Canon 40D and ISO 3200

Exactly frenchfx.. When I worked for a newspaper during both film and digital days if we got back to the office without the shots and complained we ran out of light, the pictorial editor would throw the nearest, heaviest object at us.. We all knew at times the only solution was to bump the ISO up to get the shot..

We don't live in a perfect world where things are easy,, we have to adapt and use what is planted between our ears!!

Ehhh wrote in post #6313612
I do however doubt that 1600 is worthless in the sense that it is worse than "software emulating" 800 to 1600. Do you have any examples to back that claim up?

I measured the noise. With ISO 1600 is a bit lower than with an underexposed 800. In cases, when highlight clipping is not a real problem, even this bit may be useful. However, my statement related to the example, which contains lots of clipped areas. One needs to carefully evaluate in such situation if this small gain in the shadows is worth of the full stop loss in the highlights.

frenchfx wrote in post #6313792
These were shot at ISO1600 1\200 F2.0: anything less wouldn't work

This is all right when recording JPEG in camera, but it is not so with raw. The intensity could have been increased in raw conversion. The white/red uniforms would look even better than now.

so what you are getting at is
1) that when highlights are critical, use a lower ISO, underexpose, and bring it up in post processing?
2) and when shadows are critical and highlights aren't so critical, use a higher ISO (since you said that 1600 had less noise than underexposed 800 brought up a stop)?

what about highlight tone priority, how does that figure into the equation?

also, i thought that 1600 was a real ISO? hence it not being a letter, such as H (H is 3200 equiv on 40D)?

_aravena wrote in post #6302347
ISO 3200 and some extreme edits

in such a case as his, wouldn't shooting a lower ISO and bringing it up cause more noise in the shadows than is present right now?

this is pretty enlightening, no pun intended

Sfordphoto wrote in post #6313993
in such a case as his, wouldn't shooting a lower ISO and bringing it up cause more noise in the shadows than is present right now?

I have not suggested to shoot it with ISO 100 but with 800. Perhaps 1600 (instead of 3200) would be enough to save the highlights, but that can't be determined any more.

Sfordphoto wrote in post #6313977
so what you are getting at is
1) that when highlights are critical, use a lower ISO, underexpose, and bring it up in post processing?
2) and when shadows are critical and highlights aren't so critical, use a higher ISO (since you said that 1600 had less noise than underexposed 800 brought up a stop)?

Exactly that's what I meant.

HTP is cheating with the ISO: the actual ISO will be one stop lower than the selected one (that's the reason it does not work with ISP 100). Thus HTP On causes a -1 EV exposure bias. That's all regarding the raw data.

When shooting JPEG or the raw file is processed by Canon software, the intensity will be adjusted by +1 EV, except for the very highlights. ACD/LR do the intensity adjustment, but don't keep back the highlights (one has to do that manually).

HTP too is a feature for JPEG. The one full stop exposure bias is quite brute, one can adjust it with less than one stop if required and compensate for that, falls the raw data is recorded.

please keep up this discussion. I am really gleaning a lot of helpful tips here. I am still working my way through "Understanding Exposure (Revised Ed.) and this topic is one I am really interested in understanding. I haven't learned enough yet to know what questions I have, but as soon as I get time to begin experimenting I know the comments here will be really helpful, so THANKS to everyone..

Panopeeper wrote in post #6313497
1. ISO 3200 is a numerical derivative of ISO 1600, with one stop underexposure. It creates exactly the same image, except that it clips the brightest stop, i.e. reduces the dynamic range by one stop. Identical result can be achieved by using ISO 1600 and adjusting the intensity in raw conversion, except that the dynamic range is not reduced.

2. ISO 1600 is "real" (as opposed to 3200). However, it is not better in the shadows than ISO 800 underexposed by one stop, but it clips the highlights. Again, the dynamic range is reduced in exchange for nothing.

With the 40D, 800 is the highest ISO, which offers something in the shadows in exchange for the lost stop of highlights.

So, yes, ISO 1600 and 3200 are worthless with raw recording. If someone does not like the term "worthless", one can say "negative worth".

However, when recording JPEG in-camera, one does not have the same degree of freedom for adjusting the intensity in PP; in that case, it can be reasonable to use 1600 and 3200.

Well Mr. mathematician ISO3200->ISO1600 is a FULL stop of LESS light inflated or not. At ISO1600 I *might* have gotten 1/320 *WHICH* it will not freeze the action or worst cause blurr. For sport a minimum of 1/500 shutter speed is needed to try to freeze action -- The shots get published and peeps get paid. You seem to be in rant about 40D's high ISO images being useless. Noise it is a fact of dSLR as much as it is dust sensor.

Do you have any more rants about noise? Or did we hear enough ?

Here is YET another *useless* ISO3200 image

HTTP response: 404 | MIME changed to 'text/html' | Byte size: ZERO

HTTP response: 404 | MIME changed to 'text/html' | Byte size: ZERO

silvex wrote in post #6314552
Or did we hear enought ?

The subject was, of course, using a lower ISO with underexposure (in other words: the exposure is as metered for the higher ISO). What else would the intensity adjustment in raw conversion be for?

See above:

1) that when highlights are critical, use a lower ISO, underexpose, and bring it up in post processing?

Panopeeper wrote in post #6314599
You seem not to profit from what you hear, so it does not matter. The subject was, of course, using a lower ISO with underexposure (in other words: the exposure is as metered for the higher ISO). What else would the intensity adjustment in raw conversion be for?

I clearly understand you ISO3200 statements. They *might* work for static subjects. For moving object light= contrast = sharp. So ISO800 -2EV = ISO3200 0EV. give and take.

silvex wrote in post #6314649
give and take.

I give it up.

Sfordphoto wrote in post #6298312

HTTP response: 404 | MIME changed to 'image/png' | Byte size: ZERO

Monterey bay aquarium, jellyfish (very small ones)

To give you an idea of how dimly lit it was...
50mm f/1.4 @ f/1.4, ISO 3200, 1/25 shutter speed

Good lord those ones are impossible.. good job for getting it. I love that place..

Panopeeper wrote in post #6314745
I give it up.

Come on...be a sports... Here you go both given same PP (one +2.85EV the other +0.85EV everything else the same)

ISO800 -2EV f1.4 1/125 with CWB FULL SIZE

HTTP response: 404 | MIME changed to 'text/html' | Byte size: ZERO

HTTP response: 404 | MIME changed to 'text/html' | Byte size: ZERO