Abscissa with Time scale

[sebastienvullioud78]

Hi everybody,

I just have a question relating to the default setting of X_min (in my case, value 1614412800000) / ΔX Step (value 7992000) & X_max (value 1614812400000) when working with the Algorithm "X Step w/Interpolation" and a timescale. What do the "Units" correspond to? I have a graph of a drying cycle of a wood kiln. I know that the cycle started on 02/27/2021 at 7:42 a.m. and ended on 03/05/2021 at 11:39 a.m. I would need to get the differents y variables with a fixed interval of 1 minute to be synchronized with other data set.

Thanks in advance!

[nbehrnd]

From the program's GUI running the averaging window, and from the input format of the images submitted (bitmap .png or .jpg) I assume the digitizer works in pixel units. At latest when it comes to report the work as a .csv, there is a transformation into the floats one may see, sort, and format in precision.

If you do not have access to the raw data of this control chart (Kammer_3.Dat could be a hint to a raw data file, somewhere in the recording device), then the .png shared here is of too low resolution to extract the data of interest in a one per minute interval. The file spans over 932x581 pixels only. Reading the .png with inkscape with the setup of Image DPI from file, I accessed the time bar in the top left corner of the image. Using inkscape's measurement tool (simply type M) activates a ruler (if there only two points are selected) and records a horizontal distance between the start and the end of this reference bar of 56px only. Maybe I did not got these termini right, but the very least expectation of 9 x 60 x 2 = 1080 px for 9 h of 60 min differs by a factor of about 20 from this finding. (The factor of two results from the Nyquist theorem to prevent undersampling I would add here.)

Maybe lost in translation; check the time increment used. The Unix / epoch time stamp supports time stamps down to nanoseconds. If applied to start and end of your recording as well about "ΔX Step (value 7992000)", the later would correspond to multiple recordings per second. The intended resolution of 1/min you mention earlier is much more reasonable to determine the time constant of the thermal response of the kiln.

digitizer_test.zip