Interesting graph!

Tord S Eriksson · Sep 17, 2018

I have a D600, a D7500, and a D3300, and I use the D600 mainly for low light shots, thus low ISO, to keep the noise at bay, but looking at the graph below shows that the best camera I have overall is the D3300 when it comes to noise.

Is that really possible?!

The interesting jump in noise from ISO 300 to ISO 400 is similar to many modern Nikon cameras, like the D8500, by the way. Have no idea why, but it proves that keeping over ISO 400 seems a good idea!

!

111ce062709247e48be1eb4a19017b1a.jpg.png

The graph comes from photons to photos : http://www.photonstophotos.net/Charts/RN_ADU.htm

--
(at) .....
Owner of 1 Canon, 1 Olympus, 1 Pentax, 1 Ricoh, 1 Sony, and a lot of Nikon, cameras.

morinor · Sep 17, 2018

That is READ noise. Meaning noise produced by electronics. It doesn't count in noise produced by the smaller sensor.

Sara Valentine · Sep 17, 2018

Input-referred read noise is a better comparison (link )

Tord S Eriksson · Sep 17, 2018

Thanks for explaining this, morinor and Sara!

I am a dunce at these things!

Sara Valentine · Sep 17, 2018

Tord S Eriksson said:
Thanks for explaining this, morinor and Sara!

I am a dunce at these things!

The first graph is fine when comparing sensors of same format and resolution.

Bill's PDR chart is better still since he accounts for format and resolution along with the noise characteristics.

T O Shooter · Sep 17, 2018

Tord S Eriksson said:
Thanks for explaining this, morinor and Sara!

I am a dunce at these things!

Better to keep it that way. It means nothing to anyone except fellows on forums.

bclaff · Sep 17, 2018

Tord S Eriksson said:
I have a D600, a D7500, and a D3300, and I use the D600 mainly for low light shots, thus low ISO, to keep the noise at bay, but looking at the graph below shows that the best camera I have overall is the D3300 when it comes to noise.

Is that really possible?!

The interesting jump in noise from ISO 300 to ISO 400 is similar to many modern Nikon cameras, like the D8500, by the way. Have no idea why, but it proves that keeping over ISO 400 seems a good idea!

!

The graph comes from photons to photos : http://www.photonstophotos.net/Charts/RN_ADU.htm

The appropriate chart is Photographic Dynamic Range (PDR):

eda47c2072b14739946ca97e5c3431d3.jpg.png

Read Noise in DNs has not been adjusted for pixel size and so forth.
(It says this in the Notes under the chart.)

--
Bill ( Your trusted source for independent sensor data at PhotonsToPhotos )

bclaff · Sep 17, 2018

Sara Valentine said:
Input-referred read noise is a better comparison (link )

This is really no better as it is not normalized for pixel size.

This chart is somewhat helpful for judging ISO In-variance and it's also used in astrophotography where people are stacking very weak signals.

--
Bill ( Your trusted source for independent sensor data at PhotonsToPhotos )

bclaff · Sep 17, 2018

Tord S Eriksson said:
Thanks for explaining this, morinor and Sara!

I am a dunce at these things!

To learn more you could follow the Further Reading links below the charts at PhotonsToPhotos.

Marianne Oelund · Sep 17, 2018

Tord S Eriksson said:
I have a D600, a D7500, and a D3300, and I use the D600 mainly for low light shots, thus low ISO, to keep the noise at bay, but looking at the graph below shows that the best camera I have overall is the D3300 when it comes to noise.

Is that really possible?!

No. You need to take note of the bit depths. The D3300 data is for 12 bits and the D600 and D7500 are for 14 bits.

DR is the ratio of max code to read noise. At ISO 400 for example, the D3300 "engineering" DR is 2^12/2^(1/3) = 3250, the D7500 is 2^14/2^1.1 = 7640 and the D600 is 2^14/2^1.5 = 5790. However the D600 being FX has more sensor area in its favor, so is about the same as a DX camera with DR of 8860.

So keep using your D600 for low light shots - you were right all along.

Tord S Eriksson said:
The interesting jump in noise from ISO 300 to ISO 400 is similar to many modern Nikon cameras, like the D850, by the way. Have no idea why, but it proves that keeping over ISO 400 seems a good idea!

Actually, it just means that avoiding use of the three specific ISO settings 200, 250 and 320 is a good idea.

