IRremote library sening RAW problem

Hello,

My problem is, how the arduino sending only FFFFFFFF (0 bits) and Raw (66).
My array is 68 long and sending only 66?
I don’t understand, help please.

My code:

#include <IRremote.h>
// IR led on D3 (not work any other pin)
int buttonPin = 12;
int buttonState = 0;
IRsend irsend;
 
//This raw received from tv remote: {-26992 8900 -4500 550 -550 550 -550 600 -550 550 -600 500 -600 500 -600 550 -550 550 -550 550 -1700 550 -1700 550 -1650 550 -1700 550 -1700 550 -1700 500 -550 600 -1650 550 -600 550 -550 550 -550 550 -550 600 -1650 550 -600 550 -550 550 -550 550 -1700 550 -1650 600 -1650 550 -1700 550 -550 550 -1700 550 -1650 600 -1650 550};
unsigned int powerOn[68] = {26992,8900,4500,550,550,550,550,600,550,550,600,500,600,500,600,550,550,550,550,550,1700,550,1700,550,1650,550,1700,550,1700,550,1700,500,550,600,1650,550,600,550,550,550,550,550,550,600,1650,550,600,550,550,550,550,550,1700,550,1650,600,1650,550,1700,550,550,550,1700,550,1650,600,1650,550};

void setup()
{
  Serial.begin(9600);
  pinMode(buttonPin, INPUT);
}
 
void loop() {
  buttonState = digitalRead(buttonPin);
  if (buttonState == HIGH) {
      irsend.sendRaw(powerOn,68,38);
      delay(100);
  }
}

Drop that first number from your data.

http://www.righto.com/2009/08/multi-protocol-infrared-remote-library.html

"The raw data for received IR measures the duration of successive spaces and marks in 50us ticks. The first measurement is the gap, the space before the transmission starts. The last measurement is the final mark.

The raw data for sending IR holds the duration of successive marks and spaces in microseconds. The first value is the first mark, and the last value is the last mark.

There are two differences between the raw buffers for sending and for receiving. The send buffer values are in microseconds, while the receive buffer values are in 50 microsecond ticks. The send buffer starts with the duration of the first mark, while the receive buffer starts with the duration of the gap space before the first mark. The formats are different because I considered it useful for the library to measure gaps between transmissions, but not useful for the library to provide these gaps when transmitting. For receiving, 50us granularity is sufficient for decoding and avoids overflow of the gaps, while for transmitting, 50us granularity is more than 10% error so 1us granularity seemed better."