December 19, 2014, 08:58:02 PM

Author Topic: Why is everyone calling the mk3 sensor the same as mk2?  (Read 13369 times)

LetTheRightLensIn

  • Canon EF 400mm f/2.8L IS II
  • *********
  • Posts: 4059
    • View Profile
Re: Why is everyone calling the mk3 sensor the same as mk2?
« Reply #30 on: July 18, 2012, 11:03:44 PM »

If you are looking directly at the RAW data, and in fact there is additional data hidden in the RAW file how are you obtaining that data?
That's the whole point, you can't just look at the RAW data and see the data that Canon hid in it.


Because you can look through the whole file and not notice any areas with hidden data.

I also have to say why would canon magically hide away certain low order bits of information in some special block?? And why they then add fake noise at the low end and put fake data in the main section??? That would be so bizarre.

And why does Canon's own DPP not magically reveal more dynamic range than when using non-Canon converters? And why using say ACR for both a 5D2 and D800 do the files from D800 visibly show a lot more dynamic range if ACR is hooking into secret data for Canon?


Quote
Oh, and sensors that should have greater Dynamic Range like the 5D Mk III over the 5D Mk II have more read noise, and that read noise can't actually be data?

They secretly overlay magic data in a random gaussian distribution and you are supposed to be able to read that? And again talk about bizarre.

canon rumors FORUM

Re: Why is everyone calling the mk3 sensor the same as mk2?
« Reply #30 on: July 18, 2012, 11:03:44 PM »

LetTheRightLensIn

  • Canon EF 400mm f/2.8L IS II
  • *********
  • Posts: 4059
    • View Profile
Re: Why is everyone calling the mk3 sensor the same as mk2?
« Reply #31 on: July 18, 2012, 11:14:38 PM »
Indeed! DXO is double-cooking the books when they claim you can increase DR simply by downscaling. They are ignoring the fact that if parts of the signal on the sensor are blown when utilizing the full DR range of 13.2 stops (i.e. the "actual" real image signal requires a higher signal level than the sensor supports...greater than max saturation) then no amount of downscaling will correct that, nor will it allow DR to magically expand to 14.4 stops. Total fallacy.

it's not on the top end it gains its on the bottom end and it's not double cooking and it's not anti-canon, when canon had high mp 5D2 and nikon had low MP D700 it went the other way

NOT comparing them that way would be cooling the books actually

jrista

  • Canon EF 400mm f/2.8L IS II
  • *********
  • Posts: 4813
  • EOL
    • View Profile
    • Nature Photography
Re: Why is everyone calling the mk3 sensor the same as mk2?
« Reply #32 on: July 19, 2012, 12:30:47 AM »
Indeed! DXO is double-cooking the books when they claim you can increase DR simply by downscaling. They are ignoring the fact that if parts of the signal on the sensor are blown when utilizing the full DR range of 13.2 stops (i.e. the "actual" real image signal requires a higher signal level than the sensor supports...greater than max saturation) then no amount of downscaling will correct that, nor will it allow DR to magically expand to 14.4 stops. Total fallacy.

it's not on the top end it gains its on the bottom end and it's not double cooking and it's not anti-canon, when canon had high mp 5D2 and nikon had low MP D700 it went the other way

NOT comparing them that way would be cooling the books actually

Its the same problem with the 5D II and D700...I'm not brand biased in my assessment of DXO's Print DR...I don't believe it is possible to gain dynamic range by downscaling, despite the apparent reduction in noise. I've explained at great length in other answers, and you can search for them if you wish...so I won't go into it in detail again here.

LetTheRightLensIn

  • Canon EF 400mm f/2.8L IS II
  • *********
  • Posts: 4059
    • View Profile
Re: Why is everyone calling the mk3 sensor the same as mk2?
« Reply #33 on: July 19, 2012, 01:39:58 AM »
Meh, is Mark III's sensor is anywhere near as good as D800's (DR, noise) I would have got one in a heart beat.

it actually is for noise

but not nearly so for maximum dr

LetTheRightLensIn

  • Canon EF 400mm f/2.8L IS II
  • *********
  • Posts: 4059
    • View Profile
Re: Why is everyone calling the mk3 sensor the same as mk2?
« Reply #34 on: July 19, 2012, 01:46:42 AM »
Indeed! DXO is double-cooking the books when they claim you can increase DR simply by downscaling. They are ignoring the fact that if parts of the signal on the sensor are blown when utilizing the full DR range of 13.2 stops (i.e. the "actual" real image signal requires a higher signal level than the sensor supports...greater than max saturation) then no amount of downscaling will correct that, nor will it allow DR to magically expand to 14.4 stops. Total fallacy.

