No, the recording interval is fixed at 1 minute. There is not much value in a shorter recording interval. Given the limitations on accuracy from detecting whether you are actually using the computer, trying to provide accuracy of less than a minute is misleading at best. There are almost 500 minutes in an 8 hour day - trying to track time down to the level of 20-30 seconds is not very useful.
The way that TimeSprite samples means that over a longer time, the results are reasonably accurate. If you switch to a particular window for only 15 seconds at a time, most times TimeSprite will not detect it. If you do it often however, probability means that TimeSprite will detect it often enough to give approximately the right answer. If you switch to a window for 15 seconds 100 times, you spent 1500 seconds or 25 minutes. TimeSprite will detect it about 1 in 4 times, or 25 out of 100 so TimeSprite reports about 25 minutes.
This question usually arises because people want to see results after only running the program for a few minutes. You need to run the program for a few hours at least before judging the results - a whole day is even better.
It is not possible to give a definite value for TimeSprite's accuracy. There are 2 major factors that influence the accuracy:
Which of these factors are more significant depends on your personal usage habits.
Despite this, TimeSprite is probably more accurate than times most people record manually. If you are one of the diligent people who do get better accuracy manually, you can define your own activities instead of monitoring the active window. This means that you can explicitly switch activities, and the records will have similar accuracy to the manual process. You can then use TimeSprite to view reports by day, week, month etc.
Entries can be grouped in 3 different ways:
More specific grouping overrides less specific grouping:
This is caused by the way that TimeSprite does its automatic grouping, and tends to occur when grouping a small number of windows or a short period of time.
TimeSprite only creates a group automatically when there is more than one window to go in the group. For example, if there are entries for "Microsoft Word - Document 1" and "Microsoft Word - Document 2" these will be grouped as "Microsoft Word". If you only have "Microsoft Word - Document 1", there is only 1 window so a "Microsoft Word" group will not be created. The window will be listed under its full title.
Then TimeSprite tries to find groups for leftover windows. This is the process that groups "Save As" and other miscellaneous dialogs. These will be grouped with other windows from the same application.
TimeSprite's automatic grouping works by finding similarities in the window titles, and trying to work out the best names for groups. One of the consequences is that in general, the more times and the more titles it has to work with, the better the grouping. It will occasionally show odd results for small amounts of time, when it doesn't have enough information for sensible grouping. For example, if you had 1 minute in "Microsoft Word - Document 1" and 2 minutes in the Word "Save As" dialog, they would be grouped under "Save As". Usually this isn't a problem as you don't normally spend much time in these dialogs.
When recording time against projects, you should assign time to the project groups in the Journal. When you assign blocks of time to groups, they are not affected by other grouping changes. Don't use the Time window to group the window titles into projects unless you have very specific window titles, because they will change if you change the window grouping in the future.
However, you can use the Time window to do preliminary grouping to make grouping by time easier. For example, you might group all Visual Studio windows as "Development", then you can assign the Development entries in the Journal to a specific project.
There are a couple of possibilities:
TimeSprite by default tries to split groups into subgroups. This means that it can show how much time you spent on particular websites, particular documents in an application etc. However, this can generate a lot of different groups - particularly for things like web browsing.
Changing to "Group" means that only the first level of grouping will be performed, which in most cases means times will be grouped by application.
If you want to use subgroups for most applications but have others that generate too many groups, you can use wildcards to group the problem application.
Using Internet Explorer as an example, right click on an Internet Explorer entry and group it as "Internet Explorer" (or whatever name you prefer). Go into Edit->Groups and edit the Internet Explorer entry you created. Change the entry under the group to read "*Internet Explorer" and all windows which end in "Internet Explorer" will be assigned to this group.
You can then move entries from this new group into other groups if necessary.