android NDK: Is optimizing array copy worth it? - android

For a school project I am creating an android app that involves streaming image data. I've finished all the requirements about a month and a half early, and am looking for ways to improve my app. One thing I heard of is using the android NDK to optimize heavily used pieces of code.
What my app does is simulate a live video coming in over a socket. I am simultaneously reading the pixel data from a UDP packet, and writing it to an int array, which I then use to update the image on the screen.
I'm trying to decide if trying to increase my frame rate (which is about 1 fps now, which is sufficient for my project) is the right path to follow for my remaining time, or if I should instead focus on adding new features.
Anyway, here is the code I am looking at:
public void updateBitmap(byte[] buf, int thisPacketLength, int standardOffset, int thisPacketOffset) {
int pixelCoord = thisPacketOffset / 3 - 1;
for (int bufCoord = standardOffset; bufCoord < thisPacketLength; bufCoord += 3) {
pixelCoord++;
pixelData[pixelCoord] = 0xFF << 24 | (buf[bufCoord + 2] << 16) & 0xFFFFFF | (buf[bufCoord + 1] << 8) & 0xFFFF | buf[bufCoord] & 0xFF;
}
}
I call this function about 2000 times per second, so it definitely is the most used piece of code in my app. Any feedback on if this is worth optimizing?

Why not just give it a try? There are many guides to creating functions using NDK, you seem to have a good grasp of the reasoning to do so and understand the implications so it should be easy to translate this small function.
Compare the two approaches, you will no doubt learn something which is always good, and it will give you something to write about if you need to write a report to go with the project.

Related

How can I visualize PCM data

I have a PCM datafile that I know is valid. I can play it, edit it into pieces, etc. and it will always play, as well as the individual pieces.
But when I try to translate it into shorts from bytes
bytes[i] | (bytes[i+1] << 8)
The file is 16 bit, single channel and 44100 sampling. I don't see anything that looks like a wave file visually.
As a test I record led among silencer with one very loud sound in the middle. Still the chart I made from my intake looked like every other chart when I try this. Am I somehow doing this wrong? Or misunderstanding what I'm reading/attempting?
All am looking to do is detect a very low threshold to find a word gap.
Thanks
My psychic powers suggest this is a big-endian vs little-endian thing.
If the source file stores samples in big-endian, this is likely what you want:
(bytes[i] << 8) | (bytes[i+1])
For what it's worth, WAV files are little-endian.
Other possibilities include:
I don't see your code, but maybe your code is only incrementing i by 1 instead of 2 on every loop iteration. (A common mistake I've made in my own code).
signed types or casting. Be explicit how you do the bit operations with respect to signed vs. unsigned. I'm not sure if "bytes" is an array of "unsigned char" or "char". Nor am I sure if "char" defaults to signed or unsigned. This might be better:
unsigned char b1 = (unsigned char)(bytes[i]);
unsigned char b2 = (unsigned char)(bytes[i+1]);
short sample = (short)((b1 << 8) | (b2));

Pic16F688 has no stable readings via buletooth

I have spent much time trying to find out where is my mistakes while Im flashing the PIC16F688. The Pic has successfully flashed using PicKit2. Im using the Pic to convert analog pressure sensor to digital output and sending the data via Bluetooth, but the Bluetooth is not receiving stable numbers of data. The data is consist of 4 character decimal number that is between 0 and 1023.
The problem is that the Bluetooth can't wait at specific number and keep reading it, instead, it is reading the 4 digits in random.
I think my mistake is within the configuration of internal oscillator.
I'm attaching my code, the code is written to configure the flexiforce sensor circuit that outputs analog voltage up to 5v, and then the pic duty is to convert it to digital as I mentioned above.
it might be my wiring is not correct, please If you could help out with this one
and what configuration "at edit project" do I need to choose for Mikro PRO software?
I used "Bluetooth terminal" app to see my data asynchronous from Bluetooth.
Thank you.
char *temp = "0000";
unsigned int adc_value;
char uart_rd; int i;
void main()
{
OSCCON = 0x77;
ANSEL = 0b00000100;
CMCON0 = 0X07;
TRISA = 0b00001100;
UART1_Init(9600);
Delay_ms(100);
while (1)
{
adc_value = ADC_Read(2);
temp[0] = adc_value/1000+48;
temp[1] = (adc_value/100)%10+48;
temp[2] = (adc_value/10)%10+48;
temp[3] = adc_value%10+48;
for (i=0;i<4; i++)
UART1_Write(temp[i]);
UART1_Write(13);
Delay_ms(1000);
}
}
You can use itoa function to convert ADC integer value to characters for sending over UART. If there is error in calculation then you wont get appropriate value. Below code snippet for your reference :
while (1)
{
adc_value = ADC_Read(2);
itoa(adc_value, temp, 10);
for (i=0;i<4; i++)
UART1_Write(temp[i]);
UART1_Write(13);
Delay_ms(1000);
}
Please check Baud Rate you have configured at both ends is same or not. If baudrate mismatches then you will get Random value at Bluetooth Terminal where you are reading values.
What i would suggest, if you have a logic analyser, hook it up. If you don't recalculate your oscillator speed with the datasheet. It could just be that the internal oscillator is not accurate enough. What also works, is to write a function in assembly that waits a known time (by copy-pasting a lot of NOPs and using this to blink a led. Then start a stopwatch and count, say, 100 blinks. This is what i used to do before i had a logic analyser. (They are quite cheep on ebay).

