The Leica M maxes out at 6400.
That was what I meant in my first paragraph. I had a gig in the early '90s showing a local paper how they could use Photoshop to salvage previously âunusableâ news photos, and they were amazed at what could be made printable. I was amazed by what they considered printable.
What I am wondering is what the response curve looks like in this mode and how much it emphasizes or de-emphasizes particular color ranges. And how much IR it responds to.
Iâm looking forward to seeing what a full review at dpreview.com yields too. Or maybe renting one to do some tests myself.
Not for handheld shots, no.
In the late 1970âs I interviewed for a student job with a group that was working on the APâs first âelectronic darkroomâ â a computer system that would grab wirephotos directly into storage, do some very basic manipulation on them, and go to print from there. A lot of development in that next decadeâŚ
Newspaper halftones are about 85 LPI, with glossy magazines ranging up to 300 LPI. To effectively use the dithering, the source image has to be about 1.5 to 2x that resolution, but anything more is mostly wasted. (Good discussion at Pixel density - Wikipedia). So a 1-column newspaper image is pretty much VGA resolution before losing detail to convert greyscale into halftones.
Old proverb: âThe amazing thing is not how well the bear dances, but that it dances at all.â
4 Likes
And yet costs eight thousand dollars.
Iâd classify Monster Cable as more overpriced for what you get because you need special hearing to really hear itâŚ
Leicas are nice and all, but they launched a $25,000 digital camera that only does monochrome. Thatâs overpriced for what you get because you need special vision to really see it.
Funny. B&H says that the Leica M Monochrom is 8 thousand.The Leica S is 22 thousand, but thatâs a Medium Format camera. Of course, once you have your $8000 camera, you need to get a couple of $8000 lenses, soâŚ
1 Like
Or you could strap a 400mm f/2.8 onto a Nikon V3 body for rather less and get ISO 12800âŚ
Yes, you loose two stops on the ISO, but you gain three on the lens so the V3 has a slight edge.
Only thats not really how it works. For a given sensor there is a point where the response is limited by the electronics of the A/D converter etc. Thats used to be about ISO 1600 on the D3 but Iâll believe its a bit higher on the D4s.
If you go above that point, the sensor becomes the limiting factor and each time you pump up the gain on the amplifiers by a stop you lose a stop of dynamic range. So you actually end up capturing less information rather than more.
So what it comes down to is the dynamic range of these cameras is what is really improving. So you could take the picture at 1600 and boost the shadows to see what is there or crank the ISO gain to silly levels, the results will be pretty similar.
The D4s is probably the better sensor but you would have to take other measurements to tell.
Itâs dropped in price in the past two years.
That really does depend on whether the Nikon 1v3 is capable of producing noise free images at ISO 12800 or 6400 or whatever. Iâm not sure that it is.
1 Like
Then lets look at some figures:
The DxOMark scores are adjusted for resolution but not for sensor size. Since we are comparing two lenses of the same field of view we need to compensate, by a factor of 2.6^2 (6.76)
The ISO score for the D4s is 3074
The V3 scores 384, so with compensation its 2595
However compensating the f/2.8 on the V3 gives an equivalent aperture of v/7.25 which is also a smidge worse than the f/7 that the f/5.6 lens gives with the 1.25x on it.
So the D3s is a little bit better but the real reason to go for it would be the DSLR performance on autofocus and the better dynamic range of the D3s sensor. The Aptina sensor on the V3 is better tuned for video but it only has 12 bits rather than 14.
compensate? For what?
Unlike the two previous scenarios in which light is either generous (studio) or stability is assured (landscape), photojournalists and action photographers often struggle with low available light and high motion. Achieving usable image quality is often difficult when pushing ISO.
When shooting a moving scene such as a sports event, action photographersâ primary objective is to freeze the motion, giving priority to short exposure time. To compensate for the lack of exposure, they have to increase the ISO setting, which means the SNR will decrease. How far can they go while keeping decent quality? Our low-light ISO metric will tell them.
The SNR indicates how much noise is present in an image compared to the actual information (signal). The higher the SNR value, the better the image looks, because details arenât drowned by noise. SNR strength is given in dB, which is a logarithmic scale: an increase of 6 dB corresponds to doubling the SNR, which equates to half the noise for the same signal.
An SNR value of 30dB means excellent image quality. Thus low-light ISO is the highest ISO setting for a camera that allows it to achieve an SNR of 30dB while keeping a good dynamic range of 9 EVs and a color depth of 18bits.
A difference in low-light ISO of 25% represents 1/3 EV and is only slightly noticeable.
As cameras improve, low-light ISO will continuously increase, making this scale open.
Nothing in there about compensating for sensor size.
Moreover, this essay, says this kind of compensation doesnât work in the real world.
Another surprise is that the smallest sensors manage to outperform the blue D3s scaling line. The orange scaling line shows that the diminutive Pentax Q is currently best at high-ISO if you assign a handicap for sensor size. This doesnât mean that the Pentax Q has very low noise. On the contrary: it needs to be operated at 200 ISO to get the same print quality as the D3s at 3200 ISO. But in view of the size handicap, various of the smaller cameras do a surprisingly good job[xvii].
