Experiment with high ISO on a D70

[binah06]

Hi Everyone!
I have read several posts where people are talking about how noisy the higher ISOs are on Nikon D90 and below. I generally dont have any noise issue with my D70 but I rarely shoot at that high of an ISO. So I decided to experiment. Here is a pic of my husband's falcon and I included the EXIF data. The large version in on my flickr page.


Camera: Nikon D70
Exposure: 0.008 sec (1/125)
ISO: 1000
Aperture: f/4.5
Focal Length: 70 mm
Focal Length: 71.3 mm
Exposure Bias: 0 EV
Flash: No Flash

On the largest version I can see some noise but I dont know how it would print up. Might print one next time just to see.

Like I said I dont normally use high ISOs but I thought it would be fun. I would love to know what the pros think about the noise level.
Melissa

[binah06]

I forgot to mention that this is straight out of the camera. It never saw photoshop in its little life

[OsmosisStudios]

At that size, you likely wont see too much noise. Getting close up to it, however, and you will.

Not only that, but the D70 is only 6mp and, like the D40, has the advantage here much like the D3 does over the D300. Having a lower MP count on the same sized sensor means more light to each sensor. The D90 has twice the megapixels (sites) and thus has half the light to each.

[binah06]

Thanks OsmosisStudios. I noticed that I have settings so that you cant see the larger version, so I changed it on this one to creative commons. Its good to give away images every once in a while

Ok I am trying to follow you here. I am not very good with digital terminology. I first learned photography on film SLRs. So this means that the D70 uses light more effectively the the D90?

Thanks
Melissa

[OsmosisStudios]

The size of the sensor of the D70 is the same as the D90: 18x24mm.

A D70 is a 6mp camera, which means it has 6 million photosites on the sensor
A D90 is a 12mp camera, which means it has 12 million photosites on the sensor.

Each photosite = 1pixel on the image.

The amount of light hitting a sensor during an exposure doesn't change. As such, if you have a sensor with twice as many photosites, each site gets half as much light.

You end up with a larger image, but at the expense of noise at higher ISOs. This is because, as you increase the ISO, the camera boosts the signal coming from the photosites artificially. If a sensor returns a value of 1 at 100iso, and you have the camera set to 400, it'll return a value of 4. This is great, but it means that there is alot more noise.

This is why the D3 trumps the D300 in high-iso noise. They're both 12mp, but the sensor on the D3 is twice the size of the D300. As such, each of the 12 million photosites gets twice as much light. This means that the sensor in a D3 would return a value of 2 at 100iso whereas a D300 would return a value of 1 at 100iso.

In your example, your D70 has half the megapixels over the same space as a D90, so it naturally follows the same principle. The downside to that is that you end up with a smaller picture. That's fine for most work, actually, but it means that you have a tougher time making MASSIVE prints (if you do) and it means you cant downsample for sharpness (re-size to a smaller image size).

Hope that clears things up

[whiggy]

Also, this is an interesting article. I wonder how much of this applies in real life.

More pixels offsets noise!

[vandergus]

Great article whiggy. I was just about to chime in with something similar. What Osmosis said is true but only if your examining the images at per pixel level. This isn't how we're usually viewing pictures, though. When comparing noise, I want to know how 8X10 prints from two different cameras compare or how they look at a particular screen resolution. When you're looking at the per pixel level, it's like comparing an 8X10 print to a 12X15 print, not really fair. When images are normalized to a standard viewing size, the advantages of the low pixel count sensor are lost.

Of course comparing the D3 to the D300 introduces differences in format so there are other things to consider in that case.

[binah06]

I think I got it. I am going to read it all again tomorrow when I am more awake. Thank you!

[kencaleno]

Originally Posted by binah06

Hi Everyone!
I have read several posts where people are talking about how noisy the higher ISOs are on Nikon D90 and below. I generally dont have any noise issue with my D70 but I rarely shoot at that high of an ISO. So I decided to experiment. Here is a pic of my husband's falcon and I included the EXIF data. The large version in on my flickr page.


Camera: Nikon D70
Exposure: 0.008 sec (1/125)
ISO: 1000
Aperture: f/4.5
Focal Length: 70 mm
Focal Length: 71.3 mm
Exposure Bias: 0 EV
Flash: No Flash

On the largest version I can see some noise but I dont know how it would print up. Might print one next time just to see.

Like I said I dont normally use high ISOs but I thought it would be fun. I would love to know what the pros think about the noise level.
Melissa

As long as you have enough light,and you expose correctly, noise isn't too much of a problem.Hope you don't mind,I just brough the other eye visible. Ken