Are we too critical with the Foveon sensor????

I'll have to remember when I get another Sigma that I should overexpose a little, when I shoot at high ISO settings . . . unless it's a Quattro, huh?
Assuming that el cuatro has an AFE - yep.

On ISO-less cameras, you can also leave it at your favorite ISO and under-expose. Makes no difference to the raw data itself, only to how SPP shows the review image.

In other words, in SPP, for a high-ISO 'overexposed' shot, you'd "pull back" those apparently blown highlights or, for a low-ISO 'underexposed' shot, you would "push" those overly-dark shadows
 
I'll have to remember when I get another Sigma that I should overexpose a little, when I shoot at high ISO settings . . . unless it's a Quattro, huh?
Assuming that el cuatro has an AFE - yep.

On ISO-less cameras, you can also leave it at your favorite ISO and under-expose. Makes no difference to the raw data itself, only to how SPP shows the review image.

In other words, in SPP, for a high-ISO 'overexposed' shot, you'd "pull back" those apparently blown highlights or, for a low-ISO 'underexposed' shot, you would "push" those overly-dark shadows
And which Sigma cameras are certainly known to be ISO-less ?
 
I'll have to remember when I get another Sigma that I should overexpose a little, when I shoot at high ISO settings . . . unless it's a Quattro, huh?
Assuming that el cuatro has an AFE - yep.

On ISO-less cameras, you can also leave it at your favorite ISO and under-expose. Makes no difference to the raw data itself, only to how SPP shows the review image.

In other words, in SPP, for a high-ISO 'overexposed' shot, you'd "pull back" those apparently blown highlights or, for a low-ISO 'underexposed' shot, you would "push" those overly-dark shadows
And which Sigma cameras are certainly known to be ISO-less ?
SD9, SD10, SD14, DP1, DP2 and more.

But the tone of question appears to imply some doubt. Need proof?

--
Cheers,
Ted
 
Last edited:
. . . and not enough on sensor exposure?
ISO gain is simply a multiplier factor applied in software. Exposure (= amount of light) is the sole determinant of image quality (in the context our discussion).

If the camera has an AFE, then the gain is applied in hardware and is more helpful.
 
You can set it to 1600, over expose and recover highlights.
Why would you set the ISO to 1600 and then over-expose? In low light you should always expose as long as possible (to avoid motion blur), so if you can afford a longer exposure time why not go with ISO 800 from the beginning. Anyway, the ISO is secondary, its only importance is that it tells SPP how much to denoise. It does not change anything in the raw data.
 
. . . and not enough on sensor exposure?
ISO gain is simply a multiplier factor applied in software. Exposure (= amount of light) is the sole determinant of image quality (in the context our discussion).

If the camera has an AFE, then the gain is applied in hardware and is more helpful.
Thank you.
 
In low light you should always expose as long as possible (to avoid motion blur)
An interesting assertion - quite the opposite of normal photographic practice!

--
Ted
 
Last edited:
I'll have to remember when I get another Sigma that I should overexpose a little, when I shoot at high ISO settings . . . unless it's a Quattro, huh?
Assuming that el cuatro has an AFE - yep.

On ISO-less cameras, you can also leave it at your favorite ISO and under-expose. Makes no difference to the raw data itself, only to how SPP shows the review image.

In other words, in SPP, for a high-ISO 'overexposed' shot, you'd "pull back" those apparently blown highlights or, for a low-ISO 'underexposed' shot, you would "push" those overly-dark shadows
And which Sigma cameras are certainly known to be ISO-less ?
SD9, SD10, SD14, DP1, DP2 and more.

But the tone of question appears to imply some doubt. Need proof?
As far as I know

With AFE: DP1x, DP2x, SD15, DP2Q.

The rest is without.
 
If the camera has an AFE, then the gain is applied in hardware and is more helpful.
It simplifies life, yes. And if the sensor has very low noise it also improves the image.

But - the Foveon sensor do not have very low noise.

And - applying the gain makes the A/D converter clip highlights.

So, for Foveon sensors it is not so clear cut if it is advantage or not. Maybe it is for Quattro that have better noise behaviour.
 
Roland Karlsson wrote:

And - applying the gain makes the A/D converter clip highlights.
Not necessarily. If you shoot at, say, ISO 400 and expose properly you can set the gain to 4x and you won't clip. But if you overexpose or use a gain > 1 at ISO 100 you will clip.

So the gain has to be tied to the ISO selection.
 
