)
usual definition refers to gain (response , i.e mv/photons)

It's a choice. But if you do this choice, you may know that the sensor is no more linear when it is near to saturation = you get FPN, color shift, etc..

Then, it's the choice of the manufacturer to estimate if the sensor can be used or not in this area. Some say yes, some say no and clip the signal before (i.e. Canon on the 5D). The dxo measurement , which refers to vsat, introduces a bias if it's used to compare S/N, because the way the manufacturer considers if the signal is useful or not when close to saturation has nothing to see with the sentivity in the linear zone, where S/N is measured.

Dxo guys are doing a good job, seems they have a good knowledge of the theory but ...

lol101 Veteran Member • Posts: 5,855
So do Nikon, Sony and most camera manufacturers...

ISO is a standard recommendation... not an absolute rating.

Look at Nikon's D700/D3 measurements on the site you pointed at: they are the same as 5DMkII one...

It's all a big forgery... fortunately, in real world it doesn't matter...

 lol101's gear list:lol101's gear list
Panasonic LX100 Canon EOS-1D Mark II N Canon EOS 6D Canon EOS M5 Canon EF 24mm f/2.8 +5 more
Luciano Morpurgo Contributing Member • Posts: 710
Re: Canon 5dmk2 cheats with ISO

pforsell wrote:

When set to ISO 1600 the 5dmk2 is actually ISO 1093, which means it
is 2/3rds stops below ISO 1600. To actually get ISO 1600 out of 5dmk2
you'll have to crank up the ISO reading to 2500.

For the 5D, iso 1600 is actually 1277, to have actual 1600 you should crank it up to 2000. Some difference yes, and it's bad to see that Canon is pushing further this "trick" with the new model, yet it's not exactly "2 stop advantage"...

But another thing I noticed is that in some comparison shots, the 5DII is slightly underexposed given the same iso/aperture/speed, and that's no good too.
Luc

-- hide signature --
 Luciano Morpurgo's gear list:Luciano Morpurgo's gear list
Canon PowerShot G5 Canon PowerShot G9 Canon PowerShot S95 Canon EOS 5D Canon EOS 40D +9 more
minasnd Forum Member • Posts: 54
Re: Canon 5dmk2 cheats with ISO

There are a lot of tests measuring actual sensitivity, which give different results.

For example: http://www.diwa-labs.com/wip4/compare_camera.epl?product=7118

To see the results you download tha iso sensitivity .pdf file. The actual sensitivity measured is "Saturation ISO sensitivity" in summary.

rhlpetrus Forum Pro • Posts: 25,932
Re: 5dmk2 & 1Dsmk3 & D700 Iso Sensitivity almost the same

wildlife1212 wrote:

Click ISO SENSITIVITY on this page:
http://www.dxomark.com/index.php/eng/Image-Quality-Database/Canon/EOS-5D-Mark-II

Even though the difference in these graphs are about 1/3 stop between D700 and 5DII in the high ISO range, the DxO scores take that into account, since they use the actual values. For example, the lowlight score is based on SNR at 18%, the cut point of 30DB, so that the score is the actual ISO where that cut occurs, so it doesn't matter for DxO scores. Same for DR, etc (except that cameras with ISO100 as base ISO have some advantage over those with base ISO 200, by their own account of methodology).

-- hide signature --

Renato.
http://www.flickr.com/photos/rhlpedrosa/
OnExposure member
http://www.onexposure.net/

Good shooting and good luck
(after Ed Murrow)

 rhlpetrus's gear list:rhlpetrus's gear list
Leica D-Lux (Typ 109) Nikon D7000 Nikon 1 V1 Nikon AF-S DX Nikkor 12-24mm f/4G ED-IF Nikon AF Nikkor 35mm f/2D +4 more
glowie NYC Regular Member • Posts: 255
Re: Canon 5dmk2 cheats with ISO
RedFox88 Forum Pro • Posts: 29,398
wow, who let the troll in??

And does that site have credibility for anything?

Because DPR rated the 5D to be about 1/3 stop FASTER than iso standards in their review, e.g. iso100 is really 125, iso1600 is really 2000. Yet that site says the 5D is actually SLOWER than the ISO standards. Hmm, who do you believe? A new site looking for attention or a testing site that's been around for about 10 years?

