do you believe

[stasvolik]

Another quick test (made with D50, sorry).

why sorry...

second shot with less highlights blown out is ISO 400 pushed?

Yes. I was really surprised by the difference (the exposure was longer than on the first one, yet the overall tonality is much more pleasant to my eye).
--
Cheers,
Stas.

 

[wilkman]

Here's a few shots to try and see how much practical signifigance we are facing.....

The first shot is at iso 100 and is two stops underexposed.



The second is at iso 1600 and is two stops overexposed.



So there is 4 stops of difference between the two, albeit the sensor recieved exactly the same light since both were taken at identical aperture and shutter.

Here's the "underexposed" pushed two stops in post.



Here's the "overexposed" cut two stops in post.



To my eyes, on this set and many other exaggerated sets I've tried, noise differences are usually quite small. However, contrast, detail, and color saturation almost always favor cutting exposure in post rather than pushing exposure. The detail in the wall paper seems to be the most striking difference in this set. (Focus was manual and unchanged between the two, as was exposure of course)

But this is 4 freakin stops of difference ..... when i compare more realistic sets of photos that are only a stop or two apart, it just doesn't make any signifigant difference how I trade the over-or-underexposure of in cam iso, vs post processing compensation. The picts look the same.

So I believe all that practically matters is aperture and shutter. Getting it right, while staying as close as possible to the cam's cleanest native iso, seems to be the name of the game. And more often than not, that just comes down to raising iso to get the sufficient shutter speed for whatever we're trying to shoot.... I mean who raises iso when you can lower shutter?

Interesting thread for sure, but not anything to change how we shoot or process.....

best, mark

 

[Vieri Bottazzini]

is contrast, IMHO. I added a little curve to your underexp/pushed image and looked like the overexp/pulled one...

thank you for the test, though - it shows that there is very little difference, at this small size, between the two situations. Seeing a 100% crop would have been more interesting though.

Best regards,

Vieri

--
equipment in profile

http://www.madshutter.com

more stuff here:
http://www.pbase.com/vieripbase

 

[DonBJr]

"do you believe..." seemed more like a morpheus impersonation to me.

 

[wilkman]

is contrast, IMHO. I added a little curve to your underexp/pushed
image and looked like the overexp/pulled one...

thank you for the test, though - it shows that there is very little
difference, at this small size, between the two situations. Seeing
a 100% crop would have been more interesting though.

.... it probably is more due to contrast than anything. About the only differences I ever see in these type comparisons is a little bit of satuaration/contrast .... and that is with zero effort to get the difference out in post other than adjusting exposure.

Here's a couple of 100% ers ....

the +2 push in post



the -2 cut in post



Does any of this matter????

best, mark

 

[Vieri Bottazzini]

it does matter, though very little - to me, it looks like the top image (push) has slightly higher noise & slightly higher apparent details; lower image, lower apparent noise and details. Though, both look darn good to me as far as 1600 ISO goes - wouldn't you agree?

Best regards,

Vieri

--
equipment in profile

http://www.madshutter.com

more stuff here:
http://www.pbase.com/vieripbase

 

[Flashlight]

Shhhh! Don't let Julia hear you! It's not 1600 ISO, it's four stops amplification of the base signal...

Ah well, what do I know... Still fighting uphill noise in my ivory tower...

Though,
both look darn good to me as far as 1600 ISO goes - wouldn't you
agree?

--
Philip

 

[wilkman]

Though, both look darn good to me as far as 1600 ISO goes - wouldn't > you agree?

Best regards,

Vieri

Hi Vieri, yea, I'd say both look pretty good.

But you know in my mind, they are both iso 400 shots. That's the iso I used to set aperture and shutter......(so that iso 100 became -2EV and iso 1600 became +2. )

IMO, I don't care how much pushing or pulling is done in post, whatever iso would cause the camera to meter flat at time of capture, that is the iso of the shot.

I wonder how many people see it simarly?

best, mark

 

[Flashlight]

Though, both look darn good to me as far as 1600 ISO goes - wouldn't > you agree?

Best regards,

Vieri

Hi Vieri, yea, I'd say both look pretty good.

But you know in my mind, they are both iso 400 shots. That's the
iso I used to set aperture and shutter......(so that iso 100 became
-2EV and iso 1600 became +2. )

IMO, I don't care how much pushing or pulling is done in post,
whatever iso would cause the camera to meter flat at time of
capture, that is the iso of the shot.