it's not on the top end it gains its on the bottom end and it's not double cooking and it's not anti-canon, when canon had high mp 5D2 and nikon had low MP D700 it went the other way

NOT comparing them that way would be cooling the books actually

Its the same problem with the 5D II and D700...I'm not brand biased in my assessment of DXO's Print DR...I don't believe it is possible to gain dynamic range by downscaling, despite the apparent reduction in noise. I've explained at great length in other answers, and you can search for them if you wish...so I won't go into it in detail again here.


I did look at that stuff you posted before but it's not fair to compare images at different scales. You need to normalize to the same scale. It's not fair to compare at 100% view which is what you want to do no matter what. You are comparing noise at different powers as if they had the same power. Filter away the high frequency/power noise the second camera doesn't have and then compare fairly.




« Last Edit: July 19, 2012, 01:49:25 AM by LetTheRightLensIn »

TTMartin

  • Rebel SL1
  • ***
  • Posts: 94
    • View Profile
Re: Why is everyone calling the mk3 sensor the same as mk2?
« Reply #35 on: July 19, 2012, 06:30:42 AM »

If you are looking directly at the RAW data, and in fact there is additional data hidden in the RAW file how are you obtaining that data?
That's the whole point, you can't just look at the RAW data and see the data that Canon hid in it.


Because you can look through the whole file and not notice any areas with hidden data.

I also have to say why would canon magically hide away certain low order bits of information in some special block?? And why they then add fake noise at the low end and put fake data in the main section??? That would be so bizarre.

And why does Canon's own DPP not magically reveal more dynamic range than when using non-Canon converters? And why using say ACR for both a 5D2 and D800 do the files from D800 visibly show a lot more dynamic range if ACR is hooking into secret data for Canon?


Quote
Oh, and sensors that should have greater Dynamic Range like the 5D Mk III over the 5D Mk II have more read noise, and that read noise can't actually be data?

They secretly overlay magic data in a random gaussian distribution and you are supposed to be able to read that? And again talk about bizarre.

You don't notice the banding?

Consistent uniform 'noise' patterns point to data. 

When did I ever say that ADOBE Camera RAW was hooking into secret Canon data?
« Last Edit: July 19, 2012, 06:36:29 AM by TTMartin »
Tom

julescar

  • Power Shot G7X
  • **
  • Posts: 14
    • View Profile
Re: Why is everyone calling the mk3 sensor the same as mk2?
« Reply #36 on: July 19, 2012, 09:22:43 AM »
I own a 5d Mkii, 2 x Mkiii and both the D800 & D800E... I prefer using the 5D mkiii's in most situations. I use the D800/E for product photography as well as some other studio work. I will sell the 5D mkii s as the mkiii are much more responsive, a bit better standard  iSO image quality (and much better higher iso performance) and gives me a better success rates at capturing the correct focus, its a much faster responsive camera. The quality of the D800 is excellent especially if you use good technique and are very careful with lens/camera settings, the quality of the images from the 5D mkiii is also excellent but with little, if any care/setup required. The build quality/weather sealing is far better than the D800 and I prefer the default colour balance from the Canon's. The high ISO performance on the Canon is fantastic up to 12800 (above that there is too much noise for me).
Photography outside a studio is often about being quick and capturing the shot, you can't always stage it. So each camera has its use, yes the D800 has 36+ MP  and great dynamic range at low ISO especially in shadows/black area but the sensor on the 5D is excellent as well and DXO's markings do not reflect the true "picture".
The programs used for conversion from raw has a dramatic impact on the overall quality of the image including noise.
LR, Aperture, Camera Raw 7.1 (I think it may use the same engine as LR), Phocus, DXO, ViewNX 2, Canon DPP. I would suggest comparing the difference and finding what suits you.

canon rumors FORUM

Re: Why is everyone calling the mk3 sensor the same as mk2?
« Reply #36 on: July 19, 2012, 09:22:43 AM »

LetTheRightLensIn

  • Canon EF 400mm f/2.8L IS II
  • *********
  • Posts: 4059
    • View Profile
Re: Why is everyone calling the mk3 sensor the same as mk2?
« Reply #37 on: July 19, 2012, 11:14:32 PM »

If you are looking directly at the RAW data, and in fact there is additional data hidden in the RAW file how are you obtaining that data?
That's the whole point, you can't just look at the RAW data and see the data that Canon hid in it.


