medium format and low light

[DaveRobertson]

In light of the DXOMark (which I take with a grain of salt because it seems to omit some key factors) why do the medium format sensors perform so poorly in low light situations compared to some of their small format cousins?  I use both, playing to their strengths, but would love to use medium format for everything.

(I'm talking about performance at higher ISO's, not time exposures)

Dave

[ErikKaffehr]

That is a question I also asked many times. To me it does not make any sense that MF sensors are said to have good DR but bad high ISO performance.

On the other hand, Michael Reichmann is very clear that there is a significant advantage to MF in large enlargements, and I don't think that he is dogmatic person. I would say that I'd think that Mr. Reichmann is right, until proven otherwise.

Best regards
Erik

In light of the DXOMark (which I take with a grain of salt because it seems to omit some key factors) why do the medium format sensors perform so poorly in low light situations compared to some of their small format cousins?  I use both, playing to their strengths, but would love to use medium format for everything.

(I'm talking about performance at higher ISO's, not time exposures)

Dave

[pindman]

Which is better -- a car or a boat?

Henri Cartier-Bresson used a Leica.  Ansel Adams used a view camera.  Is one camera better than the other?

If you test how fast a boat can travel on a freeway, or a car can go in the ocean you will reach a different conclusion as to which is better vehicle.

Paul

[whawn]

(I'm talking about performance at higher ISO's, not time exposures)

For the P45+, at least, it looks like the back doesn't really have higher ISO levels.  As near as I can read the charts, it seems the setting changes nothing once you get above 100 ISO, so all you're doing is underexposing (which, honestly, is all you're doing with ANY digital camera, except that the others amplify the sensor signals in-camera to fake the ISO output change) and then C1 automatically makes up the difference, so it looks like it's coming out of the camera that way.  Near as I can figure, Phase has a 'native' ISO of 100, and the 50 mark is achieved by gain reduction.  I kinda like their strategy, because it means you'll very seldom have any blown out pixels at higher ISO settings.

The way I read the charts, noise performance isn't so much worse than other digital sensors, and my own experience is that the p45 (not plused) is very clean at all the programmed levels (only 50, 100, 200, 400).

What I'm wondering:  Since they've added the MF backs to the DxOMark site, does that mean they'll be adding MF backs to the DxO converter app?  I'd like to see that.  I'd even pay for it.

[DaveRobertson]

A little pedantic in your reply, Paul, and not terribly useful.  I'm just wondering what makes the high iso response different in the sensors, which is a fair question.  I'm not asking about boats and cars.

Regards,
Dave

Which is better -- a car or a boat?

Henri Cartier-Bresson used a Leica.  Ansel Adams used a view camera.  Is one camera better than the other?

If you test how fast a boat can travel on a freeway, or a car can go in the ocean you will reach a different conclusion as to which is better vehicle.

Paul

[DaveRobertson]

Thanks Walter, that makes sense.  I've just found in the H3DII39 and the P25 (non-plus) ISO's of 200 and up are pretty useless.  I suppose the manufacturers must have considered the kind of amplification going on in the smaller formats but rejected it for some reason.

Dave

[quote name='whawn' date='Feb 4 2009, 10:46 PM' post='257510']
For the P45+, at least, it looks like the back doesn't really have higher ISO levels.  As near as I can read the charts, it seems the setting changes nothing once you get above 100 ISO, so all you're doing is underexposing (which, honestly, is all you're doing with ANY digital camera, except that the others amplify the sensor signals in-camera to fake the ISO output change) and then C1 automatically makes up the difference, so it looks like it's coming out of the camera that way.  Near as I can figure, Phase has a 'native' ISO of 100, and the 50 mark is achieved by gain reduction.  I kinda like their strategy, because it means you'll very seldom have any blown out pixels at higher ISO settings.

The way I read the charts, noise performance isn't so much worse than other digital sensors, and my own experience is that the p45 (not plused) is very clean at all the programmed levels (only 50, 100, 200, 400).