I wonder how many people see it simarly?

best, mark

I agree. The difference is that the PP latitude is gone if you shoot a scene at -2 or +2 EV. I mean, if you shoot it at -2 EV and you think later you should add another stop to the exposure, you have a problem.

--
Philip

 

[Vieri Bottazzini]

Mark,

you do have a good point here:

But you know in my mind, they are both iso 400 shots. That's the
iso I used to set aperture and shutter......(so that iso 100 became
-2EV and iso 1600 became +2. )

IMO, I don't care how much pushing or pulling is done in post,
whatever iso would cause the camera to meter flat at time of
capture, that is the iso of the shot.

However, similarly to film days, sensors and their response to push-pull in post is something we do have to take into account when deciding, after knowing that this would have been a ISO 400 exposure, which way to go (up ISO, down in post - down ISO, up in post).

Sort of using a ISO 100 film exposed to 400 or a ISO 400 film exposed properly or a ISO 1600 underexposed. Very different results in film days - not so different results today with digital, according to your finding.

Very interesting stuff, isn't it?

Best,

Vieri

--
equipment in profile

http://www.madshutter.com

more stuff here:
http://www.pbase.com/vieripbase

 

[Julia Borg]

"do you believe..." seemed more like a morpheus impersonation to
me.

think matrix... metering...

--
Julia

 

[Julia Borg]

And now for the important question: When would you recommend
pushing afterwards vs pushing in camera for the best quality?
Always?

absolutely when I need maximum dynamic range to hold details in highlights. like it is with weddings.

--
Julia

 

[Julia Borg]

it matters when I want to preserve higher tones and highlights, minimising clipping.

results depend on raw converter very much.

--
Julia

 

[Lasse Eisele]

absolutely when I need maximum dynamic range to hold details in
highlights. like it is with weddings.

I took that for given. Let me rephrase the question: Under what circumstances would you not recommend pushing afterwards?

 

[wilkman]

I agree. The difference is that the PP latitude is gone if you
shoot a scene at -2 or +2 EV. I mean, if you shoot it at -2 EV and
you think later you should add another stop to the exposure, you
have a problem.

Yes indeed.... keeping lattitude. And I have to believe we maximize latitude simply by knowing how we want to expose. ... IOW, by getting aperture and shutter right to begin with.

The way I see it, ISO really has no meaning for determining exposure, other than it is a number that carries a certain amount of noise with it, and it will lead us to choose a particular aperture and shutter.

So we simply use the least-noise ISO we can, and then we set aperture/shutter for either mid, shadows, or highlights, depending on the shot, and what we need to emphasize or protect.

I have a hard time thinking much really matters past this....

best, mark

 

[RFPhotonut]

Can you please summarize what the conclusions are for those of us that don't want to read each response in this thread? I saw your other thread in the D40/etc forum asking what the differences are between shots of a bookshelf, but never saw what the point was. I am also not sure what your point is in this thread, except that it would seem that it is better to overexpose at a higher ISO than under at a lower.

Thanks!

 

[jeffkrol]

Saturation based or noise based........
http://www.qualiteitems.com/images/ISO.pdf
--
360 minutes from the prime meridian. (-5375min, 3.55sec) 1093' above sea level.

'The exposure meter is calibrated to some clearly defined standards and the user needs to adjust his working method and his subject matter to these values. It does not help to suppose all kinds of assumptions that do not exist.'
Erwin Puts

 

[Julia Borg]

it is not ISO that brings noise, it is underexposure that brings decrease of dynamic range.

changing ISO you create fake histogram, same as with white balance.

what benefit do you gain increasing ISO?

--
Julia

 

[Michael Mohrmann]

Years from now cryptologists won't be able to decipher the meaning of this thread. There will be college courses constructed in honor of threads like this one.

j/k

 

[Julia Borg]

if normally you are on ISO 400 and higher, spot-metering from important highlights does not need exposure compensation.

or, "M" mode and Auto -2.5eV works for me quite well.

one needs to experiment to determine the "sweet ISO point" for the camera/converter combination.

I have markings on the LCD for -1, -2, -3 eV. ths way I know the actual amount of underexposure.

--
Julia