Approve the Cookies
This website uses cookies to improve your user experience. By using this site, you agree to our use of cookies and our Privacy Policy.
OK
Index  •   • New posts  •   • RTAT  •   • 'Best of'  •   • Gallery  •   • Gear  •   • Reviews
Guest
New posts  •   • RTAT  •   • 'Best of'  •   • Gallery  •   • Gear  •   • Reviews
Register to forums    Log in

 
FORUMS Photography Talk by Genre General Photography Talk 
Thread started 07 Jul 2014 (Monday) 02:54
Search threadPrev/next
sponsored links
(this ad will go away when you log in as a registered member)

ETTR - Equal noise at 3200 than at 1600?

 
sploo
premature adulation
2,414 posts
Gallery: 1 photo
Likes: 464
Joined Nov 2011
Location: West Yorkshire, UK
     
Jul 13, 2014 09:25 |  #31

BigAl007 wrote in post #17027627 (external link)
Not owning a 5DIII I had to download a manual to check this. I would expect that you would potentially see gains to ISO 12800. The ISO values above that though are designated as extended (H) values and would appear to be digitally generated so may as well be done by pushing in camera. At least that is the way that Canon seem to have designated ISO values in other older bodies. The fact that it is digital gain is usually accompanied by a resulting loss of DR. This is quite easy to see when using intermediate ISO values too, which are also pulled or pushed from the closest full value.

The testing that I'd done to compare ISO levels was subjective (i.e. I've not done any mathematical analysis of the results) but from reading up on results from other bodies I think that the analogue gains usually seem to drop off before you get to the extended ISO values.

As you rightly point out, intermediate ISO values are often the result of manipulation from other amplifier settings.

BigAl007 wrote in post #17027627 (external link)
Of course even if the amplification used is analogue there will come a time when using it is not necessarily going to gain you much in the way of additional SNR, but I would always be inclined to keep going with analogue gain, unless it actually made the SNR worse at any given combination of shutter speed and aperture. Also of course you still have the benefit of shooting at the highest ISO value at an ordinary level before you have to start pushing it.

Agreed; though of course remember that by using a higher ISO level you are risking more clipping in the highlights. So though you may not be getting worse shadow detail it may still be better to stay with a lower ISO level in some situations.

brettjrob wrote in post #17028131 (external link)
For cameras with Sony sensors, and likely many other non-Canon sensors, ETTR really is almost pointless above the base ISO. Yes, pushing the histogram as far to the right as you can without clipping highlights is desirable -- but reducing the ISO is more desirable, if you have the choice.

One other clarification: I'm interpreting ETTR to mean that you're purposely overexposing the scene, relative to how you will ultimately present it. In other words, that you plan to reduce the exposure in software after the fact.

Agreed on your clarification. Note though that there is a reason for using ETTR even when using a Sony sensor; darker pixels will be digitised into smaller numbers, with lighter pixels resulting in larger values.

Imagine you have an image with a sky that darkens by one stop in a smooth gradient from the top to the bottom. If your exposure settings will result in this sky being quite dark in the image on the sensor the pixels will have small values; e.g. 64 at the top and 32 at the bottom (the bottom being roughly half the signal strength, thus one stop darker).

If you instead exposed the scene so that the sky was very bright on the sensor the values may be, say 16384, and 8192 respectively.

When you come to edit this image in post, any attempts to increase the contrast in the sky (essentially increasing the difference between the brightness of the top and the bottom) might result in obvious banding and posterisation because there are only 32 (64-32) steps between the top and the bottom of the sky in the original image.

Doing the same with the brighter exposure gives you far more latitude because there is significantly more tonal gradation.

Obviously that is an extreme example, but even ignoring photon noise and Canon's Shadow Banding(tm) feature, ETTR is very beneficial.

