ISO 2500 vs 160

[photoarne]

Advertisement (gone after registration)

Finally, to show another product, I have converted in BibblePro, latest version. First is the picture at ISO 160 as it comes out of the camera and then the ISO 2500 adjusted in BibblePro and then the ISO 160 adjusted in BibblePro. I find this exercise rather telling. I also tried the older version of C1 3.7.7 and I could not get as good results. These are all at pixel peeping 100% as are the former. At normal size they all looked pretty good.

Thank you for an interesting test, seems underexposure is the way to go. My question is in reference to the comment you made on the "older" C1 3.7.7. I think this is the current version, were the conversions shown made with the beta C1 4? And did you find this version to be better (in this respect) than 3.7.7., or is the comparison between the Bibble and C1. I haven't downloaded the beta C1 yet due to numerous reports of bugs, so I'm curious.

[adan]

1. Actually, I was doing this (as some here may remember) as far back as the Digilux 2, where a jpeg shot at 100 and pushed to 400 in Photoshop avoided the built-in noise reduction and came out sharper (but noisier) than a straight 400 exposure.

 

2. It never hurts to remember that all digital sensors (unless someone has a counter-example) basically have ONE ISO. Shine x amount of photons on them and they will output y amount of electrons/voltage/charge. All the "higher" ISO settings just amplify the base signal.

 

3. In discussing the Nikon D3's ISO 25600, someone expressed the opinion that digital amplification (ISO "boost" in a camera - or exposure adjustments in RAW - or levels adjustments in Photoshop) are somehow different than analog amplification done on the raw electrons before digitizing. Personally, it seems to me that digital boosting or "pushing" is often at least equal to, and occasionally better than, the analog boost achieved by setting the "ISO" higher. But it depends...

 

4. As to the last point of getting blown highlights when pushing 4x - you might try using a mix of exposure and brightness settings (those are the Camera RAW terms - other RAW software may use different terms). In ACR, the exposure slider moves all tones, including the tip of the histogram, to the right, and can blow it off the end. The "brightness" setting shifts the middle of the scale to the right without having much effect on the right-hand tip. So a 2x or 3x "exposure" change - whatever just prevents blowout - combined with a midtone brightness increase, may work better than doing the entire push solely with the exposure.

 

5. Leaving aside CB - many of my pictures do not include anything approaching a pure white. It just depends on the subject. One of those rules we get taught is that the picture should have a "full tonal scale" that just touches black and white. Which is a decent guideline for beginners, who often (in film days) tended to print too flat and gray. But it really is subject-dependent - if I shoot a foggy scene and want to retain the atmosphere (literally and figuratively), that image will wind up with a "black" (on the 0-256 scale) of perhaps 60 and a "white" of perhaps 200-220. Sometimes I want high-key, bright "fog" - sometimes something darker and moodier.

[farnz]

... 3. In discussing the Nikon D3's ISO 25600, someone expressed the opinion that digital amplification (ISO "boost" in a camera - or exposure adjustments in RAW - or levels adjustments in Photoshop) are somehow different than analog amplification done on the raw electrons before digitizing. Personally, it seems to me that digital boosting or "pushing" is often at least equal to, and occasionally better than, the analog boost achieved by setting the "ISO" higher. But it depends...

Andy,

 

I thnk that ISO boost and in-pp exposure or levels adjustment are quite different. ISO boost inreases the sensors' sensitivity rather than amplifying its signal. The low signal level threshhold in sensor elements (dubbed "sensels" in another post) are set to a certain value, below which everything is determined to be noise and is discarded. (Actually the threshhold is set by the signal to noise ratio being worse than an arbitrary level, say, -45 dB.) Increasing the ISO value lowers the threshhold to another arbitrary value where more signal is captured but the signal to noise ratio is inherently worse, say, -30 dB.

 

Take the analogy of dredging in a river that's 5 metres deep: from 0 to 5 metres all you'll pick up is water (signal) and a little bit of mud (noise) as you skim the tops of the mud peaks. If you lower your bucket (threshhold) to, say, 5.5 metres then you'll pick up most of the water (signal) that you did before but also a lot more mud (noise) and also some water (signal) that was mingled with the mud or on the mud troughs. In this way you're picking up some water that you wouldn't have picked up before because it was mingled with mud but the drawback is that you have to pick up more mud to do it. Hence the signal to noise ratio is worse and the image is noisier.

 

Otoh, in-pp exposure and levels adjustments use digitised pixels (picture elements), which are processed packages of the analogue output from the sensels that have been assigned digital values based on their signal level (through an ADC) and assigned colour values based on the Bayer matrix and the raw interpretaion software's algorithms.

 

The only way that I can see for noise to enter the exposure and levels adjustment processes - and I'm guessing here - is interpolation noise; eg if you start with a 2x2 matrix of pixels and end up with a 4x4 matrix then something's been interpolated, and therefore created which will by definition be partly noise. The same concept applies to exposure and levels because you are taking pixels with set exposure values and artificially increasing the values so something that wasn't there before is added and you will worsen the signal to noise ratio.

 

So it appears to me that in theory in-pp exposure adjustment should yield less noisy results than boosting the ISO level.

 

Pete.

[Guest stnami]

,,,,,,as stated a old ploy but a good one

[Jamie Roberts]

{snipped} if I shoot a foggy scene and want to retain the atmosphere (literally and figuratively), that image will wind up with a "black" (on the 0-256 scale) of perhaps 60 and a "white" of perhaps 200-220. Sometimes I want high-key, bright "fog" - sometimes something darker and moodier.

 

Andy--absolutely--but here's where people sometimes get confused with digital. If you don't mind burying the shadows in your fog, or need to do this for other reasons, then a high-ISO underexposure from average might be perfectly atmospheric...

 

But, and this is the big but, you will never regain the data below the noise floor.

 

So, consequently, you will never, ever go wrong "properly exposing" (IOW pushing to the right till the ambient is lighter) a scene like that shooting RAW and pulling it back in post.

 

You can always create a dark mood, IMO, with a well-developed shot. Getting a high key fog in post and YMMV It certainly can be done, especially if you have something like the M8 at lower ISOs.