Can ISO vary on different cameras?

Question

We all know that [ISO] + [shutter speed] + [aperture] = a great picture. Now, let's take two different (very!) cameras: Nikon D3100 vs Nikon D800 and set them on Aperture priority.

Settings for the same objective, same picture (let's say on tripod), different bodies:
ISO 100, F5.6

Q1. Will D800 give a shorter shutter speed than D3100?
Q2. Will the picture have the same amount of noise for both cameras?

If you can provide some examples, I'd be grateful. I'm interested to know if a better camera will make better pictures or it's all about skill and a great glass.

Update // I missed the fact that D800 is full-frame. I didn't want to consider any difference at this kind (this means a lot for the light and image). So, supposing both sensors are the same (full or crop) as I can see, I should have same settings, same shutter speed and a bit low-noise for D800.

Answer 1

Theoretically you will get the same shutter speed with both cameras. In practice it will vary slightly, due to differences in metering sensors and algorithms, as well as differences in actual f-stop, f/5.6 is never exactly f/5.6. Even if you use the same lens, you will have to zoom (otherwise you're getting a different picture with different content, thus the comparison doesn't make sense) which can change the actual f-number slightly even though the camera still says f/5.6.

Whether you get the same image brightness is a separate question, but again theoretically you will but in practice it can vary depending on the camera's actual (as opposed to reported) ISO sensitivity and RAW processing algorithm (and image profile settings). Within the same manufacturer the difference in sensitivity is probably not enough to worry about, however between brands it can be more (Canon used to be 1/3 of a stop more sensitive than Nikon).

Theoretically you will get less noise with the D800 - its larger sensor will collect 2.25 times as much light from the f/5.6 lens. More light means more photons (averaging out the photon noise) and a larger signal (which helps marginalise the electrical noise from the sensor). In practice there are many other factors which differ between sensors, pixel count, quantum efficiency, read noise, many of which go against the larger sensor. However the extra light collected is a significant advantage.

Answer 2

No, they will rarely be. If they are it would be purely coincidental. Considering how much time I've spent comparing cameras, I wish images would match but most times it simply does not. There are several factors at play:

• Metering: Each camera has a different metering system and it is not uncommon to see a full stop or more of difference, particularly when faced with a high-contrast scene.
• ISO: Theoretically the same ISO yields the same sensitivity but cameras can differ up to half a stop. Most times it is less but it still happens often enough that certain people measure the ISO-equivalence of cameras.
• Lenses: The lens aperture value accounts for the size of its opening but not how much light is transmitted. For critical applications, light transmission is measure din T-stop and some high-end cinema lenses are specified that way, instead of F-step. Of course you can mount the same lens on both cameras but to get the same light, they have to have same-sizes sensors.
• Tone-Curve: The way a camera interprets its data and forms in image from it varies between models and is even customizable with image parameters. This applies in-camera if you shoot JPEG or TIFF and applies in software if you shoot RAW. In either case, there is plenty of latitude to get different results here.

Now this means that the picture will be different but not necessarily better. I have seen cheaper cameras with better metering systems then more expensive ones and vice-versa. What you can expect from a better camera (larger-sensor, more-modern, etc) is better quality images.

Do not confuse a quality image with a great image. These are unrelated, the former requires quality gear, the latter requires an outstanding photographer.

Answer 3

1. The shutter speed will be exactly the same.
2. Noise levels will be different. Though with 100 ISO you would not see any difference with the eye. The differences in IQ will be more apparent when you put up the ISO. here you can see some sample images (Nikon D800, D3s and D7000):

http://nikonrumors.com/2012/04/10/nikon-d800-vs-d3s-and-d7000-comparison-by-cary-jordan.aspx/

D800 has a full frame sensor, D3100 has a cropped sensor. Basically the bigger the pixels, the better quality with high ISO will be.

If you would compare high ISO performance of D800 (36mp) vs D4 (16.2mp) you will see, that D4 produces less noise with the same ISO settings.

Here is a good link with ISO comparison:

http://www.kenrockwell.com/tech/comparisons/2012-04-05-High-ISO/index.htm#25600

Answer 4

Given the same ISO setting and aperture, the exposure will generally be same, for any camera, with any lens. The ISO setting and aperture value are designed to be consistent across equipment.

There may of course be minor differences between different lenses and cameras. Setting one camera to ISO 100 might actually make it equivalent to ISO 97, while in another camera it may be equivalent to ISO 104.

Also, different cameras may expose the same scene differently, depending on the exposure measurement method used.

The amount of noise is mainly a function of the ISO setting. At ISO 100 there is hardly any noise at all for any modern DSLR. At a higher ISO setting you may see a difference in noise, usually a camera in a more expensive segment gives less noise.