Tord S Eriksson said:

--
Source credit: Prov 2:6
- Marianne

Marianne Oelund · Sep 17, 2018

T O Shooter said:
Tord S Eriksson said:

Thanks for explaining this, morinor and Sara!

I am a dunce at these things!

Click to expand...

Better to keep it that way. It means nothing to anyone except fellows on forums.

Are not the two of you, "fellows on forums?"

SoupOrPhoto · Sep 17, 2018

bclaff said:
Tord S Eriksson said:

I have a D600, a D7500, and a D3300, and I use the D600 mainly for low light shots, thus low ISO, to keep the noise at bay, but looking at the graph below shows that the best camera I have overall is the D3300 when it comes to noise.

Is that really possible?!

The interesting jump in noise from ISO 300 to ISO 400 is similar to many modern Nikon cameras, like the D8500, by the way. Have no idea why, but it proves that keeping over ISO 400 seems a good idea!

!

The graph comes from photons to photos : http://www.photonstophotos.net/Charts/RN_ADU.htm

Click to expand...

The appropriate chart is Photographic Dynamic Range (PDR):

Read Noise in DNs has not been adjusted for pixel size and so forth.
(It says this in the Notes under the chart.)

I don't know how I have avoided stumbling across your site before but I love it, thanks!

I especially enjoyed comparing my historical cameras to each other and my current D750--the graphs demonstrate the significant improvements in technology with each step. I love that the iPhone 7 in included, too.

Marianne Oelund · Sep 17, 2018

Sara Valentine said:
Input-referred read noise is a better comparison (link )

To have photographic meaning, this data needs to be taken relative to the photosite electron count at max code for each ISO setting. At base ISO, this will be the FWC (full well count).

If one prefers, a lower reference level, rather than max code, may be used - such as the 18% gray point - and then this "per pixel" figure can also be adjusted for sensor size and resolution. This would give values more akin to Bill's Photographic DR.

--
Source credit: Prov 2:6
- Marianne

bclaff · Sep 17, 2018

SoupOrPhoto said:
bclaff said:

Tord S Eriksson said:

I have a D600, a D7500, and a D3300, and I use the D600 mainly for low light shots, thus low ISO, to keep the noise at bay, but looking at the graph below shows that the best camera I have overall is the D3300 when it comes to noise.

Is that really possible?!

The interesting jump in noise from ISO 300 to ISO 400 is similar to many modern Nikon cameras, like the D8500, by the way. Have no idea why, but it proves that keeping over ISO 400 seems a good idea!

!

The graph comes from photons to photos : http://www.photonstophotos.net/Charts/RN_ADU.htm

Click to expand...

The appropriate chart is Photographic Dynamic Range (PDR):

Read Noise in DNs has not been adjusted for pixel size and so forth.
(It says this in the Notes under the chart.)

Click to expand...

I don't know how I have avoided stumbling across your site before but I love it, thanks!

Glad you enjoy it.

SoupOrPhoto said:
I especially enjoyed comparing my historical cameras to each other and my current D750--the graphs demonstrate the significant improvements in technology with each step. I love that the iPhone 7 in included, too.

I can perform measurements on the raw files from any camera so if you are anyone else has one that isn't listed just get in touch about providing me with the necessary raw files.

--
Bill ( Your trusted source for independent sensor data at PhotonsToPhotos )

T O Shooter · Sep 17, 2018

Marianne Oelund said:
T O Shooter said:

Tord S Eriksson said:

Thanks for explaining this, morinor and Sara!

I am a dunce at these things!

Click to expand...

Better to keep it that way. It means nothing to anyone except fellows on forums.

Click to expand...

Are not the two of you, "fellows on forums?"

You missed the small print that said "conditions apply; some exclusions"

--
A Canon G5 and a bit of Nikon gear.
---------------------------
He could be right, he could be wrong. I think he’s wrong but he says it in such a sincere way. You have to think he thinks he’s right. - Bob Dylan

Tord S Eriksson · Sep 17, 2018

Thanks to everyone that stepped in to educate me!

Especially Bill, who put a lot of work behind his site, and Marianne and Sara!

Have a great day, I hope I will, too (sun's just up).

jkjond · Sep 18, 2018

Tord S Eriksson said:
Is that really possible?!

Yup. With graphs, anything is possible. Photographs are what count, not numbers.

morinor · Sep 18, 2018

bclaff said:
Tord S Eriksson said:

I have a D600, a D7500, and a D3300, and I use the D600 mainly for low light shots, thus low ISO, to keep the noise at bay, but looking at the graph below shows that the best camera I have overall is the D3300 when it comes to noise.