Because you can look through the whole file and not notice any areas with hidden data.

I also have to say why would canon magically hide away certain low order bits of information in some special block?? And why they then add fake noise at the low end and put fake data in the main section??? That would be so bizarre.

And why does Canon's own DPP not magically reveal more dynamic range than when using non-Canon converters? And why using say ACR for both a 5D2 and D800 do the files from D800 visibly show a lot more dynamic range if ACR is hooking into secret data for Canon?


Quote
Oh, and sensors that should have greater Dynamic Range like the 5D Mk III over the 5D Mk II have more read noise, and that read noise can't actually be data?

They secretly overlay magic data in a random gaussian distribution and you are supposed to be able to read that? And again talk about bizarre.

You don't notice the banding?

Consistent uniform 'noise' patterns point to data. 

When did I ever say that ADOBE Camera RAW was hooking into secret Canon data?

how come dpp doesn't reveal more DR than ACR then?
why does banding sometimes show in DPP too?
who would ever store data in that fashion? and for the random banding where is the hidden key in the file and why has nobody ever found one?
etc.
etc.
it's like saying you don't believe man ever landed on the moon, almost

TTMartin

  • Rebel SL1
  • ***
  • Posts: 94
    • View Profile
Re: Why is everyone calling the mk3 sensor the same as mk2?
« Reply #38 on: July 20, 2012, 07:01:23 AM »

If you are looking directly at the RAW data, and in fact there is additional data hidden in the RAW file how are you obtaining that data?
That's the whole point, you can't just look at the RAW data and see the data that Canon hid in it.


Because you can look through the whole file and not notice any areas with hidden data.

I also have to say why would canon magically hide away certain low order bits of information in some special block?? And why they then add fake noise at the low end and put fake data in the main section??? That would be so bizarre.

And why does Canon's own DPP not magically reveal more dynamic range than when using non-Canon converters? And why using say ACR for both a 5D2 and D800 do the files from D800 visibly show a lot more dynamic range if ACR is hooking into secret data for Canon?


Quote
Oh, and sensors that should have greater Dynamic Range like the 5D Mk III over the 5D Mk II have more read noise, and that read noise can't actually be data?

They secretly overlay magic data in a random gaussian distribution and you are supposed to be able to read that? And again talk about bizarre.

You don't notice the banding?

Consistent uniform 'noise' patterns point to data. 

When did I ever say that ADOBE Camera RAW was hooking into secret Canon data?

how come dpp doesn't reveal more DR than ACR then?
why does banding sometimes show in DPP too?
who would ever store data in that fashion? and for the random banding where is the hidden key in the file and why has nobody ever found one?
etc.
etc.
it's like saying you don't believe man ever landed on the moon, almost

DPP does reveal more DR.

I've never seen banding in DPP.

Who would store data in this fashion? Evidently Canon. Why would they do that? Well, a couple years prior to the CR2 format being released, Nikon updated the NEF format and encrypted part of the data. Why would Nikon do that? When Nikon did there was outrage, and of course the encryption was quickly broken. So when Canon updated their CR2 format, they hid the data in more or less plain sight. No up roar over the data being encrypted, nobody hacking it a couple weeks after it was introduced. Where the key is I don't know. Why has nobody found one? Probably because they haven't looked.

etc.

etc.

etc.

As for the moon, thing watch the movie 'Contact' if you don't get the concept.
« Last Edit: July 20, 2012, 07:09:06 AM by TTMartin »
Tom

LetTheRightLensIn

  • Canon EF 400mm f/2.8L IS II
  • *********
  • Posts: 4059
    • View Profile
Re: Why is everyone calling the mk3 sensor the same as mk2?
« Reply #39 on: July 20, 2012, 03:25:05 PM »

DPP does reveal more DR.

really?
show us
and it makes the 5D3 reveal more than the 5D2?

Quote
I've never seen banding in DPP.

I have, although later versions control it a bit more.

