# If you measure the O.D. of cells at one wavelength, is it possible to find its equivalent O.D. at another wavelength?

For example, I measured the optical density (O.D.) of some cells at wavelength 550 nm but I should have used 600 nm. Is there a way of converting my O.D. to its equivalent in another wavelength?

• It would be better if you explain what's OD? Commented Aug 1, 2017 at 5:09
• It would be useful if you could tell us what you were using your measurements of optical density for? Commented Aug 1, 2017 at 9:07

O.D. can mean 30 different things, many of them being related to science and to optics (see here)! I'll assume you mean absorbance (for Optical Density I'd guess) or light scattering as @canadianer said

Each substance has its own absorption spectrum and reflectance spectrum. Knowing only the absorption is a given wavelength does not tell much about the absorption at another wavelength (at least not without some other a priori knowledge about the substance studied).

• Your answer is essentially correct except that OD600 readings of cell cultures are a measure primarily of light scattering, not absorbance. Commented Aug 1, 2017 at 3:04

As already said by Remi.b, there is no way you can directly convert the O.D.550 to an O.D.600.

However, in case you already made numerous experiments using O.D.550, you may not have to do them all over again. You can try to draw a calibration curve of O.D.600 as a function of O.D.550:

1. take a cell culture (of the same cells you used for your experiments) and prepare several dilutions of it, so that their respective O.D.550 range from 0.1 to 1.
2. then, measure O.D.600 for each one of these dilutions.
3. then, draw O.D.600 as a function of O.D.550.

If, and only if, the points form a straight line, you can apply a linear regression to your data. You can then use this linear regression to convert your O.D.550 values to O.D.600 values. However, as pointed out be Jack Aidley in comment, it can be expected that the relationship is not linear. Provided that the curve is smooth enough, you can still calibrate to a curve.

Be careful:

• this kind of conversion is totally empirical and will only apply to cell cultures of the same species in the same conditions.
• this conversion only applies for values within the measured range of O.Ds.

Depending on what kind of experiment you were doing, it can be better to try this, and even replicate step one and two, or it can be better to start your experiments from scratch... Only you can tell

• 600 is traditionally used because it gives a more linear relationship between OD and cell count. I would expect that this would give a curve rather than a linear relationship; however, providing the curve is smooth and the conversions occurs within the measured range I do not thing calibrating to a curve is a problem Commented Aug 1, 2017 at 9:07
• @JackAidley Thank you for the feedback. I've edited my answer accordingly.
– Flo
Commented Aug 1, 2017 at 9:43