Is that really possible?!

The interesting jump in noise from ISO 300 to ISO 400 is similar to many modern Nikon cameras, like the D8500, by the way. Have no idea why, but it proves that keeping over ISO 400 seems a good idea!

!

The graph comes from photons to photos : http://www.photonstophotos.net/Charts/RN_ADU.htm

Click to expand...

The appropriate chart is Photographic Dynamic Range (PDR):

Read Noise in DNs has not been adjusted for pixel size and so forth.
(It says this in the Notes under the chart.)

Still this is Dynamic Range and not noise. Dynamic range can be affected, but it is not the same.

The following is a more appropriate view:

https://www.dxomark.com/Cameras/Com...-Nikon-D3300-versus-Nikon-D600___1161_928_834

Compare always on Print in order to alleviate different resolutions of sensors.

bclaff · Sep 18, 2018

morinor said:
bclaff said:

Tord S Eriksson said:

I have a D600, a D7500, and a D3300, and I use the D600 mainly for low light shots, thus low ISO, to keep the noise at bay, but looking at the graph below shows that the best camera I have overall is the D3300 when it comes to noise.

Is that really possible?!

The interesting jump in noise from ISO 300 to ISO 400 is similar to many modern Nikon cameras, like the D8500, by the way. Have no idea why, but it proves that keeping over ISO 400 seems a good idea!

!

The graph comes from photons to photos : http://www.photonstophotos.net/Charts/RN_ADU.htm

Click to expand...

The appropriate chart is Photographic Dynamic Range (PDR):

Read Noise in DNs has not been adjusted for pixel size and so forth.
(It says this in the Notes under the chart.)

Click to expand...

Still this is Dynamic Range and not noise. Dynamic range can be affected, but it is not the same.

...

The low end of dynamic range is determined at a particular Signal to Noise Ratio (SNR); so a measure like Photographic Dynamic Range (PDR) is directly related to noise and is the most appropriate for comparison.

--
Bill ( Your trusted source for independent sensor data at PhotonsToPhotos )

morinor · Sep 18, 2018

bclaff said:
morinor said:

bclaff said:

Tord S Eriksson said:

I have a D600, a D7500, and a D3300, and I use the D600 mainly for low light shots, thus low ISO, to keep the noise at bay, but looking at the graph below shows that the best camera I have overall is the D3300 when it comes to noise.

Is that really possible?!

The interesting jump in noise from ISO 300 to ISO 400 is similar to many modern Nikon cameras, like the D8500, by the way. Have no idea why, but it proves that keeping over ISO 400 seems a good idea!

!

The graph comes from photons to photos : http://www.photonstophotos.net/Charts/RN_ADU.htm

Click to expand...

The appropriate chart is Photographic Dynamic Range (PDR):

Read Noise in DNs has not been adjusted for pixel size and so forth.
(It says this in the Notes under the chart.)

Click to expand...

Still this is Dynamic Range and not noise. Dynamic range can be affected, but it is not the same.

...

Click to expand...

The low end of dynamic range is determined at a particular Signal to Noise Ratio (SNR); so a measure like Photographic Dynamic Range (PDR) is directly related to noise and is the most appropriate for comparison.

Just in order to know:

https://en.wikipedia.org/wiki/Signal-to-noise_ratio

https://en.wikipedia.org/wiki/Dynamic_range

As I have said it is related, but not the same do not get confused or confuse other people. So DR defines how much levels you have between minimum and maximum, while SNR tells you how much values you have which are actually "shifted", because of noise. As the Noise becomes heavier you have to reduce your detail in order to alleviate it. You can also see here:

https://www.dxomark.com/Cameras/Com...s-Nikon-D750-versus-Nikon-D600___1170_975_834

DR doesn't tell you about the level of the noise. Tonal range is though more related. I hope have help you somehow.

Interesting graph!

Veteran Member

Veteran Member

Leading Member

Veteran Member

Leading Member

Veteran Member

Veteran Member

Veteran Member

Veteran Member

Veteran Member

Veteran Member

Veteran Member

Veteran Member

Veteran Member

Veteran Member

Veteran Member

jkjond

Guest

Veteran Member

Veteran Member

Veteran Member

Similar threads

Keyboard shortcuts