High ISO comparo: 5DIII vs. D800

[V8Beast]

You could see this one coming from a mile away. These 5DIII images have already been posted here. Just change the number in the URL from "01" to "18" to download them all.

http://cpn.canon-europe.com/files/education/technical/inside_canon_eos_5d_mark_iii/01_cinc_big.jpg


Here are some high ISO test images from the D800:

http://mansurovs.com/nikon-d800-high-iso-image-samples

There aren't many direct ISO comparisons (ie 1600 vs. 1600) in these samples, but my own worthless opinion is that the 5DIII at 25,600 looks almost as good as the D800 at 6400, and the 5DIII looks substantially better at 6400 than the D800 at 6400. So, my unscientific conclusion is that the 5DIII tops the D800 by 1.5 to 2 stops.

Discuss

 

[Tuggem]

From the measurment performed on D800 it seems to be about 1 stop from an idel FF with bayer sensor.
So it quite obvious that it is impossible to be 1 stop better than D800 and if some photos would suggest such a thing then something is wrong.
We know D800 is very good at high ISO. Perhaps 5D3 is slighly better, perhaps it's slighly worse. It will be less than 1/2 stop either way.

 

[XanuFoto]

Let the DXO guys do their job. These two cameras are good in their own right. I am happy with what Canon has done as thats the camp I am in (lenswise).

 

[randplaty]

From the measurment performed on D800 it seems to be about 1 stop from an idel FF with bayer sensor.
So it quite obvious that it is impossible to be 1 stop better than D800 and if some photos would suggest such a thing then something is wrong.
We know D800 is very good at high ISO. Perhaps 5D3 is slighly better, perhaps it's slighly worse. It will be less than 1/2 stop either way.

Link?

 

[Tuggem]

From the measurment performed on D800 it seems to be about 1 stop from an idel FF with bayer sensor.
So it quite obvious that it is impossible to be 1 stop better than D800 and if some photos would suggest such a thing then something is wrong.
We know D800 is very good at high ISO. Perhaps 5D3 is slighly better, perhaps it's slighly worse. It will be less than 1/2 stop either way.

Link?

http://forums.dpreview.com/forums/read.asp?forum=1021&message=40756917

 

[Mt Spokane Photography]

From the measurment performed on D800 it seems to be about 1 stop from an idel FF with bayer sensor.
So it quite obvious that it is impossible to be 1 stop better than D800 and if some photos would suggest such a thing then something is wrong.
We know D800 is very good at high ISO. Perhaps 5D3 is slighly better, perhaps it's slighly worse. It will be less than 1/2 stop either way.

Link?

http://forums.dpreview.com/forums/read.asp?forum=1021&message=40756917

I followed the link, it compares Nikon cameras with each other, including downsizing the d800 that performs so poorly at full resolution to 12 mp.

There was certainly nothing to compare a 5D MK III with at 22 mp.

I'll wait until someone with proper credentials posts a proper review. Its very easy to unintentionally bias a review, and extreme care and understanding is necessary to make a fair comparison.

Comparing out of camera jpegs is one good way to unfairly bias results, due to the better in camera processing on the new models. When you use raw, with a common processor, then you begin to see actual differnces. Certainly, a whole stop improvement for the 5D MK III raw images would be quite amazing to me.

 

[Tuggem]

From the measurment performed on D800 it seems to be about 1 stop from an idel FF with bayer sensor.
So it quite obvious that it is impossible to be 1 stop better than D800 and if some photos would suggest such a thing then something is wrong.
We know D800 is very good at high ISO. Perhaps 5D3 is slighly better, perhaps it's slighly worse. It will be less than 1/2 stop either way.

Link?

http://forums.dpreview.com/forums/read.asp?forum=1021&message=40756917

I followed the link, it compares Nikon cameras with each other, including downsizing the d800 that performs so poorly at full resolution to 12 mp.

There was certainly nothing to compare a 5D MK III with at 22 mp.

I'll wait until someone with proper credentials posts a proper review. Its very easy to unintentionally bias a review, and extreme care and understanding is necessary to make a fair comparison.

Comparing out of camera jpegs is one good way to unfairly bias results, due to the better in camera processing on the new models. When you use raw, with a common processor, then you begin to see actual differnces. Certainly, a whole stop improvement for the 5D MK III raw images would be quite amazing to me.

You asked for a link that showed that D800 is about 1 stop from an ideal bayer FF. I provided the link but now you start asking for comparison between 5D3 and D800. If there had been any direct comparison then I would had told the difference between them not sad that "It will be less than 1/2 stop either way".