EDIT: An old article, but hopefully useful: http://www.luminous-landscape.com/tutorial​s/expose-right.shtml (external link)


Camera, some lenses, too little time, too little talent

  
  LOG IN TO REPLY
sponsored links
(this ad will go away when you log in as a registered member)
tonylong
...winded
Avatar
54,657 posts
Gallery: 60 photos
Likes: 544
Joined Sep 2007
Location: Vancouver, WA USA
     
Jul 13, 2014 17:37 |  #32

Wow, holy cow!

I'm not a digital image scientist, so please note that I'm chiming in as a hobbyist, for fun, and because it's interesting!!

So, as I've seen things, there are a couple aspects we are looking at here, and we, especially the "newbies", should be clear about things:

1) There is a clear advantage to "ETTR" in the sense that if we expose "To The Right" to the extent that our shutter speed and aperture allow and if we are shooting Raw we are not blowing out highlights, we can fully recover all our image data, and since the lower light areas will capture less photons and so we benefit from more photons being captured, well, ETTR is a win-win, which in fact differs from the film days, but that's a different discussion...

2) And then, there is the ISO discussion, but that differs. It's not about the "science" of digital imagery, but it's about the specifics of how technology handles digital imaging!

For example, we've seen that at higher ISOs (needed in lower light scenarios) Canon cameras do deliver images that are "cleaner" when regarding digital "noise". I don't know the science around this, but I have seen this...but then, I've also seen that bodies from Nikon and also Sony have done better in delivering low ISO noise levels!

So, in effect, it may be a "toss-up", as to whether you choose to "ETTR" and include the Canon-biased "HAMSTTR" approach or not, or you just want to stick with the digital aspects of ETTR, or to do like I typically do, which is to seek to "get it right" (I know, that's a silly term) in the camera, but I do shoot Raw and I do make the best use of my Raw software to optimize my photos!!

Well, that's just me, blabbering as usual...and as others have observed, well...responding??


Tony
Two Canon cameras (5DC, 30D), three Canon lenses (24-105, 100-400, 100mm macro)
Tony Long Photos on PBase (external link)
Wildlife project pics here (external link), Biking Photog shoots here (external link), "Suburbia" project here (external link)! Mount St. Helens, Mount Hood pics here (external link)

  
  LOG IN TO REPLY
sploo
premature adulation
2,414 posts
Gallery: 1 photo
Likes: 464
Joined Nov 2011
Location: West Yorkshire, UK
     
Jul 14, 2014 05:07 |  #33

tonylong wrote in post #17029631 (external link)
1) There is a clear advantage to "ETTR"...

Yep. That's spot on.

tonylong wrote in post #17029631 (external link)
2) And then, there is the ISO discussion, but that differs. It's not about the "science" of digital imagery, but it's about the specifics of how technology handles digital imaging!

For example, we've seen that at higher ISOs (needed in lower light scenarios) Canon cameras do deliver images that are "cleaner" when regarding digital "noise". I don't know the science around this, but I have seen this...but then, I've also seen that bodies from Nikon and also Sony have done better in delivering low ISO noise levels!

As far as I understand it:

Tonal gradation benefits of ETTR are pretty universal. That is, any imaging sensor that detects light and then digitises it, such that darker pixels will result in smaller values and brighter pixels result in larger values, should get some benefit if you can set your exposure such that the image histogram is pushed towards the right. In some circumstances the difference may not be visible (or feasible, due to having a scene with a large dynamic range).

The issue with Canon sensors is that the read-out electronics unfortunately adds more noise to the image. Worse, this noise takes on the appearance of vertical bands, which makes it particularly obvious and objectionable.

For any pixels that are quite bright on the sensor, the added noise will be negligible and unlikely to be visible. For the darker pixels the added noise could be quite large in comparison to the signal for the pixel, and therefore detail in the darker stops in the image are overwhelmed by this pattern noise. E.g. 30 units of random noise on a pixel with a reading of 8000 isn't a problem. 30 units of random noise on a pixel with a reading of 25 means that pixel is essentially junk.

