Post your 'before' and 'after' pictures

[Rick M]

What do you mean with amount of data recorded?

You record less data with the under exposed shot. I just confirmed this with My V2. I shot 2 identical shots in RAW, one exposed as normal (requiring ISO 400) and one at native ISO of 160 and increased shutter speed to under expose. The properly exposed image was 13.6 MB, the under exposed was 11.0 MB. This results in less data to work with. I was still able to pull out an acceptable image, but I want the best quality a camera can produce.

This is a good tool under circumstances where you need more speed and can still get the shot; Although at potentially less quality.

 

[Blade Canyon]

Okay, I understand your point, J-see. You say that, shutter speed and aperture being the same, it doesn't matter what ISO you shoot because for higher ISOs the camera is only doing a software adjustment, and you can do that same thing in post. In fact, your computer processor and software are probably more powerful than whatever's built into the camera anyway.

I can possibly believe this, but have never heard it before. I will experiment with the idea tonight.

That said, this book "The Digital SLR Expert: Landscapes" by Tom Mackie et al. says that the reason you want a good histogram in your original shot is because 75% of your sensor is used for the highest two stops, and doing it your way would not do that. I'll just let them explain it in the attached scan:

 

[Blacktop]

It doesn't affect detail. That depends on how close I can get and how much light there is for my shutter/aperture.

View attachment 134724

Let me jump in here.
You shot this duck at 11:34 in the morning, which means plenty of light, even on a cloudy day.
The duck is not flying, but rather leisurely swimming about, so you don't need 1/1000th of a second shutter speed. even with a shutter speed of 1/250th which is 2 stops lower than 1/1000th you would be fine shooting this duck handheld at ISO 100.
Even if you're not, why not bump up the ISO to 200 and get yourself at 1/500th of a second shutter speed.

I can understand if you were shooting a D200 or a D70 that has no high ISO capability whatsoever, but OMG!!

I'm not trying to knock down your method, because quite frankly I don't understand it, and you're the one doing all the experimenting, I'm just somewhat baffled is all.

 

[J-see]

You record less data with the under exposed shot. I just confirmed this with My V2. I shot 2 identical shots in RAW, one exposed as normal (requiring ISO 400) and one at native ISO of 160 and increased shutter speed to under expose. The properly exposed image was 13.6 MB, the under exposed was 11.0 MB. This results in less data to work with. I was still able to pull out an acceptable image, but I want the best quality a camera can produce.

This is a good tool under circumstances where you need more speed and can still get the shot; Although at potentially less quality.

The data size doesn't matter Rick, if you used the exact same settings, the initial data the higher ISO worked with is identical to the data you shoot at native ISO. The only difference is the light the sensor received not yet being adjusted to our liking.

 

[J-see]

Let me jump in here.
You shot this duck at 11:34 in the morning, which means plenty of light, even on a cloudy day.
The duck is not flying, but rather leisurely swimming about, so you don't need 1/1000th of a second shutter speed. even with a shutter speed of 1/250th which is 2 stops lower than 1/1000th you would be fine shooting this duck handheld at ISO 100.
Even if you're not, why not bump up the ISO to 200 and get yourself at 1/500th of a second shutter speed.

I can understand if you were shooting a D200 or a D70 that has no high ISO capability whatsoever, but OMG!!

I'm not trying to knock down your method, because quite frankly I don't understand it, and you're the one doing all the experimenting, I'm just somewhat baffled is all.

My shutter speed was a result of me being lazy. I could have shot it much slower but it didn't matter much. The more extreme my test shots, the faster I would encounter issues if there were.

 

[Rick M]

The data size doesn't matter Rick, if you used the exact same settings, the initial data the higher ISO worked with is identical to the data you shoot at native ISO. The only difference is the light the sensor received not yet being adjusted to our liking.

I guess I prefer more data to work with than less. The difference was 13.6 MB vs. 11.0 MB. To each his own, but to me more is better !

 

[wud]

How will you be able to see if you nailed focus, if all your images are so dark on the screen?

I like getting a shot as right as possible in camera. Only way I can correct an error.

 

[rangioran]

I'd just like to jump in here to note that, of all the forums on 'Nikonites' this one has captured my attention and interest more than any other - thanks!

 

[Blacktop]

My shutter speed was a result of me being lazy. I could have shot it much slower but it didn't matter much. The more extreme my test shots, the faster I would encounter issues if there were.

I think it does matter. You can clearly see how much noise you're introducing to this poor duck by pulling up all this blackness. With that D750 set at ISO 200 or 400 you would have a properly exposed shot with no noise whatsoever.
I am clearly not understanding your method of madness.

 

[J-see]

How will you be able to see if you nailed focus, if all your images are so dark on the screen?

I like getting a shot as right as possible in camera. Only way I can correct an error.

I see my shot in the viewfinder exactly the same way anyone does. It doesn't change that. Whether I see it in my shot before processed is not important since by then it is too late anyways. Regardless what settings I use, I close to never check my shots.

 

[hark]

I am old school. Nail the shot in camera as much as possible. That's my motto.