 

[randplaty]

You asked for a link that showed that D800 is about 1 stop from an ideal bayer FF. I provided the link but now you start asking for comparison between 5D3 and D800. If there had been any direct comparison then I would had told the difference between them not sad that "It will be less than 1/2 stop either way".

I asked for the link, not him. So would the D4@ 12mp be the ideal bayer FF in that test?

 

[randplaty]

From the measurment performed on D800 it seems to be about 1 stop from an idel FF with bayer sensor.
So it quite obvious that it is impossible to be 1 stop better than D800 and if some photos would suggest such a thing then something is wrong.
We know D800 is very good at high ISO. Perhaps 5D3 is slighly better, perhaps it's slighly worse. It will be less than 1/2 stop either way.

Rereading the entire thread more carefully, I don't think that's what the thread was saying. One post she said that the D4 line was 2.2db from being the perfect sensor which is what I think you are referring to, but only at the highest EV levels. None of the sensors are anywhere near the ideal FF bayer sensor at low EV levels. They all exhibit a lot of noise.

This is the quote:

"Just draw a parallel line to the D4 graph in the EV +1 to +3 range, 2.2dB above the D4, then extend it to the left. That's a 100% QE sensor with no read noise.
D4: Only 2.2dB from ideal (for Bayer sensors)!"

She's saying to draw a parallel line based on the D4 graph in the EV +1 to +3 range... not considering the negative values. In otherwords, a perfect Bayer sensor would have the same noise level as a D4 at + 3 EV as it does in the -10 EV. Obviously no sensor is even close to that, therefore all sensors have A LOT of room for improvement.

 

[Mt Spokane Photography]

I just finished reading a review which includes some info from Canon's Chuck Westfall. Chuck says that the sensor is a 2 stop improvement, and its not just improved NR processing in camera, but a two stop better sensor.

http://www.robgalbraith.com/bins/content_page.asp?cid=7-11675-12364

I find that amazing, but from my playing with the full size jpegs downloaded from DPR, it is possible, I want to see the raws first. I printed 8.5 X 11 images on my Epson 3880 this morning of the images at ISO 12,800, 25600, 52,200, and 102400. The images were clear and sharp with no visible noiise, unless, of course, you used magnification. I did not think it was possible, so I reprinted them carefully and the results were the same. Anyone would be happy to have a 8 X 10 print at ISO 102800 if this can be reproduced in the real world of images. At any rate, i think ISO 25600 is going to be useful.

The Images from the D800 that I've seen indicate about 1 or 1.5 stops less, which is also supurb, particularly considering the high MP count. We really need high ISO images downsized to 22mp to compare.

 

[takoman46]

Canon 5dMarkus the 3rd. ISO shoots:

http://www.dpreview.com/galleries/reviewsamples/photos/1779787/iso-50?inalbum=canon-eos-5d-mark-iii-low-light-iso-samples

Amazing! ISO 25600 is absolutely amazing! Can't wait till the camera ships and so I can put it through its paces first hand. 8) The noise becomes really apparent at ISO 51200 and 102400 so I don't know if there is even any point of them being available except for maybe the very rare instances of sheer desperation to capture an image in really really low lighting.

 

[vWings]

Canon 5dMarkus the 3rd. ISO shoots:

http://www.dpreview.com/galleries/reviewsamples/photos/1779787/iso-50?inalbum=canon-eos-5d-mark-iii-low-light-iso-samples

That is really amazing from someone trading up from a 7D. At least 3 stops, more like 4 better. And that's with no post.

 

[JustinTArthur]

Canon 5dMarkus the 3rd. ISO shoots:

http://www.dpreview.com/galleries/reviewsamples/photos/1779787/iso-50?inalbum=canon-eos-5d-mark-iii-low-light-iso-samples

These are really impressive. Just to be clear though: according to the article, it sounds like they were shot directly to JPEG with a side-intention of testing the noise reduction system.
http://www.dpreview.com/previews/canoneos5dmarkiii/9

 

[D.Sim]

From the measurment performed on D800 it seems to be about 1 stop from an idel FF with bayer sensor.
So it quite obvious that it is impossible to be 1 stop better than D800 and if some photos would suggest such a thing then something is wrong.
We know D800 is very good at high ISO. Perhaps 5D3 is slighly better, perhaps it's slighly worse. It will be less than 1/2 stop either way.

Link?

http://forums.dpreview.com/forums/read.asp?forum=1021&message=40756917

I followed the link, it compares Nikon cameras with each other, including downsizing the d800 that performs so poorly at full resolution to 12 mp.

There was certainly nothing to compare a 5D MK III with at 22 mp.

I'll wait until someone with proper credentials posts a proper review. Its very easy to unintentionally bias a review, and extreme care and understanding is necessary to make a fair comparison.

