I took a picture of a black hole with this camera at the highest ISO setting. Here it is:
I used a relatively fast shutter speed to avoid overexposure of the entire frame.
9 Likes
Now with Neutrino sensor!!!
5 Likes
Right, the quantum efficiency is some large fraction of 1. But with larger sensors each pixel bucket can get larger, so there is a higher count per pixel.
I may have been born at night, but it wasnāt LAST night.
EDIT: @newliminted looks like I was misinformed about that being Uranus. It was Venus. but thatās still Jupiter up there. It was this night: http://www.iceinspace.com.au/63-675-0-0-1-0.html
They are getting better; Iād be quite willing to use up to ISO 12,800 or even 25,600 on the D4S, based on the sample images Iāve seen, whereas a camera I have from about 6-7 years ago I tried to keep below 800 whenever possible.
But these extremely high ISO levels? They seem to be mainly for bragging rights and marketing copy. Certainly not for taking photos someone who shelled out $6500 for a body would likely want to put in their portfolio.
4 Likes
Itās one more than Iso 204800, which was available (as Hi 2) with the Canon 1D-X and the Nikon D4. The march of Progress.
http://www.fredparker.com/ultexp1.htm
may be useful
letās say that you have a f1.4 lens, 1/250 is about the slowest speed you feel comfortable with, and youāre using iso3200 film.
Pody says that these settings are good for
11
EV11 Sunsets. Subjects in deep shade.
409600 is 7 stops better than 3200.
whatās EV 4?
Candle lit close-ups. Christmas lights, floodlit buildings, fountains, and monuments. Subjects under bright street lampsā¦
And thatās at 1/250s.
At 1/60, you might be able to get pictures of a total lunar eclipse.
of course, this assumes that 409600 isnāt grainy. My guess is that 25600 will be the practical maximum.
2 Likes
as others have pointed out, many of these extended ISOs are just digital multiplication of the data that comes off the A-D converter. in that sense they are truly a gimmick because you can easily multiply the data yourself with photoshop or similarā¦
the ārealā ISOs are essentially gain settings for the (analog) read amplifiers on the CCDs themselves.
increasing these ISOs lowers the available dynamic range of the camera. having said that in low light situations the high, real ISOs are sometimes the difference between getting an image and not getting an image, dynamic range be dammed.
if you are interested in this nonsense (as i am) you can read http://clarkvision.com, and in particular, this page:
http://clarkvision.com/articles/digital.sensor.performance.summary/index.html
4 Likes
I checked out DPReviewās āfirst lookā at the camera. Yeah, the noise in the RAW images at that top ISO was badā¦ but not so bad that you wouldnāt use it to get the shot if you were a pro shooting something that mattered.
It looks like the additional settings past the 25,600 are binning some number of photo sites to increase the effective sensitivity at the sacrifice of resolution. If you look at the Hi-4,3,2,1, you can see the noise blocks shrinking at each slower speed. Itās a trick that has been done with astronomical CCDs for about as long as theyāve been around, I hadnāt realized it was this entrenched in regular photo gear now (shows how Iāve been out of it for a while.)
Advantages? How about shooting something at the silly-high ISO to confirm your framing and then reducing the ISO and increasing the exposure time to get a good, solid picture. Or even to just find something you are looking for that is faint. The last set of images I took of Hyakutake (on film) I couldnāt see the comet; but it showed up fine on film. Now if I had something like this I could have moved my setup to avoid the wire I couldnāt see in the darkā¦
What I found boggling was the 11 frame per second full-resolution for āapproximatelyā 200 frames. Thatās huge.
Oh, other reasons for the high ISO? Shooting F1 cars through a long telephoto on a dark, cloudy day in the rain.
And then thereās concert photosā¦
Want. Donāt need; I havenāt used my D2H in almost a year. But, stillā¦ want.