Neil Enns Regular Member • Posts: 241
It's more accurate than my T-Max!

Sweet! Canon does a better job of measuring ISO than Kodak.

After testing it turns out that (depending on how I process it) I get between ISO 160 and ISO 250 using T-Max 400.

Neil

Iliah Borg Forum Pro • Posts: 26,188
I'm awfully sorry

But you are completely confused.

Please read the appropriate standard on camera sensitivity.

-- hide signature --
ThomasMiller Senior Member • Posts: 2,735
I did the real tests

In my tests of the 5D MK2, using a Kodak spectra 5200 Leve3 without the graphic spider interface activated, we found that the new Canon was actually at ISO 1613.597 when set to 1600, but only in situations where overall light value were below 950 CL. At that point, Canon's propriety firmware engaged a boost with integrated NR which essentially elevated the sensor to ISO 1790.995. This would often change based on the type of lighting used, shifting heavily in color ranges above 2200K and at the lower end of the inputs if the spider was activated.
Switching to the Fuji based analyzer we found nearly the same results.

I know, don't attack me just because the Fuji has a stage 1 interface and the interociter is based on Metaluna technology rather than Krell drivers. I'm well aware of the value changes and it CAN be dialed in to match if there are no magnetic sources in the room and the traps are above 70 degrees.

In the end we found that most people who spoke about ISO values and black dots until they turned blue couldn't actually take a good photo to save their shaved poodles.

But then a lot of those same folks sold everything in their trailer already and can't afford a camera as good as the new 5D anyway.
So in the end there is justice afterall.

Iliah Borg Forum Pro • Posts: 26,188
Re: I did the real tests

In my tests of the 5D MK2, using a Kodak spectra 5200 Leve3

Good for you. Not funny enough to spill my coffee.

Those who are actually testing, are you testing as per the standard? Are you mixing up ISO and EI?

-- hide signature --
f_stops Contributing Member • Posts: 565
Re: Canon 5dmk2 cheats with ISO

pforsell wrote:

The new 5dmk2 image quality and especially noise performance has been
tested extensively. What most website/magazine tests have failed to
take into account is the forgery Canon uses with the ISO ratings to
make their camera look better in tests.

I noticed my 5d and 5dII meter differently, sometimes - roughly by 1/4 stop - 1/2 stop (5dII exposing more). I do have the 5d set up with the ee-s screen and the 5dII viewfinder is much brighter.

When set to ISO 1600 the 5dmk2 is actually ISO 1093, which means it
is 2/3rds stops below ISO 1600. To actually get ISO 1600 out of 5dmk2
you'll have to crank up the ISO reading to 2500.

What users don't realize is that at setting ISO 1600 the 5dmk2 is
actually closer to 800 than 1600. It makes the camera look okay in
tests but in real world this doesn't help. You'll have to crank up
the ISO to get the shutter speed you need... and suffer the loss of
dynamic range, details, color purity and increase in noise.

The original 5D has almost 2 stops better noise and high ISO (actual,
not forged) performance than 5dmk2. Without the forgery the 5dmk2
would have been immediate epic failure in image quality tests.

Here is the same exposure at 3200 on the 5d and 5dII 1/125 - f3.5. 5dII downsampled . NR off. Note the blacks and reds.

The four individual shots are here (at 100%)

http://revolve.smugmug.com/gallery/6803974_TpUFU/2

 f_stops's gear list:f_stops's gear list
Leica M9
mpixel Veteran Member • Posts: 5,779
Re: Canon 5dmk2 cheats with ISO

Remember that the actual ISO if the original 5D is about 25% higher than listed:

http://www.dpreview.com/reviews/canoneos5d/page21.asp

f_stops wrote:

pforsell wrote:

The new 5dmk2 image quality and especially noise performance has been
tested extensively. What most website/magazine tests have failed to
take into account is the forgery Canon uses with the ISO ratings to
make their camera look better in tests.

I noticed my 5d and 5dII meter differently, sometimes - roughly by
1/4 stop - 1/2 stop (5dII exposing more). I do have the 5d set up
with the ee-s screen and the 5dII viewfinder is much brighter.

