Optical drawbacks

Optical methods track light scattering, not cells. Four studies show where that breaks down, and how BactoBox® compares.

Optical methods measure how much light your culture blocks. Optical density (OD) and backlight scattering are the two common ones. Both are often used as proxies for cell concentration. You run them on-line as a sensor, or diluted off-line in a cuvette.

They are fast, cheap, and already in your lab. They also share one blind spot: light scattering responds to much more than the number of cells. In this section we will focus on optical density at 600 nm (OD₆₀₀) because it is the predominant light scattering proxy method for cell concentrations.

OD depends on cell size

Cell size shows the problem most clearly. Cells shrink as nutrients run out. The same number of cells therefore scatters less light in stationary phase than in early-exponential phase.

Optical methods are strongly dependent on the size of objects. The two samples scatter the same amount of light, but the bacteria in exponential stage (left) are considerably bigger than the cells in stationary stage (right). Same OD but different cell concentrations.

OD is affected by pigments and particulates

Optical methods like OD can't tell a cell from a bead. Less light reaching the detector may be due to scattering from bacteria, but it may also be due to scattering from non-bacterial objects or absorption by pigments.

Optical methods like OD can't tell if a signal is due to cells, particulates or color absorption. Left: A bacterial culture. Right: Molasses containing black pigments and non-bacterial particulates

Four studies demonstrate optical shortcomings

We ran four studies to put numbers on this. Each one compares OD₆₀₀ against BactoBox®, plate counts, or both. The table sums up what we found. The four pages below hold the study designs, the data, and the implications. Click the links for to see the results in the subsections.

BactoBox® compared to optical methods

Question	Optical methods (OD600)	BactoBox®
What is the output?	absorbance units (AU)	cells/mL
Does it distinguish cells from other objects?	No Every object that scatters or absorbs light adds to the signal.	Yes It classifies each object, then reports cells/mL apart from total/mL.
What is the lower limit of quantification (LLOQ)?	≈2×10⁸ cells/mL	3×10⁶ cells/mL at 1:100 dilution. 3×10⁵ cells/mL at 1:10 dilution.
Is a calibration factor needed?	Yes One per strain, medium, and growth phase.	No
Is the calibration factor stable?	No Cells-per-OD moved 11.4× across a single growth curve.	N/A The ratio to plate counts held at geomean of 1.01, span 2.2×.
Does cell size skew the result?	Yes Cells-per-OD scales with size of the objects.	No It counts one object at a time, and reports CIZE separately.
Usable in lag and early-exponential phase?	No The first nine readings of an 11.6 h run sat at 0.00 ± 0.01 AU.	Yes It resolves a 1×10⁶ cells/mL seed.
Does the method agree with plate counts when screening growth media?	No (R² = 0.12).	Yes (R² = 0.83).
Does it rank growth media correctly?	No OD600 called one glycerol concentration uniquely best. Direct cell counts and CFUs disagreed.	Yes It tracked plate counts. Better precision gives higher statistical power than CFUs.
Is it available as online method?	Yes	No
Is it available for high-throughput screening	Yes	No

Optical methods still earn their place

Use OD₆₀₀ or backlight scattering for on-line monitoring and high-throughput screening. Both are fast and cheap. Then refine your shortlist with direct cell counts.

For the individual studies, see the following subpages

• OD is inflated by background

• OD misses low cell concentrations

• OD calibration factors drift

• OD confounds medium optimization