[mattdm]

Argh! False precision.

We don't call it f/1.41421356237309504880, or f/22.62741699796952078083. Or even f/23, even though that's closer to correct.

And look at shutter speeds: 1, 2, 4, 8, 15!, 30, 60, 125, etc.

So, c'mon, Pentax (and everyone else). Maybe it's technically ISO 51,200, but just call it ISO 50k and be done with it. Same for 25k.

---

Last edited by mattdm; 09-22-2010 at 10:08 AM. Reason: wrong title. :)

[dgaies]

Argh! False precision.

We don't call it f/1.41421356237309504880, or f/22.62741699796952078083. Or even f/23, even though that's closer to correct.

And look at shutter speeds: 1, 2, 4, 8, 15!, 30, 60, 125, etc.

So, c'mon, Pentax (and everyone else). Maybe it's technically ISO 51,200, but just call it ISO 50k and be done with it. Same for 25k.

100% agree.

Flase precision is one of my pet peeves

[AngryCorgi]

100% agree.

Flase precision is one of my pet peeves

But false-precision is a fundamental concept upon which Pentax AF systems are based...




...jokes...jokes...

[16N64]

So, c'mon, Pentax (and everyone else). Maybe it's technically ISO 51,200, but just call it ISO 50k and be done with it. Same for 25k.

With these high sensitivity values, I wish ISO to be shown in the still valid DIN representation which is a log scale instead of the linear one in ASA. The 51200 ASA would read 48 DIN - much more readable, especially as sensitivity goes higher and higher.

So can we please have DIN back in advertising, EXIF data and tech talk?

Stefan

[Spongefingers]

With these high sensitivity values, I wish ISO to be shown in the still valid DIN representation which is a log scale instead of the linear one in ASA. The 51200 ASA would read 48 DIN - much more readable, especially as sensitivity goes higher and higher.

So can we please have DIN back in advertising, EXIF data and tech talk?

Stefan

Unlikely - not just because it's an entrenched standard, but because advertisers wouldn't like it. 51200 sounds a lot bigger than 48.
But... In the near future, when maximum ISO speeds get ridiculous (1.6 million is only a few years off, I reckon,) the avoidance of a log scale may become impractical!

[16N64]

Unlikely - not just because it's an entrenched standard, but because advertisers wouldn't like it. 51200 sounds a lot bigger than 48.
But... In the near future, when maximum ISO speeds get ridiculous (1.6 million is only a few years off, I reckon,) the avoidance of a log scale may become impractical!

It is already impractical, not only do 12800 and 102400 look too similar, the EXIF specification is already useless for any ASA values above 65535, as only a 16 bit value is stored. Some 102400 ISO Canon camera already has reached this limitation and saturates to 65535.

An alternative EXIF tag with DIN and the possibility to display sensitivity in DIN in the camera would be reasonable.

Stefan

[jogiba]

ISO 100 – 12800 (up to 102400 in H3 mode) Photography Blog

[imtheguy]

Dont forget about the 16.931234567890 total pixels sensor with effective 16.2812345678790 pixels.

[thibs]

Argh! False precision.

We don't call it f/1.41421356237309504880, or f/22.62741699796952078083. Or even f/23, even though that's closer to correct.

And look at shutter speeds: 1, 2, 4, 8, 15!, 30, 60, 125, etc.

So, c'mon, Pentax (and everyone else). Maybe it's technically ISO 51,200, but just call it ISO 50k and be done with it. Same for 25k.

Funny, our teacher just talked about that today during the course

[fzwo]

Yes, it's ridiculous. It's like saying a lens has a focal length of 31 mm (make it 30), or an aperture of 1.9 (how about 2?)

[ytterbium]

12k and 52k would be enough for easier communication and backwards compatibility. In future you could extend it to M's and G's, just like with computer drives.
But this still requires a new exif line. Better it be some standardized value than manufacturer dependent string representation.

I find it would be simpler to take iso 100 as base and count the EV's.
like ISO 0EV for 100, ISO 1EV for 200, 2 for 400, 3-800, 4-1600, ... , ISO 7EV for 25600. ISO -1EV for ISO 50. Or you could put 1 at iso 100 for better resemblance.. (just dont put it at sensor base iso, will make a mess when between cameras).
This would make it base 2 logarithmic, but in more easy to convert/remember way.

[Adam]

But false-precision is a fundamental concept upon which Pentax AF systems used to be based...

See what I did there?

[bymy141]

Argh! False precision.

We don't call it f/1.41421356237309504880, or f/22.62741699796952078083. Or even f/23, even though that's closer to correct.

And look at shutter speeds: 1, 2, 4, 8, 15!, 30, 60, 125, etc.

So, c'mon, Pentax (and everyone else). Maybe it's technically ISO 51,200, but just call it ISO 50k and be done with it. Same for 25k.

In that case our friends here complaining about flash sync speed will have a free upgrade!
The shutter mechanics will deliver a binairy precision:

1/1, 1/2, 1/4, 1/8, 1/16, 1/32, 1/64, 1/128, 1/160, 1/196, 1/256, ...

- Bert

[mattdm]

Yes, it's ridiculous. It's like saying a lens has a focal length of 31 mm (make it 30), or an aperture of 1.9 (how about 2?)

31mm vs. 30mm is quirky — less than 2° FOV difference, or about 1% shaved off the edge of the frame.

f/1.9 is 0.15 of a stop -- small, but enough to be worth noting.

But ISO 50,000 vs 51,200 is only 0.03 of a stop!

[Arpe]

Just divide all ISOs by 100 (call them metric ISOs! No wait, those in the USA couldn't handle that ), they start at 100 anyway, usually now. so 1, 2, 4, 8....