 

[Blade Canyon]

Well, J-see, my own quickie experiments don't prove you dead wrong, but they don't prove you right. Here are two shots, from a tripod, shot on Manual at 1/50th of a second, F5. One was shot at ISO 100 (and thus very underexposed in every way that we usually mean it), and the other was shot at ISO 1600 which the camera read as properly exposed. In ACR, I increased the "exposure" on the ASA 100 by exactly four stops.

The color was not at all the same as the normal looking ISO 1600 shot. It was very yellow. So I adjusted that and cropped in on both of them. The "underexposed" shot looks noisier than the properly exposed shot. Here are extreme crops of both. Other than the color and slightly more noise, however, I do admit they are fairly similar. What do you think?

 

[J-see]

I think it does matter. You can clearly see how much noise you're introducing to this poor duck by pulling up all this blackness. With that D750 set at ISO 200 or 400 you would have a properly exposed shot with no noise whatsoever.
I am clearly not understanding your method of madness.

I don't know how to explain this since it seems so simple to understand but I apparently can't explain it as simple.

Let's agree that at a certain shutter and aperture, we capture a certain amount of light. We probably won't disagree about that. Regardless what ISO I shoot, the amount of light hitting the sensor is defined by A and S. In manual mode, ISO does not close or open the lens or affect the shutter. We can agree about that too. So whatever the DR range of the scene, whatever the brightness and shadows, ISO does not change those before they hit the sensor.

I hope we agree this far.

What ISO does is change the analog data of the sensor into an exposure increase we desire. But whether you increase ISO or not, the analog data is identical.

What I do is not use ISO and change the analog data (after being digitalized) towards the exposure I prefer. Higher ISO does the exact same thing but during the conversion. But the data it starts with is still the same. It doesn't add light to the light captured.

So whatever I do, I have the same data it all started with; the amount of light defined by A and S. I use it unmanipulated, high ISO provides it to me manipulated.

Here are two shots I've show before with the histograms. The one at high ISO, the other native. I manipulated both in post until they have the exact same exposure and set black and white to the maximum possible so they don't clip. Their file sizes differ but check the histograms.





Post:

 

[J-see]

Well, J-see, my own quickie experiments don't prove you dead wrong, but they don't prove you right. Here are two shots, from a tripod, shot on Manual at 1/50th of a second, F5. One was shot at ISO 100 (and thus very underexposed in every way that we usually mean it), and the other was shot at ISO 1600 which the camera read as properly exposed. In ACR, I increased the "exposure" on the ASA 100 by exactly four stops.

The color was not at all the same as the normal looking ISO 1600 shot. It was very yellow. So I adjusted that and cropped in on both of them. The "underexposed" shot looks noisier than the properly exposed shot. Here are extreme crops of both. Other than the color and slightly more noise, however, I do admit they are fairly similar. What do you think?

It's pretty hard to see in this shot since there's a rather uniform brightness. The moment the DR range is wider than the cam, the difference between both is that low ISO will pay a price in the shadows while high ISO pays one in the brightness levels.

Btw, WB is affected by ISO but not always. I don't know why it happens but I need to adjust them differently.

 

[wud]

Changing ISO 100 to 6400 (in this last post) doesn't prove anything. No one would use ISO 6400 for the last shot, if looking at the camera screen it would have been pretty clear it was a wrong setting.

I agree the cameras are amazing at correcting errors, but not every bad shot can be saved in pp.

 

[Rick M]

The data size doesn't matter Rick, if you used the exact same settings, the initial data the higher ISO worked with is identical to the data you shoot at native ISO. The only difference is the light the sensor received not yet being adjusted to our liking.

When you crank up the shutter speed to underexpose, you limit the amount of light to the sensor, hence limiting the data recorded. Less data, less quality, it's that simple.

 

[J-see]

Good for J-See now STFU instead of posting this and that post your good stuff

You throw alotta crap out there if you like BLACK pics take them we don't

I'm answering replies. If you don't like that, there's a simple solution; don't read them.

 

[J-see]

Changing ISO 100 to 6400 (in this last post) doesn't prove anything. No one would use ISO 6400 for the last shot, if looking at the camera screen it would have been pretty clear it was a wrong setting.

I agree the cameras are amazing at correcting errors, but not every bad shot can be saved in pp.

6400 is not an unusual setting when trying to grab stars during light pollution. It's actually after reading articles about it, I myself used it high ISO too. Until I noticed 100 does an as good job.

 

[Rick M]

I think your theory needs parameters to be successful. Native ISO in many circumstances will force a higher shutter speed to get a sharp image, which in turn may result in data loss from the underexposed image.

 

[J-see]

When you crank up the shutter speed to underexpose, you limit the amount of light to the sensor, hence limiting the data recorded. Less data, less quality, it's that simple.

No disagreement there but my birds in flight limit my options. I need f/6.3 or f/7.1 as aperture with the Tam and 1/1000s for shutter. My technique with the long lens still requires a lot of practice so high shutter is my only option. In the past I upped the ISO with those settings. During this season, light is too low here. But now I discovered I can do the same without the ISO change.