However, the ISO setting on your camera is effectively an amplifier that multiplies the signal on the sensor before it is read out. At higher ISO levels (where you expect the readings on the sensor to be low because your scene is dark) the amplifier has increased the signal in the pixels, thus making them larger before you go to the read-out stage. E.g. amplify our pixel with a signal of ~25 by 16 times (ISO 100->1600) and you get ~400. 30 units of noise on 400 is far less damaging.

Note that in any individual pixel there will also be an element of noise as well as signal (this is true of any sensor, and can be from any stray energy such as heat). When you amplify the signal in a pixel with a high ISO setting you are of course also amplifying the noise - because there is no way to know what portion of the signal for that pixel came from light through the lens or from other sources. This means that a dark scene amplified up with a high ISO setting will naturally have more noise, and less dynamic range because for the darkest areas of the image the pixels will have a poor signal-to-noise ratio.

With a Canon sensor, the noise inherent with a dark scene+high ISO setting is enough to mask the pattern noise that gets added in the read-out stage.

If you look at sensor tests you'll see that the dynamic range of a Sony sensor drops as you increase the ISO - because of the greater pixel noise. However, Canon sensors show a fairly consistent (lower) dynamic range reading for the lowest ISO stops, because although the pixel noise isn't large the detail in the darker areas of the image is being damaged by the pattern noise.

Once you get into the higher ISO settings the pattern noise is masked by the pixel noise (because it becomes small in comparison), and for those settings the Canon sensors are as good as (and in some cases slightly better than) rival sensors. From what I understand this indicates that the sensor itself is fundamentally good, and if Canon could improve their read-out electronics (such that it's not adding pattern noise) they'd be competitive with other sensors.

An ideal workaround to this would be to be able to change the ISO setting on a per pixel basis, such that the darker pixels get amplified up so they overcome the pattern noise issue, but you don't use a higher ISO setting on brighter pixels as this would cause them to become too large and clip. The MagicLantern DualISO feature does something a bit like this; in the sense that it reads out every other line from the sensor at one of two different ISO settings. The pixel data from lines that are read out with a lower ISO (e.g. 100) are used for the medium and brighter areas of the image, and the lines read at the higher ISO (e.g. 800) are used for the shadow detail. The data gets combined in post such that I believe the shadow detail is shifted back down so you effectively end up with an image as if it were taken at the base ISO, but with cleaner shadows because having read them at a higher ISO the pattern noise is less evident.

So in summary, lower ISO values should always get you a cleaner image, but you don't get the advantage you should when using a Canon sensor at a lower ISO setting.


Camera, some lenses, too little time, too little talent

  
  LOG IN TO REPLY
tzalman
Fatal attraction.
Avatar
13,476 posts
Likes: 197
Joined Apr 2005
Location: Gesher Haziv, Israel
     
Jul 14, 2014 08:01 |  #34

I agree entirely with what sploo has written, but I do take issue with several comments by bettjrob, perhaps written because he is not familiar with Canon cameras.

But if overexposing requires you to bump up the ISO, the extra noise that comes with that will more than offset the advantage of ETTR. That's all I was trying to say. Bumping up the ISO for the sole purpose of ETTR isn't really worthwhile. To minimize noise, first and foremost use the lowest ISO possible.

Look at this chart of analog ISO noise levels in the 5D2 (measured by John Sheehy).

IMAGE: http://photos.imageevent.com/elied/linked/5D2%20noise.jpg

Clearly the increase in noise is not linear. The increase with a 2X amplification from ISO 100 to ISO 200 is very small. Going from ISO 1600 to ISO 3200 the noise is almost doubled. That is why the 5D2's analog amplifier only goes up to a 32X gain. Beyond that there is no advantage to using costly analog amplification and further increases in ISO are done digitally (virtually free). Digital amplification always doubles the noise with a doubling of the ISO.

