05-18-2016 10:03 AM - edited 05-18-2016 10:09 AM
I am fairly new to DSLRs, with just over 2 years under my belt, but not SLR photography. I am still learning all of the nuances that digital photography brings to the table, compared to my film world from 30 years ago. Let's say I took a sabbatical from the hobby, so I could raise a couple of sons. The drug store disposables were more economical than using an SLR.
Enough of that. I have recently rediscovered that my 6D has ISO expansion at the top end, like most Canon DSLRs. But, it also has an expansion in the low, "L", direction, too. The higher ISOs tend to introduce noise, and freely admit that I have never used them.
However, I recall going through the menus early on, and making a decision not to make ISO 50 available for use. I have run into several occasions where a wide open aperture [f/1.4], ISO 100, and a 1/4000 shutter speed, added up to an overexposure. I found that I either had to stop down the aperture, which upset the bokeh I wanted, or resort to using an ND filter, which frequently introduce their own set of WB and CA problems.
I do not recall the rationale behind my decision to make ISO 50 unavailable for use. Does the ISO 50 setting put extrain strain on the sensor or electronics? At the moment, I cannot think of how it could be harmful. Obviously, there is some drawback associated with it. Otherwise, why would it be made an option, like high ISO expansion, instead of simply making it available at all times?
Solved! Go to Solution.
05-18-2016 01:17 PM
@Waddizzle wrote:Why would Dynamic Range be reduced? I would not be surprised if would be. Obviously, there is some down side to it. Maybe, the sensor is being operated at such a low power, or amplification, that noise creeps in. Sort of like how driving at a very slow speed dramatically reduces your gas mileage? That's the best GUESS that I can come up with.
I guess because the image at ISO 100 is darkened to make it ISO 50, you end up losing some highlight end...Dpreview did a review on the 5DMark II and mentioned this... I'm cutting an excerpt from it below:
http://www.dpreview.com/reviews/canoneos5dmarkii/25
05-22-2016 04:35 AM - edited 05-22-2016 05:38 AM
ISO L is ISO 100 overexposed. You will get less noise (due to more photons) and less dynamic range. I have a picture showing the difference in the highlight from a 5D at ISO L and ISO 100.
To get the best out of your sensor, ETTR at ISO 100 is the key. After the exposure you make the raw file darker in your raw converter so it doesn't look too bright.
If you shoot at f/1,4 and 1/4000 and ISO 100 with your 6D you may cut highlights if you ETTR because of the hidden ISO boost at f/1,4. With lenses that the camera can't read the aperture from, for example a M42-adapter, you will never get the hidden ISO boost.
You can do the same just loosing your EF lens a little bit so that the camera and the lens can't communicate.
The attached pictures below show the hidden ISO boost at ISO 100, f/1,4 and 1/50 second from a 7D. This is just the boost, not the light loose due to wide aperture. The ISO boost will get stronger with f/1,2 or f/1. First picture shows the ISO boost, the second picture shows without ISO boost with stoped communication between lens and camera.
This is how to get the most out of the sensor without installing Magic Lantern. With Magic Lantern it is possible to get real ISO 66 or ISO 77 and that will will give more photons (less noise) and higher dynamic range (around 0,38 EV). Still in development mode.
What is true about L and ISO 100 is also true for ISO 320 (ISO 400) and ISO 500 (ISO 400). ISO 320 will get less noise but lower dynamic range. ISO 500 will get more noise and less dynamic range.
I use the knowledge above when I shoot landscapes. Then I have plenty of time to make all the settings. For weddings and portraits, I don't really have the time to fix all those settings more than that I never ever shoot at ISO 125, 250, 500, 1000 because of more noise and less dynamic range.
Will your consumers be able to notice any difference? Probably not, but this is how it works and knowledge about your camera equipment is never a bad thing to understand.
05-22-2016 09:41 AM
Peter,
The native vs normal ISO was a big thing sometime ago but I thought it was settles that it is a bogus thing. I always knew it was bogus. But others didn't. I know some folks seem to see what you try to describe but is it really there? I think no it isn't.
When the ISO is changed your camera just adjusts the signal gain. You are changing the amount of amplification applied to whatever the sensor saw. Each sensors data can be amplified in this way a certain amount before Canon says it is unusable.
Shooting with one of the extended ISOs, H or L, is the same as editing your photographs in LR then pushing the exposure ahead or back a stop or two.
This is a place where the advertising department has taken control because saying their camera can shoot fro 50 to 1000,000 sounds real impressive. The sensor isn't doing any of it. The circuitry is.
05-22-2016 01:36 PM
@ebiggs1 wrote:Peter,
The native vs normal ISO was a big thing sometime ago but I thought it was settles that it is a bogus thing. I always knew it was bogus. But others didn't. I know some folks seem to see what you try to describe but is it really there? I think no it isn't.