2 Likes
For everyone asking why you who would need extreme sensitivity to freeze action in the dead of night, even if it has piles of digital grain:
Itâs likely you read boingboing for articles with pictures of people dancing in the moon at burning man, standing at the barricades in the night in the Arab Spring, or testing some new contraption in a dark warehouse. Advances in high-ISO capabilities let you see more moments like that, or see them more clearly. The important news here isnât that the extreme maximum got higher, its that when the extreme max got higher the usable max also got better, making it easier for everyone to show you something special.
7 Likes
His point was that Leica charges extremely high prices for objectively pedestrian performance specs, much like Monster Cable (although an emphasis on quality may be why they donât offer high or boosted ISOs).
And if high ISO = Monster Cable, then pretty much every other DSLR manufacturer is in an arms race to be crowned Monster Cable.[quote=âjerwin, post:72, topic:27189â]
compensate? For what?
[/quote]
Donât disagree with him: he has an Oxbridge PhD in nuclear physics, and this makes him an authority in digital cameras.
2 Likes
Buy now, find out later, Jason, itâll catch the afterglow.
ISO equivalence for digital cameras is a guide to setting exposure. Comparing digital cameras with different sized sensors raises a problem. The field of view changes so from full to crop sensor so there are two cases
- We use the same lens on two different cameras ignoring the difference in field of view
- We compare lenses with the same field of view.
The first is the most useful for doing photography because we have a bag of lenses and it is inconvenient to think about the f/2.8 becoming a f/4 when we put it on the crop sensor. But that is exactly what the effect is on both exposure and depth of field.
Exposure = [subject] * ISO * Shutter / focal-ratio
Since we want the bodies to perform the same and donât want to adjust the focal-ratio, we adjust the ISO equivalence instead.
So to compare the performance of two sensors with a crop factor of x, we multiply the ISO and the focal ratio by x^2 and work out the exposure for the same subject.
Anybody know (or know how to calculate) the effective âISOâ (obviously each one is only a single pixel, so the comparison is nonsense) of the photomultiplier tubes used in Cerenkov-type neutrino detectors? I think that some of those are actually capable of detecting single photon events, which blows my mind a bit.
2 Likes
Well, in the first case itâs really hard to compare the images because there are two different fields of view. So comparing the depth of field makes little sense because of the differences in field of view. And an f/2.8 doesnât become an f/4 when you put it on a crop sensor: the f-stop is a physical property of the lens and has nothing to do with the sensor.
What the hell is this formula supposed to mean? What is â[subject]â?
And why are you adjusting the âISO equivalenceâ?
If a noonday scene is 1/100 at f/16 and ISO 100 on an 8x10 view camera, itâs also 1/100 at f/16 and ISO 100 on a 35mm FF camera, ad the same again on a DX camera. The focal length and field of view are irrelevant. Or take a look at a hand-held light meter. It will give shutter and aperture readings, and all the user has to do is set the ISO. The user does not have to specify if she is shooting 35mm or MF or LF, nor does she have to specify the focal length she is using.
As far as the amount of light that reaches the sensor and the depth of field goes, that is the effect.
If you have a 100mm f/1.4 and put it on a 1.5x crop sensor the effective focal length becomes 200mm. But the depth of field that you get is equivalent to a f/2. And you are getting 1/2.25th of the light hitting the sensor.
On a bellows or using a view camera the the distance between the lens and the focal plane can be changed. And when you move the sensor out you have to compensate for the fact that you are spreading the light out more.
If you want to compare like with like you need to compensate just like you would when using a teleconverter.
100mm/1.4 (if such a lens exists) on a DX format does not become an âeffective focal lengthâ of 200mm. The focal length remains 100mm because this is a physical property of the lens. The field of view has the FF equivalent of 150mm. The depth of field, assuming the âprintâ size is the same, is the same as it would be on a 150/2.0 on a FF sensor, sure. And itâs also the same as doing a 66% crop on the FF 100/1.4 image.
Sure. But thatâs not how ISO is measured. ISO is the film/sensor sensitivity (per unit area). A swimming pool may catch more rainfall than a bucket, but that doesnât mean it was raining harder over the swimming pool than it was over the bucket.
You might have an actual point if you actually changed the distance between lens and film when switching between 4x5 and 8x10 film. But you donât. A 300mm lens, by definition, is a lens that focuses at infinity when the focal node (of the simple-lens equivalent) is 300mm away from the film plane. So if you have a 300mm lens on a large format camera, it doesnât matter if you use an 8x10 film holder or a 4x5 film holder: the lens is always the same distance from the film plane (300mm away, if focused at infinity). Similarly, a DX format sensor is just as far from the lens as a FF format sensor is.
This is wrong. A teleconverter changes the actual focal length of a lens, but does not change the physical aperture size. With a physically longer focal length but same physical aperture size, the ratio of focal length:aperture changes, which is why the f-stop changes. For example, a 100/2.0 lens has a focal length of 100mm and a physical aperture of 50mm, for a f-stop of 100/50 or f/2.0. Slap on a 1.4x teleconverter and the focal length becomes 140mm but the aperture remains 50mm, for an f-stop of f/2.8. This represents a very real reduction in the intensity of light transmitted by the lens to the film. In effect, it is like trying to fill the swimming pool with the bucket.
Changing film/sensor formats, on the other hand, does not change the focal length. It does not change how far away the sensor is from the lens. It does not change the lensâs f-stop. It does not change the ISO.
1 Like
Iâm just confused as to why the example photographs used on the Amazon page are so incredibly crappy.
Horrible noise and pixelation, and thatâs just (supposedly) at 3200.
