在 Javascript 中读取带符号的 16 位数据

Question

I have been banging my head to solve this:

I received a raw data from an embeded device. From the documentation, the way to read it into a single value is:

Every two bytes of data can be combined to a single raw wave value.Its value is a signed 16-bit integer that ranges from -32768 to 32767. The first byte of the Value represents the high-order byte of the twos-compliment value, while the second byte represents the low-order byte. To reconstruct the full raw wave value, simply shift the first byte left by 8 bits, and bitwise-or with the second byte.

short raw = (Value[0]<<8) | Value[1];

One of the 2 bytes that I received is "ef". When I used the bitwise operation above the result does not seems right as I noticed I never get a single negative value (its ECG data, negative values are normal). I believe using Javascript to do this is not straight forward. The way I did it was like this:

var raw = "ef"; // just to show one. Actual one is an array of this 2 bytes but in string.
var value = raw.charAt(0) << 8 | raw.charAt(1)

Please Advice. Thanks!

EDIT:

I also did like this:

let first = new Int8Array(len); // len is the length of the raw data array
let second = new Int8Array(len);
let values = new Int16Array(len) // to hold the converted value

for(var i=0; i<len ; i++)
{
   //arr is the array that contains the every two "characters"
   first[i] = arr[i].charAt(0);
   second[i] = arr[i].charAt(1);
   values[i] = first[i] << 8 | second[i];  
}

But still all is positive result. no negative. Can someone verify if I am doing this correctly, just in case maybe the values are actually all positive :p

Answer 1

It's two's complement: Check the top bit of the high byte - byte[high]>>7. If it's 0, do byte[top]<<8 | byte[low]. If it is one, do -((byte[top]^0xff)<<8 | byte[low]^0xff) - 1. See https://en.wikipedia.org/wiki/Two%27s_complement for an explanation.

Also check out https://developer.mozilla.org/en-US/docs/Web/JavaScript/Typed_arrays . It has Int16 arrays which is what you want. It might be a ton faster.

Answer 2

You can use property that the string is already 16 bit and then make it signed.
Also instead of reading 8bit at time just read one unsigned 16bit using charCodeAt.

var raw = "\u00ef"; //original example
var buf = new Int16Array(1);
buf[0] = raw.charCodeAt(0); //now in the buf[0] is typed 16 bit integer
//returns 239, for \uffef returns -17

 var raw = "\￯"; //original example var buf = new Int16Array(1); buf[0] = raw.charCodeAt(0); //now in the buf[0] is typed 16 bit integer console.log(buf[0])

Answer 3

[EDIT]

Use Int8Array to store byte data:

function str2ua(str) {
    var buf = new Int8Array(str.length);
    for (var i = 0, strLen = str.length; i < strLen; i++) {
        buf[i] = str.charCodeAt(i);
    }
    return buf;
}

var raw = 'ef';
var buf = str2ua(raw);
var val = buf[0] << 8 | buf[1];
console.log(val);  // output 25958

raw = '¢f';
buf = str2ua(raw);
val = buf[0] << 8 | buf[1];
console.log(val);  // output -23962

Answer 4

For first byte take two's complement first then shift by 8 bit

let x = raw.charCodeAt(0); //ASCII value of first character

then flip x for 1's complement and add +1 for 2's complement and finally do

var value = x << 8 | bytevalueof(raw.charCodeAt(1))

Answer 5

This is a question about the raw wave data coming from a Neurosky Mindwave Mobile EEG headset .

You should find three values in the buffer when you read from the device. Perform this operation on the second two to get a correct reading:

    var b = reader.buffer(3);
    var raw = b[1]*256 + b[2];
    if(raw >= 32768) {
      raw = raw - 65536;
    }