When exporting from DiVa, you have two options: "Export FCS" or "Export Experiment". FlowJo will read both formats. With "Export Experiment", the raw data is exported, and FlowJo will apply a slight transformation to all area parameters. This removes picket fencing and allows visually appealing display of events below the axis.

A minor transform is applied to the uncompensated data when users "export experiment" from the DiVa because the "export experiment" command generates data files that can have negative values for area parameters (not for height parameters). FlowJo uses a single automatic transformation for these parameters.

Thus, when you then compensate these parameters (i.e., those that are already transformed), the compensated parameters by default will get the same minor transformation. This might be enough transformation--but probably isn't. Therefore, you should still choose to create a custom transformation for your data manually.

For this purpose, you should select a fully-stained sample (that has fluorescence in all channels), gated to remove dead or unwanted cells (i.e., lymphocyte gate), and then choose "Define Transformation..." from the compensation menu. A custom transformation is computed based on the distribution of the data, which results in a better visualization.

The custom transformation will be applied to all samples that are compensated the same way in the workspace; hence it only need be computed once for each unique compensation matrix (and when computed for other matrices, requires a sample compensated with that matrix as an input to the customization).

The default parameters generally will work well for 18 bit acquisition where the lymphocytes are pinned to a top value of about 100 (10^2).

If the "negative" population is splitting in two or is too wide, then decrease the positive decades. Positive decades should be roughly the number of decades from the top of your autofluorescent lymphocytes to the top of the scale plus 1. So, if you put lymphocytes topping out at 10^2 (3.5 decades from the top), this value is typically 4.5. If the "negative" population is too compressed, then increase this value.