INT 16 conversion

[R13]

Hi,

We are working on some test data that we collected which are in .DAT format. When converting the .DAT to .I16 format we face a problem in the change in time. All our tests are conduted and data are stored at an interval of 0.01sec. But when we call the data in DIAdem and convert them to INT16 format, the time gets grouped together in the single common time (eg. for time 43.01, 43.02, 43,03... we get the I16 format data stored as 43.000012*). So we further have some problems in calling them again in other softwares as TIME data is most important to us.

So can u pls suggest me how to retain the data as original one even after converting to I16 format?

Thanks in advance.

Best Regards,

Ranga

[kikiduu]

Hi,

How can I16 show 43.01? Probably you need DBL or SGL. Are you storing the time channel also in the .DAT file? If yes in what format is it stored? Normally time in seconds is stored in U32 format. Or are u creating the time channel in DIAdem?

[R13]

Hi,

Wat i meant was tat the data stored in I16 format is of long integer value (but our data are in the increasing order of 0.01sec and we want to retain the same). Moreover I am very new to DIAdem and would to know how to create a time channel and the retain the same value (eg, 48.01, 48.02..... etc) in I16 format?

[Ingo_Schumacher]

Hi Ranga,

the problem you are describing is quite expected bahaviour. Converting a channel to I16 means that you are loosing resolution. The data format you choose has a resolution of just 2^16 steps, which will be used to express each time stamp between t0 and tmax. Internally, the time stamps are converted and stored to integer and when loading restored with a factor / offset scaling. 

You need to increase the resolution for the channel if you do not want to loose information, I'd recommend to store the data as r32 or even r64 which is the standard.

If the time channel is equidistant, that means if delta t always is 0.01, the channel could be stored implicitely, which will save the most harddisk space.

Best regards

[R13]

Hi Mr Schumacher,

Thanks a lot for your help. But i too tried with the other available formats and they do work well. But the problem is that the other software which we use to read the data from, reads only the *.I16 fromat data. So we need to be accurate with the data in this format, which we find it difficult.

I even tried converting from R32 and R64 formats to I16, but still encounter the same problem. And there is where we are struck with. Any suggestions from you on this highly appreciated....

Regards

Ranga

[Ingo_Schumacher]

Hi Ranga,

I have got 3 questions:

1. Which version of DIAdem are you using?
2. Can your timestamps be expressed implicitely, which means by the parameters Start time and delta time? For that, the time between two measured points always must be the same, which is quite often the case if the data was acquired continuously.  
3. Can the software you want to export the data to handle implicite channels of DIAdem .DAT and .i16 files?

If the file format i16 is not precise enough to handle your time channel you might have to try to exchange the data via ASCII. I don't know what you are doing with the data once it is imported to the other program. If you explain a bit more details I moght be able to advice how to use DIAdem to do the same, then you would not have to export.

Best regards,

[R13]

Hi,

Thanks for your suggestions. But we have found out the solution for this problem. We were able to capture the time in our software using the time channel that was actually generated by DIAdem and now we dont have a problem interpreting the time channel. Moreover we tried with the *.i32 format which was more precise.

Btw we use version DIAdem10.2

Thanks a lot for ur help and ur suggestions.

Ranga