40D at ISO 3200 - why is it not on by default

Any ideas why canon leaves the ISO 3200 feature off and hidden away on the 40D? Does it generate to much heat or reduce battery lifetime if left on and not used? Or what?

Shot taken: F4, 1/125. ISO 3200, Noise reduced with Canon Digital Photo Professional 3.4.1.1.

Because is not a native ISO, it´s just an electronic game ... but you can select its activation to be always on by Cf-n ...

If you didn't understand that explanation, then I'll expand it a bit. This is still a bit simplified, but anyway.

The ISO levels 100-1600 are achieved by changing the amplification when reading the data from the CMOS sensor.
When you use expansion modes, in this case H, which is equivalent to ISO 3200, the camera actually exposes according to ISO 1600, then just multiply by two. This reduces the range from the sensor, effectively making the 14-bit sensor in the 40D a 13-bit instead.
So there's a quality loss from a different reason, and that's why Canon makes you think once more, before you can use it.

ahhh makes perfect sense now

apersson850 wrote in post #6596653
...
When you use expansion modes, in this case H, which is equivalent to ISO 3200, the camera actually exposes according to ISO 1600, then just multiply by two. This reduces the range from the sensor, effectively making the 14-bit sensor in the 40D a 13-bit instead.....

Hmm.. this is not how I understand the expanded modes. M y understanding has always been that the H mode is basically dialing in an exposure compensation value of -1 then when the raw file is processed the exposure is pushed by +1 thus you gain a full stop in speed. You can accomplish the same thing by shooting EC-1 and in DPP add exposure +1 and will get the exact same pic.

gjl711 wrote in post #6597180
Hmm.. this is not how I understand the expanded modes. M y understanding has always been that the H mode is basically dialing in an exposure compensation value of -1 then when the raw file is processed the exposure is pushed by +1 thus you gain a full stop in speed. You can accomplish the same thing by shooting EC-1 and in DPP add exposure +1 and will get the exact same pic.

Originally Posted by apersson850

HTTP response: NOT FOUND | MIME changed to 'text/html'

So why would you loose 1/2 of the data? Something is not making sense.

gjl711 wrote in post #6598166
So why would you loose 1/2 of the data? Something is not making sense.

You don't. You lose 1/14 of the dynamic range. A shift of one bit is the same as multiplying ISO by 2.

-js

Thanks, it clarifies the concept and I will keep this in mind when moving into the expanded range. That may also be the reason why Canon shows H in the LCD and not 3200 anymore. It is not real 3200 The 50D obviously uses the same method to go from 3200 to its expanded mode.

number six wrote in post #6598953
You don't. You lose 1/14 of the dynamic range. A shift of one bit is the same as multiplying ISO by 2.

-js

But loosing 1 bit does not loose 1/14, it looses 1/2 What you guys are saying is that you go from 2 to the 14th to 2 to the 13th or 1/2 of the data. Isn't the data just remapped to a different range?

gjl711 wrote in post #6600230
But loosing 1 bit does not loose 1/14, it looses 1/2 What you guys are saying is that you go from 2 to the 14th to 2 to the 13th or 1/2 of the data. Isn't the data just remapped to a different range?

Arithmetically it is 1/2 of the data...2^13 vs 2^14...but photographically it is taking 14 levels of range to 13 levels...the top and bottom are both the same, but you lose an intermdiate value, and that level is mapped (rounded) up or down to the next range instead. You're both right!