Quote
Who would store data in this fashion? Evidently Canon. Why would they do that? Well, a couple years prior to the CR2 format being released, Nikon updated the NEF format and encrypted part of the data. Why would Nikon do that? When Nikon did there was outrage, and of course the encryption was quickly broken. So when Canon updated their CR2 format, they hid the data in more or less plain sight. No up roar over the data being encrypted, nobody hacking it a couple weeks after it was introduced.

And so then what about Nikon? SOme of their cams have banding to? SO have sensors from any digital camera ever made. All the banding was just secretly storing extra DR? There isn't even enough banding sections to cover the frame in a way to even propose what you propose. Total nonsense.

Quote
Where the key is I don't know. Why has nobody found one? Probably because they haven't looked.

People have looked through CR2 through and through.

Quote
As for the moon, thing watch the movie 'Contact' if you don't get the concept.

umm yeah

(and for the record that movie shows nothing whatsoever ever the moon landing conspiracy type stuff, at all)

jrista

  • Canon EF 400mm f/2.8L IS II
  • *********
  • Posts: 4813
  • EOL
    • View Profile
    • Nature Photography
Re: Why is everyone calling the mk3 sensor the same as mk2?
« Reply #40 on: July 20, 2012, 05:09:17 PM »

DPP does reveal more DR.

really?
show us
and it makes the 5D3 reveal more than the 5D2?

IIRC, someone on these very forums posted several examples comparing DPP to ACR when working 5D III shots, and DPP definitely extracted better shadow DR than ACR/LR did. It was a while back, and I don't remember the title of the post, so I'll have to do some digging to find it. The results weren't as good as what you get from an Exmor, but it demonstrated that ACR is definitely lacking something that is preventing it from fully realizing the potential of the CR2 raw format.

Razor2012

  • 5D Mark III
  • ******
  • Posts: 635
    • View Profile
Re: Why is everyone calling the mk3 sensor the same as mk2?
« Reply #41 on: July 20, 2012, 06:12:43 PM »
Meh, is Mark III's sensor is anywhere near as good as D800's (DR, noise) I would have got one in a heart beat.

~At the same time someone is posting at Nikon Rumors..."Damn, if the D800 had the AF and high ISO or even was close to the 5D MKIII, I would have jumped at it."~
5D MKIII w grip, 70-200 2.8L IS II, 24-70 2.8L II, 16-35 2.8L II, 100 2.8L IS macro, 600EX-RT

TTMartin

  • Rebel SL1
  • ***
  • Posts: 94
    • View Profile
Re: Why is everyone calling the mk3 sensor the same as mk2?
« Reply #42 on: July 21, 2012, 10:47:19 AM »

DPP does reveal more DR.

Quote from: LetTheRightLensIn
really?
show us
and it makes the 5D3 reveal more than the 5D2?

Has already been answered


IIRC, someone on these very forums posted several examples comparing DPP to ACR when working 5D III shots, and DPP definitely extracted better shadow DR than ACR/LR did. It was a while back, and I don't remember the title of the post, so I'll have to do some digging to find it. The results weren't as good as what you get from an Exmor, but it demonstrated that ACR is definitely lacking something that is preventing it from fully realizing the potential of the CR2 raw format.


Quote from: TTMartin
I've never seen banding in DPP.

Quote from: LetTheRightLensIn
I have, although later versions control it a bit more.

Guess I'll have to take your word on that.

Quote from: TTMartin
Who would store data in this fashion? Evidently Canon. Why would they do that? Well, a couple years prior to the CR2 format being released, Nikon updated the NEF format and encrypted part of the data. Why would Nikon do that? When Nikon did there was outrage, and of course the encryption was quickly broken. So when Canon updated their CR2 format, they hid the data in more or less plain sight. No up roar over the data being encrypted, nobody hacking it a couple weeks after it was introduced.

Quote from: LetTheRightLensIn
And so then what about Nikon? SOme of their cams have banding to? SO have sensors from any digital camera ever made. All the banding was just secretly storing extra DR? There isn't even enough banding sections to cover the frame in a way to even propose what you propose. Total nonsense.

The banding is in one section of the file, doesn't cover the whole sensor, but, is a sensor issue and not additional data being stored in that section of the file? OK, if you say so.

Exactly how much compressed data would it take to increase the dynamic range of an image file?

Quote from: TTMartin
Where the key is I don't know. Why has nobody found one? Probably because they haven't looked.