And - applying the gain makes the A/D converter clip highlights.
Not necessarily. If you shoot at, say, ISO 400 and expose properly you can set the gain to 4x and you won't clip. But if you overexpose or use a gain > 1 at ISO 100 you will clip.

So the gain has to be tied to the ISO selection.
I'm confused. How does a User "set the gain to 4X" in, for example, a Quattro?

Equally, how does one "use a gain > 1"?

These are not trick questions, I just don't understand your post, sorry.
 
And - applying the gain makes the A/D converter clip highlights.
Not necessarily. If you shoot at, say, ISO 400 and expose properly you can set the gain to 4x and you won't clip. But if you overexpose or use a gain > 1 at ISO 100 you will clip.

So the gain has to be tied to the ISO selection.
So - I have to be more precise then :)

If you amplify the analogue signal then the headroom gets smaller and you might clip. This is something you fully avoid if you do not amplify.
 
If you amplify the analogue signal then the headroom gets smaller and you might clip. This is something you fully avoid if you do not amplify.

--
OK. But when shooting at high ISO you don't really worry about clipping, and the analog amplifier is beneficial, because it will reduce the ADC quantization noise.

Of course, in low light, if you can use a tripod and there is no motion, then you want to use long exposure time and set ISO=100, effectively disabling the amplifier.
 
And - applying the gain makes the A/D converter clip highlights.
Not necessarily. If you shoot at, say, ISO 400 and expose properly you can set the gain to 4x and you won't clip. But if you overexpose or use a gain > 1 at ISO 100 you will clip.

So the gain has to be tied to the ISO selection.
I'm confused. How does a User "set the gain to 4X" in, for example, a Quattro?
By setting ISO=400 in the camera--it's automatic.
Equally, how does one "use a gain > 1"?
Setting the ISO to > 100 will do that.

(Assuming the Q. analog gain does work like that )

These are not trick questions, I just don't understand your post, sorry.

--
Cheers,
Ted
 
OK. But when shooting at high ISO you don't really worry about clipping,
Why? Why cannot high ISO images have high contrast?
and the analog amplifier is beneficial, because it will reduce the ADC quantization noise.
Yes, but if the sensor is very noisy (like the Foveon is) then you do not gain so much.
Of course, in low light, if you can use a tripod and there is no motion, then you want to use long exposure time and set ISO=100, effectively disabling the amplifier.
 
And - applying the gain makes the A/D converter clip highlights.
Not necessarily. If you shoot at, say, ISO 400 and expose properly you can set the gain to 4x and you won't clip. But if you overexpose or use a gain > 1 at ISO 100 you will clip.

So the gain has to be tied to the ISO selection.
I'm confused. How does a User "set the gain to 4X" in, for example, a Quattro?
By setting ISO=400 in the camera--it's automatic.
Equally, how does one "use a gain > 1"?
Setting the ISO to > 100 will do that.

(Assuming the Q. analog gain does work like that )
I did some initial measurements of AFE for Quattro. The amplification is not linear with ISO, it increases more slowly. I plan to do more accurate measurements when I get time.
 
OK. But when shooting at high ISO you don't really worry about clipping,
Why? Why cannot high ISO images have high contrast?
The quoted statement is either too general or it is over-simplified - whichever applies. If, for example, I owned a SD15, I would be very worried about clipping (assuming that the worthy member is referring to ADC clipping).
 
OK. But when shooting at high ISO you don't really worry about clipping,
Why? Why cannot high ISO images have high contrast?
The quoted statement is either too general or it is over-simplified - whichever applies. If, for example, I owned a SD15, I would be very worried about clipping (assuming that the worthy member is referring to ADC clipping).
I am a little amazed where discussions leads.

All I said was that if you do not amplify the signal, then you have better headroom. That is obvious.

And - It is a great advantage of having no AFE. At least I think so.
 
OK. But when shooting at high ISO you don't really worry about clipping,
Why? Why cannot high ISO images have high contrast?
High dynamic range? Sure they can. But there are at least 2 reasons to use high ISO: low light scenes and fast moving objects. Not sure that in such cases you worry about some blown highlights (like the lights in a dark bar)

and the analog amplifier is beneficial, because it will reduce the ADC quantization noise.
Yes, but if the sensor is very noisy (like the Foveon is) then you do not gain so much.
Of course, in low light, if you can use a tripod and there is no motion, then you want to use long exposure time and set ISO=100, effectively disabling the amplifier.
--
/Roland
X3F tools:
http://www.proxel.se/x3f.html
https://github.com/rolkar/x3f
 

Keyboard shortcuts

Back
Top