When the ISO is changed your camera just adjusts the signal gain. You are changing the amount of amplification applied to whatever the sensor saw. Each sensors data can be amplified in this way a certain amount before Canon says it is unusable.
Shooting with one of the extended ISOs, H or L, is the same as editing your photographs in LR then pushing the exposure ahead or back a stop or two.
This is a place where the advertising department has taken control because saying their camera can shoot fro 50 to 1000,000 sounds real impressive. The sensor isn't doing any of it. The circuitry is.
I have no opinion about Peter's claims, but I don't think I believe your conclusion. You seem to be conflating analog amplification (in camera) with digital amplification (via editing in LR) without any supporting physics to justify the equivalency. Even if we assume, for the sake of argument, that the in-camera amplification is nominally digital, there's no obvious justification for the assumption that the algorithms applied to accomplish the amplification in camera and in LR would be the same.
05-22-2016 03:53 PM - edited 05-22-2016 04:35 PM
I am not talking about native ISO. I am talking about intermediate ISO and why you shouldn´t use ISO 125, 250, 500 and 1000. Maybe you missunderstood when I wrote "real ISO 66" and Magic Lantern? Real compared to L that what I know is not allowed to be called ISO 50 by the ISO standard. If you define ISO based on the clipping point, your camera's L is actually just ISO 100. This ISO 66 is as true as ISO 100, 200 and 400 are. It allows capturing more photons without clipping, than does ISO 100. Canon L and ISO 100 clip at the same number of photons captured.
@ebiggs1 wrote:Peter,
The native vs normal ISO was a big thing sometime ago but I thought it was settles that it is a bogus thing. I always knew it was bogus. But others didn't. I know some folks seem to see what you try to describe but is it really there? I think no it isn't.
.I haven´t seen any proofs of the opposite about the intermediate ISO 125, 250, 500 and 1000. I posted a very good link, but it was erased.
@ebiggs1 wrote:
Shooting with one of the extended ISOs, H or L, is the same as editing your photographs in LR then pushing the exposure ahead or back a stop or two.
.
I wrote that. "ISO L is ISO 100 overexposed". I just didn´t write Lightroom or H.
05-23-2016 02:37 AM
Peter and Bob from Boston,
My point is, I know of the theory about the ISO's, but it doesn't exist. It has gone on for years. But my contention is, and I thought it was conceded, to be false. The result is the same. LR or in camera pushing makes no difference.
The primary reason for L and H and H1, H2, or H3 on some cameras is, the product isn't at Canon's standards but is usable if the photographers absolutely wants the shot.
Personally what I have seen seems to prove it to me. Of course this is an area where I love to delve in to. Love the discussion guys.
Perhaps a Moderator that knows Canon can shed 'light' on this?
05-23-2016 09:10 AM - edited 05-23-2016 09:12 AM
"Shooting with one of the extended ISOs, H or L, is the same as editing your photographs in LR then pushing the exposure ahead or back a stop or two. .... The sensor isn't doing any of it. The circuitry is. "
Sorry, but I simply do not see how that assertion could hardly be true. Note, this does not mean I disagree. It means your explanation to support it is pretty thin.
As I stated in my original post, I've run up against a wall where my settings are f/1.4, 1/4000, and ISO 100 add up to an overexposure. What you're saying is that if I capture the overexposed image at ISO 100 and post process, then that would be the same as shooting the originial shot at ISO 50?
I never knew that LR contained any circuitry.
05-23-2016 08:44 AM
05-23-2016 10:27 AM
Peter I am not trying to say there isn't a difference, I am trying to say, possibly not clearly, the difference is in the processing. It doesn't come from the sensor. It doesn't really matter whether it is done in camera or in post. The results are or can be the same.
After all the resulting product is what is important not how it got there. As you change the ISO, you are simply adjusting the signal gain. This amount of analog gain or amplification applied to what the sensor saw. Absolutely this is analog vs digital, as Robert pointed out, gain but in the end it results in the same product.
True I am talking the H side of the ISO range so what happens at the L side? The camera effectively cuts the shutter speed from ISO100 in half. In other words the digital gain is .5 instead of 1. It then adjusts the tone curve and BAM, you have ISO 50. Obviously twice as much light results in less noise. But other areas suffer, like highlights as they will clip much faster. This same thing can be applied in post and shot at ISO 100.
This is easy to test. I don't need to tell you or Bob from Boston how but you might want to do it. I guess what I am trying to say is all ISO numbers are just that numbers. They are values of amplification in the circuitry.
05-23-2016 10:27 AM
05-23-2016 10:49 AM
12/18/2024: New firmware updates are available.
EOS C300 Mark III - Version 1..0.9.1
EOS C500 Mark II - Version 1.1.3.1
12/05/2024: New firmware updates are available.
EOS R5 Mark II - Version 1.0.2
09/26/2024: New firmware updates are available.
EOS R6 Mark II - Version 1.5.0
Canon U.S.A Inc. All Rights Reserved. Reproduction in whole or part without permission is prohibited.