Answer 5

I think it'd be hard to beat the Wikipedia article on film speed for an understandable introduction to the topic. There's a section on ISO 12232:2006, which is the standard that applies to determining ISO-equivalent sensitivity levels for digital cameras. It turns out that there are several allowable techniques, ranging from Recommended Exposure Index, in which the manufacturer determines appropriate sensitivity levels for good exposure at various illumination levels, to Standard Output Sensitivity, in which a calibrated light source must produce a well defined result. Other (less used) techniques involve measuring sensor noise or saturation. Here's an excerpt:

The ISO standard ISO 12232:2006[55] gives digital still camera manufacturers a choice of five different techniques for determining the exposure index rating at each sensitivity setting provided by a particular camera model. Three of the techniques in ISO 12232:2006 are carried over from the 1998 version of the standard, while two new techniques allowing for measurement of JPEG output files are introduced from CIPA DC-004.[56] Depending on the technique selected, the exposure index rating can depend on the sensor sensitivity, the sensor noise, and the appearance of the resulting image. The standard specifies the measurement of light sensitivity of the entire digital camera system and not of individual components such as digital sensors, although Kodak has reported[57] using a variation to characterize the sensitivity of two of their sensors in 2001.

I think the best way to understand this is to expect that if an image is properly exposed at a given combination of shutter speed, aperture, and ISO, other cameras should also produce a pretty good image at those same settings. At the same time, realize that manufacturers have some wiggle room in this regard and expect minor variations between models and manufacturers.

I'm interested to know if a better camera will make better pictures or it's all about skill and a great glass.

That's a much different question. My own opinion is that skill trumps hardware -- if you know what you're doing, you can compensate for many hardware shortcomings. Check out this Strobist post in which David Hobby takes some decent photos with a Buzz Lightyear camera. That said, better hardware is usually called better because other things being equal, it takes better photos. Better sensors really do produce less noise at higher sensitivity levels.

Answer 6

Here is a DxO Mark comparison of the D3100 and D800. If you scroll down to the "View comparison for selected cameras" section and click the "Measurements" tab you can see the actual sensitivities at each ISO setting. When set at ISO 1600, for example, the D3100 is actually performing with a sensitivity of 1442 and the D800 is performing at a sensitivity of 1099. The D800 is almost 2/3 of a stop off! Many cameras are between 1/3 and 1/2 stop less sensitive than advertised. You can search the entire DxO sensor database for the results of any camera they have tested.

Unfortunately, DxO Mark does not test the varying sensitivities of each camera's metering system. It is the metering sensor and associated firmware that decides the shutter speed to use with your selection of ISO 100 and aperture of f/5.6. We should be able to assume that Nikon is aware of the actual ISO and has calibrated the exposure metering system to match it. If this is the case then some, but not all, scenes will vary by 1/3 stop between the D800 and D3100, since the DxO Mark database tells us at ISO 100 the D3100 is really ISO 89 and the D800 is ISO 74. This is just under 1/3 stop difference.

As others have already pointed out, nothing is exact. Shutter speeds vary a little from the number displayed, so do apertures.

For your specific question #1: If the metering systems are similarly calibrated, the D800 would give a shutter speed 1/3 stop longer to compensate for the lower sensitivity.

For your specific question #2: There are many factors that determine the noise level of a particular photo. But all other things being equal larger pixel size results in less noise for the same ISO with the same scene being photographed at the same settings. When we say Full Frame sensors have better high ISO noise performance than APS-C sensors it is because most FF cameras with a few more pixels than their APS-C counterparts still have much lower pixel density and thus each pixel can be larger. In the case of the D800 versus the D3100, however, the 36MP resolution of the D800 comes into play. The pixel size of the D3100, at 4.94 micrometers is actually slightly larger than the 4.7 micrometer pixel size of the D800! The sensor of the D800 is 2.25 times the area of the APS-C sensor of the D3100, but it also contains 2.5 times as many pixels. If all other things were equal, the D3100 would have slightly better noise performance at 100% cropped pixel-to-pixel comparisons. When both photos are displayed to the same size, however, the much higher resolution of the D800 allows for much more noise reduction to be performed without sacrificing visible detail.

Answer 7

There are two parts to the question here: better quality, and different exposure.

The better quality answer is simple: yes, there are significant differences. Cameras with larger sensors have an inherent advantage, but technology is advancing so quickly that it's often the case that technology generation is more important. Within sensors of the same size, it's definitely the case that different technology gives an edge in noise.

The other question is a bit harder, because while in basic theory it's the same, the standards for what exactly constitutes correct exposure are subjective. And, I mean that literally: the actual standards allow some latitude, and so you'll find that the brightness of the result given by the same combination of shutter speed, aperture, and ISO varies from brand to brand. (Maybe from camera to camera, as in your Nikon-to-Nikon example, but more tellingly between companies.) Dpreview used to measure this and provide a comparison to their own standard in their reviews. For example, this chart indicates that certain cameras from Pentax and Nikon match their ideal, where the Canon model was actually using a higher nominal value.

But, they don't do this anymore, probably because it's easily misinterpreted as a value-judgment about quality (or even honesty!) when it's really not. The same camera in that review often generated review complaints about a "tendency to underexpose": some people didn't like like that the chosen brightness was what DPreview calls exactly accurate in their chart.

But, having gone into some technical detail here, it's important to stress that these differences are very small, measured at less than a stop. Modern sensors have dynamic range such that this is unlikely to matter at all if you want to adjust the result slightly in post-processing. (Or, if you know that your camera tends to meter more darkly or more brightly than you like, remember that and dial in EV compensation.) In any case, unlike with ISO noise, it is not an indicator of "better pictures.