Part 1 Principles
1. Fluorescence microscope
2. Filterset in FL-Mic
3. How concocal differs?
4. What is confocal?
5. Resolution in confocal
6. Optical sectioning
7. Confocal image formation
    and time resolution
8. SNR in confocal
9. Variations of confocal
      microscope
10. Special features from
     Leica sp2 confocal
Part 2 Application
1. Introduction
2. Tomographic view
    (Microscopical CT)
3. Three-D reconstruction
4. Thick specimen
5. Physiological study
6. Fluorescence detecting
       General consideration
       Multi-channel detecting
       Background  correction
       Cross-talk correction
            Cross excitation
            Cross emission
            Unwanted FRET

Part 3 Operation and
             Optimization
 1. Getting started
 2. Settings in detail
      Laser line selection
      Laser intensity and 
         AOTF control
      Beam splitter
      PMT gain and offset 
      Scan speed
      Scan format, Zoom
        and Resolution
     Frame average, and
         Frame accumulation
     Pinhole and Z-resolution
     Emission collecting rang
        and Sequential scan

When Do you need confocal?
FAQ
Are you abusing confocal?

Confocal Microscopy tutorial

Part 3 operation, optimization of Leica SP2 LSCM

Average and Accumulation

As discussed above, using Gain and offset to balance the signal intensity and background noise level is often insufficient. And using high settings of some other parameters such as scan speed, scan format, pinhole reduces signal influx and affect SNR and image quality adversely. This side effect can not be compensated by Gain and offset of PMT. In these cased, average is a good option to reduce noise, increase pixel time and compensates the lowered SNR, thus improve image quality.

Average takes the sum of pixel values from the specified number of scan and use arithmetic mean as the final value in the image. Average preserves those persistent pixel values that are mostly specific signal while divide away along times those fluctuated values that are mostly noise; Average also extends the time of scan thus extends the pixel time and improve SNR. So, average reduces noise and improve image quality significantly. Average does not suffer from pixel saturation since the it divides the sum into mean.

Average has two forms, line average and frame average, both have same effect but are executed differently. Line average takes specified number of scan first on the same line then go to next line, repeat for all lines in a frame and finally produces the frame. Frame average takes the whole frame first then repeat scan the specified number of frames, each frame is added consecutively, the result is shown immediately after each addition. Frame average is used more often in frame based acquisition since user can see the image quality improve progressively after each frame is acquired.
Usually, four to eight times of average is enough for scan format 512 or 1024 at speed 400 Hz. But if higher format, higher speed is used, that means there are smaller pixel size, less pixel time, and less photon influx per pixel, thus less SNR,  then more times of average is needed. For 2048 or 4096 format, 800 Hz or 1000 Hz speed, 16 or 32 average has to be used to compensate the loss of SNR, that in turn bring severe bleaching. (refer SNR and image generation in confocal).

Frame accumulation also takes specified number of scan and adds up the pixel values, but it use the sum as final value in the image. Obviously, accumulation increase signal intensity and extends pixel time thus improve SNR. But accumulation adds both signal and noise thus does not have the effect of preserving persistent signal while canceling signal with varied intensity. Besides, accumulation suffer from saturation on those high intensity value pixels.

In extreme difficult case, line average and frame accumulation can be combined to reduce noise, increase signal intensity, increase SNR and improve image quality. (frame average and accumulation can not be used at the same time because they reject mutually ).