When set to ISO 1600 the 5dmk2 is actually ISO 1093, which means it
is 2/3rds stops below ISO 1600. To actually get ISO 1600 out of 5dmk2
you'll have to crank up the ISO reading to 2500.

What users don't realize is that at setting ISO 1600 the 5dmk2 is
actually closer to 800 than 1600. It makes the camera look okay in
tests but in real world this doesn't help. You'll have to crank up
the ISO to get the shutter speed you need... and suffer the loss of
dynamic range, details, color purity and increase in noise.

The original 5D has almost 2 stops better noise and high ISO (actual,
not forged) performance than 5dmk2. Without the forgery the 5dmk2
would have been immediate epic failure in image quality tests.

Here is the same exposure at 3200 on the 5d and 5dII 1/125 - f3.5.
5dII downsampled . NR off. Note the blacks and reds.

http://www.6887.com/5dII/5d-5dIImacbeth3200.jpg

The four individual shots are here (at 100%)

http://revolve.smugmug.com/gallery/6803974_TpUFU/2

-- hide signature --

GMT

Camera Dog Senior Member • Posts: 1,851
OP - you better have real proof

or risk having Canon come after you after such a statement
or hope that the laws in Finland protect you.

-- hide signature --

Certified Professional Master Photographer, Level 3.
Certified Expert on 30 D and on Zoom L Lenses
Working on L Prime Certification

ozcreations Forum Member • Posts: 67
Who would know more about ISO? Dx0 or Canon.... (nt)
bronxbombers Forum Pro • Posts: 18,226
no.

you are not looking at the graph correctly, the 5dmkii DR and noise plot line is always above the 5d line so even taking into account that the 5d understates the ISO and the 5dmkii over states it compared to standard the 5d 1 is still behind (although barely at the 100% view level, but noticeably at the per print view).

pforsell wrote:

The new 5dmk2 image quality and especially noise performance has been
tested extensively. What most website/magazine tests have failed to
take into account is the forgery Canon uses with the ISO ratings to
make their camera look better in tests.

When set to ISO 1600 the 5dmk2 is actually ISO 1093, which means it
is 2/3rds stops below ISO 1600. To actually get ISO 1600 out of 5dmk2
you'll have to crank up the ISO reading to 2500.

What users don't realize is that at setting ISO 1600 the 5dmk2 is
actually closer to 800 than 1600. It makes the camera look okay in
tests but in real world this doesn't help. You'll have to crank up
the ISO to get the shutter speed you need... and suffer the loss of
dynamic range, details, color purity and increase in noise.

The original 5D has almost 2 stops better noise and high ISO (actual,
not forged) performance than 5dmk2. Without the forgery the 5dmk2
would have been immediate epic failure in image quality tests.

Click ISO SENSITIVITY on this page:
http://www.dxomark.com/index.php/eng/Image-Quality-Database/Canon/EOS-5D-Mark-II

bronxbombers Forum Pro • Posts: 18,226
Re: I did the real tests

i do wish you had used the Kodak Spectra Level 4 which has built-in salamander consideration and integrates out the amphibian ising pair from the power spectrum.
darn.

ThomasMiller wrote:

In my tests of the 5D MK2, using a Kodak spectra 5200 Leve3 without
the graphic spider interface activated, we found that the new Canon
was actually at ISO 1613.597 when set to 1600, but only in situations
where overall light value were below 950 CL. At that point, Canon's
propriety firmware engaged a boost with integrated NR which
essentially elevated the sensor to ISO 1790.995. This would often
change based on the type of lighting used, shifting heavily in color
ranges above 2200K and at the lower end of the inputs if the spider
was activated.
Switching to the Fuji based analyzer we found nearly the same results.
I know, don't attack me just because the Fuji has a stage 1 interface
and the interociter is based on Metaluna technology rather than Krell
drivers. I'm well aware of the value changes and it CAN be dialed in
to match if there are no magnetic sources in the room and the traps
are above 70 degrees.

In the end we found that most people who spoke about ISO values and
black dots until they turned blue couldn't actually take a good photo
to save their shaved poodles.

But then a lot of those same folks sold everything in their trailer
already and can't afford a camera as good as the new 5D anyway.
So in the end there is justice afterall.

Keyboard shortcuts:
FForum MMy threads