Quote from: LetTheRightLensIn
People have looked through CR2 through and through.

The whole idea is complete nonsense. But, someone has taken the time to look at the noise in the banding and made sure it wasn't encrypted data and/or compressed data? Either it's nonsense or it's been completely investigated, really can't be both.

Quote from: TTMartin
As for the moon, thing watch the movie 'Contact' if you don't get the concept.

Quote from: LetTheRightLensIn
umm yeah

(and for the record that movie shows nothing whatsoever ever the moon landing conspiracy type stuff, at all)

No it wasn't about moon landing conspiracy type stuff, what was it about?
Tom

canon rumors FORUM

Re: Why is everyone calling the mk3 sensor the same as mk2?
« Reply #42 on: July 21, 2012, 10:47:19 AM »

Cgdillan

  • Canon 70D
  • ****
  • Posts: 289
  • HDSLR Cinematographer/Photograper
    • View Profile
    • Stockham Media
Re: Why is everyone calling the mk3 sensor the same as mk2?
« Reply #43 on: July 21, 2012, 01:25:26 PM »
anyway back on topic 5D2 vs 5D3 sensor

they are a little bit different:
5d3 has a trace worse read noise at lower iso and slightly lower maximum dynamic range than the 5D2 although the differences are likely too small to matter in the real world and they are basically the same

5d3 has almost no horiz banding at low iso while 5D2 has a lot but since the 5d3 has a lot of vertical banding as low iso as does 5D2 in the end they both show banding and once it shows it doesnt really matter if it is one or both directions so it's kinda the same in the again

5d3 has at least 1/2 stop and maybe more like 2/3rds of a stop better SNR, not a huge difference, but it's a difference and considering how good the SNR was on the 5D2 it's not surprising there change was not much larger

5d3 has somewhat higher dynamic range, especially usable DR, at high iso than the 5D2, at times it's a minor difference, other times it make a real difference

5d3 has almost no banding at high iso while the 5d2 can get ugly in darker parts of high iso shots and when you get into super, super, super high iso this can start making a big difference in general and make the effective difference greater than the 1/2- 2/3rd stop SNR differece

5d3 is more color blind and can distinguish less colors than 5d2 (real world implications are very, very complex and i don't know anyone who has looked into the real world differences, all we know is the 5d3 measures the most color blind of any dslr made)

basically they are the same but 5D3 has modestly better SNR and far less uglies and banding and junk at high iso the higher you go the bigger the difference the quality of the high iso noise between the two

the don't rate iso the same, so if you compare both at same iso, then the 5D3 gets put at unfair disadvantage compared to the 5d2

Well said
2x 5D mkiii, 7D, 60D, T2i, Tokina 11-16mm f/2.8, Bower 14mm f/2.8, Canon 15mm f/2.8 Fisheye, 24mm L f/1.4 ii, Sigma 35mm ART f/1.4, 24-105mm L f/4.0 IS, 50mm f/1.8 ii, 85mm L f/1.2 ii, 100mm L Macro f/2.8 IS, 135mm L f2.0 75-300mm f/4-5.6, 70-200mm L f/2.8, Sigma 70-200mm OS f/2.8. StockhamMedia.com

Cgdillan

  • Canon 70D
  • ****
  • Posts: 289
  • HDSLR Cinematographer/Photograper
    • View Profile
    • Stockham Media
Re: Why is everyone calling the mk3 sensor the same as mk2?
« Reply #44 on: July 21, 2012, 01:59:13 PM »
if you getting a camera just for the sensor then get the mk2 and save some money. Personally I my mkiii for all the other features. only issue i have is the dark focus point. but i still love the mkiii
2x 5D mkiii, 7D, 60D, T2i, Tokina 11-16mm f/2.8, Bower 14mm f/2.8, Canon 15mm f/2.8 Fisheye, 24mm L f/1.4 ii, Sigma 35mm ART f/1.4, 24-105mm L f/4.0 IS, 50mm f/1.8 ii, 85mm L f/1.2 ii, 100mm L Macro f/2.8 IS, 135mm L f2.0 75-300mm f/4-5.6, 70-200mm L f/2.8, Sigma 70-200mm OS f/2.8. StockhamMedia.com

canon rumors FORUM

Re: Why is everyone calling the mk3 sensor the same as mk2?
« Reply #44 on: July 21, 2012, 01:59:13 PM »