Three Quick Questions vs The Enjoyment of Photography

Started Jan 2, 2013 | Discussions thread
Great Bustard
Forum ProPosts: 23,567
Like?
Re: Three Quick Questions vs The Enjoyment of Photography
In reply to grumpyolderman, Jan 2, 2013

grumpyolderman wrote:

Ironically, Equivalence works against larger formats, as the noise advantage for larger sensor systems disappears for Equivalent photos.

GB, enlighten me here, same scene, same metering, FF show less noise (maybe not if the sensor is 7 years old) in real pictures, why is this then? Assuming same number of pixels per sensor, the larger pixels will get more light per pixel at same exposure settings, making it easier for the sensor to generate a final picture with less noise, I would say latest sensors compared FF still has about a 2 f-stop advantage in noise alone, and some in DR.

Equivalent photos have the same perspective, framing, DOF, and shutter speed.  The consequence of this is that the same total amount of light is projected on the sensor.  Thus, if the sensors are equally efficient, the noise will be the same.

However, often there is enough light so that the FF photographer can achieve the same DOF with a lower shutter speed, thus putting much more light on the sensor, and thus less noise.  Alternatively, the FF photographer can use a more shallow DOF for a given shutter speed, if the more shallow DOF is desirable, or, at least, more desirable than the higher noise.  The consequence of the more shallow DOF for a given shutter speed once again projects more total light on the sensor, and thus less noise.

So, if you see FF photos having a noise advantage, it is because the FF photographer is taking photos that are not Equivalent.

Conversely, if the mFT camera has sensor IS, or is using an IS lens and the FF photographer does not have an IS lens (e.g. EM5 + 45 / 1.8 vs 5D3 + 85 / 1.8), then, in low light, the mFT system can use a lower shutter speed for a given DOF (so long as motion blur is not an issue or is desirable for artistic effect), and it will be the mFT system that has less noise.

The effect of pixel count on noise is a secondary, not primary, effect.  The primary causes of noise in the photo are:

  1. The total amount of light falling on the sensor (Total Light = Exposure x Effective Sensor Area)
  2. The amount of that light that is recorded (Total Light Recorded = Total Light x QE)
  3. The additional noise added by the sensor and supporting hardware (read noise)

For Equivlalent photos (same DOF and shutter speed), the total amount of light falling on the sensor is the same.  For equally efficient sensors, the QE and read noise are the same.

Where pixel size comes into this is read noise and NR (noise reduction).  If pixels had the same read noise regardless of size, then more pixels would result in more read noise per area, and thus more total noise.  However, smaller pixels tend to have less noise, as a general rule, so, this reduces the noise penalty considerably, and, in some cases, reverses it.

More pixels also mean more recorded detail, which means that additional detail can be traded for less noise via NR software.  In an interesting corollary, this means that sharper lenses can result in less noise.

Hopefully, that answers your questions without being overly technical.  By the way, I have to ask:  older than what? 

Reply   Reply with quote   Complain
Post (hide subjects)Posted by
Keyboard shortcuts:
FForum PPrevious NNext WNext unread UUpvote SSubscribe RReply QQuote BBookmark post MMy threads
Color scheme? Blue / Yellow