4 Likes
amazing - same happened to me with hyakutake. went way out somewhere dark with a pentax K50 and some kodachrome film. developed the pictures to find a telephone wire across the image that i just could not see in the dark.
binning makes sense; i did not do any reading about this camera. hardware binning does have the advantage of getting 4 pixels worth of signal for 1 read noise hit, but then again thatās for CCDs. generally speaking you canāt bin an OSC CCD sensor, but CMOS sensor readout circuitry is different and perhaps you can actually bin an OSC CMOS sensor.
1 Like
That looks like a folk etymology. Every reference Iāve seen says it is because the standard was established by the I.S.O., aka the International Organization for Standardization. (No, I donāt know when it stopped being referred to as the International Standards Organization, which makes a lot more senseā¦)
As in: āThe ISO stands for International Organization for Standards, and ISO 5800:1987 is the standard assumed when the context is film speed.ā
1 Like
As you get into newer generations of cameras with these hight ISO numbers (and the D4S is at the top of that currently) ISO setting that were unusable on the last generation of camera are suddenly usable. I have a D7100 and the pictures at ISO3200 and even ISO6400 are very usable. I wouldnāt go up to Hi2 aprox ISO=25,600 but on the D4S Iād expect ISO 25,600 to be very usable.
My first DSLR I wouldnāt shoot over ISO800.
3 Likes
Itās almost always a new sensor. This sensors highest native ISO is 25,600 so 25,600 * 16 = 409,600.
Baring a collapse of the DSLR market in three generations there will be sensors that might have a native ISO almost as high as this extended ISO and will certainly have extended ISOās in the millions.
1 Like
I almost never shoot with my D3s above 6400. I wonder realistically how much more the D4s gives? One stop? Two at most, Iāll bet. Still good, though.
Those crazy high 6 digit ISOs are only good for snapping a noisy shot of a fleeing getaway car at night.
dxomark says:
Low light capabilities are exceptional, but the Nikon D4s ranks third in this category after the Nikon Df, and the older, now discontinued Nikon D3s.
Hmm.
Unless, you know, physics actually kicks in (and weāre already knocking at its door).
1 Like
The day-to-day application of higher ISO is often a lot less dramatic or technical-seeming than many of the examples Iāve seen given here. Being able to shoot at, say, 6400 without any hesitation allows faster shutter speeds and smaller apertures and a greater use of natural light than in the past. Having a few more up-clicks of sensitivity at the ready is simply better for capturing scenes in marginal light in a naturalistic manner, better approximating what we see with our eyes.
For example, the image linked below is not astounding. Itās in an antique shop in Madrid lit from daylight pouring through the window. The boy was bored as his parents browsed, and his companion Great Dane squatted down next to him. At 6400 ISO I was able to shoot quickly, capturing the small moment with the shutter sharpness and focal depth that tells the story a lot better than in the old days, shooting with Tri-X (400 ASA). But even here, I would have liked another stop of shutter speed, eliminating the slight movement detectable across the Great Daneās eyes.
For me, this is where higher ISO makes my work better.
6 Likes
Thatās odd because the D4s goes to 400,000+ while the D3s maxes out at 102400, although I start to dislike the noise it produces north of 6400.
The D3s is a somewhat lower resolution camera, which makes good high ISO performance easier as I understand it.
Some sample imagery?
http://froknowsphoto.com/nikon-d4s-sample-images-iso-409600/
http://nikonrumors.com/2014/03/08/nikon-d4-vs-d4s-high-iso-comparison.aspx/
409600 looks like a cheap parlor trick
In some sense the reverse is true: silicon sensors always suffer from some degree of nonlinearity (āhot pixelsā, with some pixels being slightly more sensitive than others) and some degree of thermal interference (on some models, you can actually see which edge of the sensor die the support/interface circuitry is on on long exposures. This is also why the delightfully crazy people in astrophotography are commonly found peltier-cooling their sensors).
Under ideal photographic conditions, lots of light, modest exposure time, the signal largely swamps the noise. Underā¦less idealā¦ conditions, all the defects that are usually hidden come out to play.
3 Likes
Do digital camera manufacturers introduce dither noise, like the audio folk do to linearise low resolution errors?
1 Like
