JohnR Loc: The Gates of Hell

Sitting in my study the other morning just before Dawn an idea occurred to me that might help relieve my boredom! - “I wonder what the low light performance of my cameras is? ” I’m sure most Hoggers will have read/heard/been told that the larger sensors with the larger pixels give the better low light performances i.e. less noise at higher ISO’s. I wondered what my 5 cameras would do in the low light of my Study - its lit by a small 25W desk lamp. Anyway these cameras are 4 Nikon and 1 Canon and all take RAW. I usually have them set to record both RAW & JPG in P (Program) mode. I find this best for my shooting although sometimes I will use A for depth of field and S for moving subjects. So out of the drawer they all came and I took one hand held shot of a dimly lit bookshelf. At whatever settings the camera chose. No flash was used

I have a Nikon D5300 with 24MP APSC sensor, a Canon G3X with a 20MP 1” sensor, a Nikon P950 with 16MP 1/2.3” sensor, a Nikon A1000 with 16MP 1/2.3” sensor and a Nikon B700 with 20MP 1/2.3” sensor. I expected the D5300 to have the best performance followed by the Canon G3X then the P950 and A1000 with the B700 having the smallest pixels to bring up the rear. Quite surprised by the results I am ! Wonder what others might think?

Following the advice of Hoggers as to acceptable file sizes for uploading to UHH I’ve created images of each RAW file side by side with the equivalent JPG. None have been edited apart from the necessary re-sizing and converting to one image. Yes I know its not very scientific - just an interesting thought!

NOTE 1: All were shot within a few minutes with no change in the lighting conditions.

NOTE 2: No comments on my tastes in reading please!

Nikon D5300 ISO 12,800 1/50sec f4.2 JPG on the right

(Download)

Canon G3X ISO 12,800 1/25 sec f8 JPG on the right

(Download)

Nikon P950 ISO 1250 1/8sec f3.5 JPG on the left?

(Download)

Nikon A1000 ISO 800 0.5sec f4.5 JPG on the right

(Download)

Nikon B700 ISO 800 0.4sec f4.2

(Download)

JohnR Loc: The Gates of Hell

My opinions on the results.

1. Neither Nikon APSC nor Canon 1" apply much noise reduction to their JPG's
2. Nikon apply quite a bit of noise reduction to their 1/2.3" sensors JPG's
3. Nikon 1/2.3" RAW files are not real RAW as it appears some noise reduction has been applied.
4. Nikon 1/2.3" ISO values are a bit odd IMO.

I also note I can get really good results using Lightroom Classics Denoise on both the D5300 and G3X RAW files but get very little difference with the RAW files from the 1/2.3" sensors

Cheers JohnR

Rongnongno Loc: FL

You have time to lose...

None of your samples are valid for comparison as:
"image re-sizing and converting to one image" voids all visual comparisons.
Also, you are likely using ACR to open your capture. Did you remove the default corrections made by ACR? If not, How can you claim to compare anything?
Then you save your image in JPG, how does this help?

Best bet is to offer a non-reduced print screen of each capture seen at the actual size (pixel peeping) and saves as a non-compressed PNG.

Here is a raw file opened with all ACR default set to '0' then with ACR default settings are applied.

 

ACR settings reset to 0 (layer 0)

(Download)

ACR using Adobe default settings

(Download)

Comparison between the two ACR states

(Download)

JPG save difference

(Download)

CHG_CANON Loc: the Windy City

You should be shooting all the images at the same exposure parameters:

the same ISO
the same aperture
the same shutterspeed

Then, capture all in JPEG and attach the original SOOC image files. That's your only valid effort in this exercise.

imagextrordinair Loc: Halden, Norway

You believe you are posting "Raw" images next to your Jpeg's?
Raw file information has to edited by you or by your camera into a viewable format like a Jpeg to be seen...

JBRIII

Your last point about the effect of the program opening the raw file can't be emphasized enough. I have an eclipse image from the 2012 eclipse. Nice image, but there was/is a red band in it due to an internal reflection? It showed up in every program I tried to open the raw file with except with PixInsight. Without any effort on my part to remove the band of red, it just does not appear using PixInsight on the raw image.

PixInsight is a program sold for a astrophotigraphy and at least when I got was a one time purchase. I've never paid for an update in over ten years.

Canisdirus

Low light has more to do with sensor size...as well as pixel size and capability.

A7SIII is only 12 MP...but spread out on a FF sensor...low light monster.

Sensor size and iso is easy math...and consistent.

A FF sensor at iso 400...is iso100 on a micro sensor.
AFF sensor at iso 225...is iso 100 on a crop sensor.

You start in the hole with smaller sensors.

btbg

That's not the only reason the comparison is not valid. All of the images are shot at different exposure settings. Some at far lower ISO's than others and they have wildly varying shutter speeds. In order to be valid they should all have the same ISO and shutter speed settings as well.

Rongnongno Loc: FL

You are correct, but having this simple issue completely off is interesting.

Bill_de Loc: US

John apparently relieved more than his own boredom, otherwise he would not have gotten any response.

Good job John

---

btbg

Yes, it is. However, to me the entire experiment is interesting since it looks like it was set up to get a predetermined result.

MJPerini

I respect your curiosity, and understand the motivation to find differences between the cameras.
But there was not much that can be learned from your methodology.
The essence of a comparison is equal conditions for each and to remove as many variables as possible.
Yours was an exercise in Adding variables.
You do understand that RAW files are not an "Image Format" and thus not directly viewable, right?. You are viewing the embedded JPEG that all RAW files have, possibly changed by default parameters of your file converter.
The first step of any comparison, is understanding what you are trying to compare.

It is not my intention here to 'Pile on' negative comments, but more to say curiosity is good, but figure out how to get to some base level, so the comparison might be useful to you.
You could use this experience to design a more meaningful comparison
Good luck

JohnR Loc: The Gates of Hell

Everyone - including you, has missed the point completely! I was simply looking at the difference in noise reduction straight out of the camera between JPG & RAW for the different sensors.

Obviously people skip through the text, briefly look at the pics and then think of a criticism which they think will make them look good

CHG_CANON Loc: the Windy City

Oh well, as others have noted, you did nothing and accomplished nothing. We all tried to help you consider and accomplish something relevant. We see now that wasn't your intent, and trying to add some helpful ideas was a waste of our time and effort.

TonyP Loc: New Zealand

Yep, pretty much par for the course these days.
Just the same protagonists each time.
Good for a laugh