However, the converse is also true - digital reduction of tonality (brightness) reduces visual noise. If shooting at a metered f/2, 1/60 at ISO 200 leaves the high end of the histogram two stops below clipping and doing ETTR by dropping the shutter speed to 1/15 is out of the question, increasing ISO from 200 to 800 will cause a 1.25X increase in noise. Subsequently digitally pulling the tonality down will produce a reduction in visual noise that more than compensates for the 1.25 increase.

One other clarification: I'm interpreting ETTR to mean that you're purposely overexposing the scene, relative to how you will ultimately present it. In other words, that you plan to reduce the exposure in software after the fact.

This is a fundamentally faulty understanding of the term ETTR. All too often ETTR is equated with "overexposing" relative to the traditional metering method inherited from film days, i.e. metering to place medium grey. Even more improperly, many people think that it means setting your camera for an automatic bit of positive EC in every shot. Properly defined ETTR means placing the brightest significant highlight (the one in which you want to retain detail) just short of clipping. Whether this is overexposure or not depends on the subject and its contrast range. In the language of Ansel Adams it is placing the brightest textured highlight in Zone VIII instead of metering for Zone V.
In a Raw capture middle grey is around three stops below clipping. If the tonal range of the subject has only two stops from medium grey to highlight, ETTR says to increase exposure by one stop above a spot meter reading of the medium grey object. If there are three stops difference from medium to highlight, a meter based exposure will already be ETTR without any change needed. And if the differential is greater then that (say, a shaded building against a sky full of bright white clouds), ETTR would dictate reducing exposure, provided that those clouds are really important enough to you that you are willing to obtain a less than optimum exposure of the building.


Elie / אלי

  
  LOG IN TO REPLY
BigAl007
Cream of the Crop
Avatar
8,011 posts
Gallery: 543 photos
Best ofs: 1
Likes: 1617
Joined Dec 2010
Location: Repps cum Bastwick, Gt Yarmouth, Norfolk, UK.
     
Jul 14, 2014 08:01 |  #35

Of course Canon's big problem with the read out noise may be that they are impeded by someone else's patents. It can be very difficult to find a way around limitations placed by patents, as it can place blocks on whole areas of innovation that Canon could potentially use to improve their systems.

Alan


My Flickr (external link)
My new Aviation images blog site (external link)

  
  LOG IN TO REPLY
sploo
premature adulation
2,414 posts
Gallery: 1 photo
Likes: 464
Joined Nov 2011
Location: West Yorkshire, UK
     
Jul 14, 2014 08:58 |  #36

tzalman wrote in post #17030699 (external link)
Properly defined ETTR means placing the brightest significant highlight (the one in which you want to retain detail) just short of clipping. Whether this is overexposure or not depends on the subject and its contrast range.

Yes, that's a much better - and succinct - description of ETTR. I guess the end result may often be a bit of mild overexposure (vs. what you want for the appearance of the final shot), but I agree that it doesn't guarantee to mean manual overexposure with regard to what the camera meters.

The daft thing is that with modern DSLRs we have great manual + auto ISO + minimum shutter speed controls, so it shouldn't be that hard for camera firmware developers to come up with an algorithm that would choose the best ISO and exposure settings (for the body being used) to give a shot with the photographer's chosen aperture and acceptable shutter speed range. Metadata in the resulting image could then be used to ensure the image appears roughly as it would have if traditionally metered when initially processed by a raw converter in post - but with potentially increased latitude for manipulating the file.

Done well, ETTR could be mostly hidden from the shooter, allowing him/her to concentrate on the artistic elements. Of course, then we'd have less stuff to talk about on forums ;)


Camera, some lenses, too little time, too little talent

  
  LOG IN TO REPLY
sploo
premature adulation
2,414 posts
Gallery: 1 photo
Likes: 464
Joined Nov 2011
Location: West Yorkshire, UK
     
Jul 14, 2014 09:00 |  #37

BigAl007 wrote in post #17030700 (external link)
Of course Canon's big problem with the read out noise may be that they are impeded by someone else's patents. It can be very difficult to find a way around limitations placed by patents, as it can place blocks on whole areas of innovation that Canon could potentially use to improve their systems.