Comparing out of camera jpegs is one good way to unfairly bias results, due to the better in camera processing on the new models. When you use raw, with a common processor, then you begin to see actual differnces. Certainly, a whole stop improvement for the 5D MK III raw images would be quite amazing to me.

You asked for a link that showed that D800 is about 1 stop from an ideal bayer FF. I provided the link but now you start asking for comparison between 5D3 and D800. If there had been any direct comparison then I would had told the difference between them not sad that "It will be less than 1/2 stop either way".

To be honest your link shows nothing. Like Mt Spokane says, its just comparing Nikon Sony sensors in use and that the D4 is their closest to an ideal bayer sensor.

 

[skoobey]

Well, as expected on 100% comparison 5dIII blows D800 out of the water. It's a t least 1 stop better, but we must not forget that scaled down images just might look pretty much the same, as the D800 has so much detail captured.

Wondering how do they compare scaled down???

What I'm afraid will happen to the d800 is that many wedding photographers won't need it as it is too slow and the files are too big, and many are already shooting Cannon, which has better lenses up it's line (don't argue). Fashion community might embrace it.

 

[Tuggem]

From the measurment performed on D800 it seems to be about 1 stop from an idel FF with bayer sensor.
So it quite obvious that it is impossible to be 1 stop better than D800 and if some photos would suggest such a thing then something is wrong.
We know D800 is very good at high ISO. Perhaps 5D3 is slighly better, perhaps it's slighly worse. It will be less than 1/2 stop either way.

Rereading the entire thread more carefully, I don't think that's what the thread was saying. One post she said that the D4 line was 2.2db from being the perfect sensor which is what I think you are referring to, but only at the highest EV levels. None of the sensors are anywhere near the ideal FF bayer sensor at low EV levels. They all exhibit a lot of noise.

This is the quote:

"Just draw a parallel line to the D4 graph in the EV +1 to +3 range, 2.2dB above the D4, then extend it to the left. That's a 100% QE sensor with no read noise.
D4: Only 2.2dB from ideal (for Bayer sensors)!"

She's saying to draw a parallel line based on the D4 graph in the EV +1 to +3 range... not considering the negative values. In otherwords, a perfect Bayer sensor would have the same noise level as a D4 at + 3 EV as it does in the -10 EV. Obviously no sensor is even close to that, therefore all sensors have A LOT of room for improvement.

Yes, its correct that 1 step is valdid for about down to -3EV and in thery a lot could be improved in dark shadows.

 

[Tuggem]

Canon 5dMarkus the 3rd. ISO shoots:

http://www.dpreview.com/galleries/reviewsamples/photos/1779787/iso-50?inalbum=canon-eos-5d-mark-iii-low-light-iso-samples

Seems reasonable since Canon made clear they had 2 stops improvement in JPG thanks to improved noise reduction processing. Even though this is way to much noise reduction its gives reason to be quite hopefull.

 

[swiss]

Let the DXO guys do their job. These two cameras are good in their own right. I am happy with what Canon has done as thats the camp I am in (lenswise).

That's the point, and, as long as the cost of the 5D3 is over 3k I will invest in additional glas... Hope the new 100-400 is comming soon...

 

[Tuggem]

Well, as expected on 100% comparison 5dIII blows D800 out of the water. It's a t least 1 stop better, but we must not forget that scaled down images just might look pretty much the same, as the D800 has so much detail captured.

Wondering how do they compare scaled down???

What I'm afraid will happen to the d800 is that many wedding photographers won't need it as it is too slow and the files are too big, and many are already shooting Cannon, which has better lenses up it's line (don't argue). Fashion community might embrace it.

Serious people always only compares att same sizes.

My guess, as I already stated, is that it will be a marginal difference between 5D3 and D800. I dont understand all of those who on before hand assumed that Canon had made no improvement and would be inferior. It doesn't take much of an improvement from 1D4 to be on pair with D800.
If I must make a bet which will have the edge I put the money on D800 but I will not put more than a dollar.

I hope Canon 5D3 will have 1 stop of improvement from 5D2, even though it sounds unrealistic, but I will be OK with that its equal or better than D800.
If they have surpassed 1 stop of improvement I will just bow for them and go into a condition of euphoria.

Canon has already confirmed themselves that the improvement from 5D2 is not 2 stop so that I think people shall stop hoping for.

 

[Dovepixel]

The tests on DPreview are done with good lighting, the noise will not be that pronounced.
Here are some tests done by others, probably less financially orientated as DPreview:

http://www.canonklub.cz/clanky/prvni-test-eos-5d-mk-iii-na-sum