Re : Re: Re : Re: Re : Re: Converting 32bit Array to Byte array

[jaffer]

Hello Everybody,

We are using moveblock function to get the pointer value in that we are getting 3,07,200   (32)bit values but we want the format of 8 Byte (12,28,800) values .How to convert this.

[TCPlomp]

The easiest is to use a Typecast function with an Array of U8 as Type:

 

This will instruct LabVIEW to interpret the data as an array of U8.

Ton

[jaffer]

Hi

In this i can able to get only 3,07200  Byte values only but i want 12,28,800 Byte values

[TCPlomp]

That would surprise me very much, but what code do you have?

Here is a little code snippet of what I have in mind and which shows that the number of elements has changed by a factor of 4.

Ton

Message Edited by TonP on 05-12-2009 08:25 AM

[JB]

Another, less efficient, solution.

PS : Be kind with me and don't mention it in the Rube Goldberg Code forum !

[TCPlomp]

Hi JB,

I have no idea what is inside your VI but it maxes LabVIEW CPU's usage to 100% for several minutes before I decide to kill LabVIEW.

Could you attach a screenshot?

Ton

[altenbach]

No problem here. It autoindexes the U32 array on a FOR loop, inside the loop, it  uses split number twice to get the four bytes, then uses built array to built the output array into a shift register. It's about 10x slower than the typecast. So why do it?

[JB]

Like mentionned in my original post, I expected it to be slower then the typecast but I admit to be astonished by the 10 times factor !

 

Message Edité par JB le 05-12-2009 02:37 PM

[craigc]

You could also get rid of that second for loop and just use the interleave 1D array function 🙂

[JB]

The performance ratio is even higher than 10 (about 12 on my machine) ! Obviously, the performance would be quite better with Initialize Array and Replace Array Subset but probably still far away from the typecast ! Don't do it this way...