High ISO Shooting Tips - Mechanical Shutter Reduces Noise (Robin Wong)

wjiang

Mu-43 Legend
Joined
Sep 7, 2013
Messages
7,257
Location
Christchurch, New Zealand
That's a funny way of looking at it - isn't it more like choosing to use electronic shutter increases noise? That's what I first thought when I tried electronic shutter when they added the feature to my old E-M1. In low light, it's typically artificial lighting, so there's also banding to contend with if electronic shutter is used.
 

RR Jonny

Mu-43 Veteran
Joined
Jan 13, 2017
Messages
247
Location
China or the U.S. - one of the two
The main reason I chose the EM5 MKII as my starter camera was because I knew I would be taking a ton of bad shots and didn't want to wear the shutter mechanism out. I have developed the habit of using the electronic shutter in almost all situations except for high speed shutter applications like birds in flight. I never expected the mechanical shutter to be better with noise. I have never even thought of it or tried it. I am looking forward to trying it because I prefer low light conditions for static birds that have some lightness to their plumage. I still have a lot to learn.
 

mfturner

Mu-43 Rookie
Joined
Aug 6, 2019
Messages
12
Nice video, thanks for the link. I like the Olympus specific tips, like the discussions of gradation and hand held starlight mode. I haven't experimented with electronic shutter and high iso, that is worth knowing.
 

Mack

Mu-43 Top Veteran
Joined
Jan 14, 2018
Messages
972
Odd they left out the starlight mode in the E1 MK II and E-M1X. No mention about the Pen-F if it has it.
 

mfturner

Mu-43 Rookie
Joined
Aug 6, 2019
Messages
12
I'm playing with it now, I may find a use for it, but on my m10.3 it uses mechanical shutter and changes the "keep warm color" to on, which I'm unsure about given the jpeg output. So I'll have to think about it. But it's nice to know that image stacking is what it does, I can see using it if I'm in a hurry and it's very dark.
 

archaeopteryx

Gambian sidling bush
Joined
Feb 25, 2017
Messages
1,229
For anyone who doesn't feel like devoting more minutes of their life to yet another video which would be more efficiently presented as an article (or at least abstracted by an OP), this thread is presumably intended to be about the first of the five tips presented:
  1. 1:32 Don't use silent shutter
  2. 2:53 Hand-held starlight mode
  3. 4:39 Do not underexpose
  4. 6:49 Shoot raw
  5. 8:43 Gradation normal
In a nutshell, this reduces to the remark at 1:47: "I don't really know what causes the silent shutter to produce more noise." The specific objections made are electronic shutter produces "way more" noise, grain, and artifacts plus a green cast in shadows. The differences I see in the two examples presented, at 2:32 and 2:51 (both ISO 6400), appear primarily attributable to different object positions and correspondingly different lighting between the exposures compared. There's no indication this was a controlled test so, potentially, any adverse selection on electronic shutter could be a sampling artifact from small n (in some ways the electronic shutter images in the comparisons appear subjectively preferable to me, though). The pixels I measured RGB tuples on show no difference in green values, which suggests any cast may be too small to pass through video encoding (a difficulty publishing stills in an article would avoid, but I digress).

From the limited data available in the video, it seems one plausible interpretation would be "way more" amounts to "no significant difference" for publication via Youtube.

isn't it more like choosing to use electronic shutter increases noise?
Depends on the noise mechanism and how sloppily terminology is being used---I think we covered much of that a year and a half ago. If you still have comparative images from the E-M1 perhaps we could work from those (I have only a Panasonic body, which isn't a particularly relevant data source).
 

b_rubenstein

Mu-43 All-Pro
Joined
Mar 20, 2012
Messages
1,434
Location
Melbourne, FL
Since Robin is an engineer, one would think that he would do some research into how to test for noise. But no, he's just another social media influencer, with lots of followers and a tenuous grasp of the underlying science and technology of digital photography.
 

PakkyT

Mu-43 Hall of Famer
Joined
Jun 20, 2015
Messages
2,771
Location
New England
My problem with these types of examples/demonstrations is that they don't use realistic evaluations of the shooting situation (shooting at night a tiny plastic figure off to one side with a blurred background of a city in most of the frame can be a cool looking shot when you look at all of it at once, but...) and try to use a 100% crop of a tiny section of the whole screen to show more or less noise, color casts, etc. when I would never think to zoom in on a section of such a photo. Why would I care about the noise close up of a tiny element of a photo that is not important to the shot by itself but is only as a part of a larger scene?

Also, often when someone does this and tries to show a side by side comparison, I am often not convinced one is clearly better than the other. I can often see they are different, but being different doesn't necessarily mean better. If I stare at them long enough I can convince myself which ever one I want to be better will start to look better.
 

Stanga

Mu-43 Top Veteran
Joined
Oct 16, 2016
Messages
971
I am trying to get my head round the conclusion that mechanical shutter reduces noise. Is it not more accurate to say that electronic shutter produces more noise? There are two types of mechanical shutters. Both operate on a different principle. So a general statement that a mechanical shutter reduces noise is in fact a throw away comment with zero credibility.
 

archaeopteryx

Gambian sidling bush
Joined
Feb 25, 2017
Messages
1,229
Is it not more accurate to say that electronic shutter produces more noise?
It's probably most accurate to say mu since the data gets read off the sensor using the same circuitry regardless of the shutter type used. I'm aware of two differences between the shutter types:
  1. With electronic shutter, the sensor remains illuminated during read. Exactly what happens depends on the structure of the sensor and the implementation of the read operation but the basic thing is pixels can continue to charge during read. This implies shifts in read, shot, and fixed pattern noise contributions overall noise may occur. I'm not aware of any evidence indicating these changes are significant but, in principle, I can't entirely exclude the possibility visible differences might occur in special cases.
  2. With electronic shutter, the ADC bit depth may be reduced to obtain a faster sampling rate and correspondingly faster curtain times. At low ISOs this truncation can reduce dynamic range but 1) it's not the electronic shutter's fault, 2) it's formally noise either, and 3) the Sony a7S is the only camera I know where it might be an issue (Kasson 2014). I am, however, aware of several cases where photography forum participants have assumed degradation a priori and referred to it as noise. ;)
This apparent lack of noise disadvantage is consistent with use of electronic shutter in aerial imaging (Lumenera). I also don't recall any finding of difference in Jim Kasson's series on the a6300. None of this precludes greater noise with electronic shutter---Jim documented it on the a7S---but it does indicate a noisier electronic shutter might be best interpreted as a design flaw and isn't an intrinsic disadvantage.

one would think that he would do some research into how to test for noise
As I understand it, the research was done mostly in the 1960s through 1980s and is collected in Janesick 2007. There's a DPR thread where Bill Claff indicates photon transfer curves from his NefUtil tool follow the methodology in Janesick. For those disinclined to purchase the book, there's a summary of transfer curves and noise mechanisms in Crisp 2009.

One way to investigate would be to give NefUtil two sets of well controlled test images, one from mechanical shutter and one from electronic shutter, and see if meaningful changes occur in σread, Kadc, or Pn. I'm not aware of a case where this specific assessment method has been used, though.
 

Latest posts

Links on this page may be to our affiliates. Sales through affiliate links may benefit this site.
Mu-43 is a fan site and not associated with Olympus, Panasonic, or other manufacturers mentioned on this site.
Forum post reactions by Twemoji: https://github.com/twitter/twemoji
Copyright © 2009-2019 Amin Forums, LLC
Top Bottom