Alan

Absolutely. That, or they've focussed resources on their dual pixel phase AF, video, and the high ISO sports/reportage market. In any event, they don't seem to be able/willing to fix the problem.


Camera, some lenses, too little time, too little talent

  
  LOG IN TO REPLY
edge100
Goldmember
1,920 posts
Likes: 16
Joined Jan 2010
Location: Toronto, Canada
     
Jul 15, 2014 14:31 |  #38

http://theory.uchicago​.edu …/noise/noise-p3.html#ETTR (external link)


Street and editorial photography in Toronto, Canada (external link)
Mirrorless: Fujifilm X-Pro1
Film: Leica MP | Leica M2 | CV Nokton 35/1.4 | CV Nokton 40 f/1.4 | Leitz Summitar 50 f/2 | Canon 50 f/1.2 LTM | Mamiya 7 | Mamiya 80 f/4.0 | Mamiya 150 f/4.5 | Mamiya 43 f/4.5
How to get good colour from C-41 film scans (external link)

Digitizing film with a digital camera (external link)

  
  LOG IN TO REPLY
sploo
premature adulation
2,414 posts
Gallery: 1 photo
Likes: 464
Joined Nov 2011
Location: West Yorkshire, UK
     
Jul 15, 2014 17:02 |  #39

Good article. It does appear to be from some time back, so I don't know if the modern DSLRs are producing low enough noise levels to justify 14-bit data files.

It's interesting that he's chosen to compare ISO 1600 vs. 3200 on the 1D3, as looking at the graph tzalman posted (for the 5D2) that also hints to there being little advantage in making the jump (but that's not true of the lower ISO levels).

The passage "going from ISO 100 to ISO 200 yields a big improvement at the lower end of exposure (at the cost of some latitude at the upper end); on the other hand, going from ISO 800 to 1600 doesn't make much difference at all in shadow S/N, and in addition one loses an entire stop of raw headroom" matches my earlier comments about the trade-off of trying to improve your shadow detail vs. clipping highlights (and when using a higher ISO setting doesn't improve shadow detail but merely risks your highlights).


Camera, some lenses, too little time, too little talent

  
  LOG IN TO REPLY
sponsored links
(this ad will go away when you log in as a registered member)

3,607 views & 0 likes for this thread
ETTR - Equal noise at 3200 than at 1600?
FORUMS Photography Talk by Genre General Photography Talk 
AAA
x 1600
y 1600

Jump to forum...   •  Rules   •  Index   •  New posts   •  RTAT   •  'Best of'   •  Gallery   •  Gear   •  Reviews   •  Member list   •  Polls   •  Image rules   •  Search   •  Password reset

Not a member yet?
Register to forums
Registered members may log in to forums and access all the features: full search, image upload, follow forums, own gear list and ratings, likes, more forums, private messaging, thread follow, notifications, own gallery, all settings, view hosted photos, own reviews, see more and do more... and all is free. Don't be a stranger - register now and start posting!


COOKIES DISCLAIMER: This website uses cookies to improve your user experience. By using this site, you agree to our use of cookies and to our privacy policy.
Privacy policy and cookie usage info.


POWERED BY AMASS forum software 2.1forum software
version 2.1 /
code and design
by Pekka Saarinen ©
for photography-on-the.net

Latest registered member is kbcl0ttery
848 guests, 234 members online
Simultaneous users record so far is 15144, that happened on Nov 22, 2018

Photography-on-the.net Digital Photography Forums is the website for photographers and all who love great photos, camera and post processing techniques, gear talk, discussion and sharing. Professionals, hobbyists, newbies and those who don't even own a camera -- all are welcome regardless of skill, favourite brand, gear, gender or age. Registering and usage is free.