Android NFC Card emulation with fixed UID

I've downloaded NFC parts from AOSP and I'm looking for the method used by Android to generate the random UID used by card emulation. My goal is to fix the UID instead of having a different one each time there is a communication with the target. I found inside "libnfc-nci" module the file "nfa_ce_act.c" containing this:
void nfa_ce_t3t_generate_rand_nfcid (UINT8 nfcid2[NCI_RF_F_UID_LEN])
{
UINT32 rand_seed = GKI_get_tick_count ();
/* For Type-3 tag, nfcid2 starts witn 02:fe */
nfcid2[0] = 0x02;
nfcid2[1] = 0xFE;
/* The remaining 6 bytes are random */
nfcid2[2] = (UINT8) (rand_seed & 0xFF);
nfcid2[3] = (UINT8) (rand_seed>>8 & 0xFF);
rand_seed>>=(rand_seed&3);
nfcid2[4] = (UINT8) (rand_seed & 0xFF);
nfcid2[5] = (UINT8) (rand_seed>>8 & 0xFF);
rand_seed>>=(rand_seed&3);
nfcid2[6] = (UINT8) (rand_seed & 0xFF);
nfcid2[7] = (UINT8) (rand_seed>>8 & 0xFF);
}
This method generate an UID for FeliCa tags. I'm not able to find the one for ISO14443 cards (MIFARE) which generate an UID beginning with 0x08 by default. According to Martijn Coenen, as explained in his G+ Post, it's something possible.
Sorry, I realize many people wanted this, but it's not possible in the official version. (You could of course do it with some AOSP hacking). The reason is that HCE is designed around background operation. If we allow apps to set the UID, every app possibly wants to set their own UID, and there's no way to resolve the conflict. We hope that with HCE, NFC infrastructure will move to higher levels of the protocol stack to do authentication instead of relying on the UID (which is easily cloned anyway).
https://plus.google.com/+MartijnCoenen/posts/iX6LLoQmZLZ
Is anyone know how to achieve it?
Thanks
One important thing to know is that the UID transfered at a very low level of the nfc protocol. This means that it is done independently by the nfc firmware and not within the android operating system.
We had the same problem in our NFCGate project and found a solution for Broadcom BCM20793 chips like the ones in the Nexus4/5 and others by writing the UID with NFC_SetConfig directly into the chip firmware.
You can see a working version in our repository on github. Here is a non-tested version to show the principle:
uint8_t cfg[] = {
CFG_TYPE_UID, // config type
3, // uid length
0x0A, // uid byte 1
0x0B, // uid byte 2
0x0C // uid byte 3
};
NFC_SetConfig(sizeof(cfg), cfg);
Our tests revealed that android sometimes sets the UID back to random (length=0 if I recall correctly), so you need to find a good place to set it when you need it or do something similar as we did and intercept NFC_SetConfig calls from android to re-set our own UID.

Sending data from arduino through bluetooth serial

I’m currently trying to design a controller that would communicate through Bluetooth to an android phone (Galaxy Nexus). I’m facing a few challenges. Also, I don’t have much practical programming experience.
At the core of the controller resides an Arduino microcontroller who scans the state of 8 digital pins and six analogue pins (10 bits) and sends the data through serial, to an HC-05 Bluetooth chip.
The android phone should then read the serial information sent through Bluetooth and either transmit the packet to another phone - That will take me a while to implement as I know very little of how the internet works - or analyse and interpret it for further action to be taken
What I’m asking for are insight as to the best way to go about doing this. Now what does best mean?
We’ll I want it to be real time so when I press a button or a combination of bottoms, boom, the android phone takes action fast enough for a human not to perceive a delay.
Then I want to be able to associate the information the phone reads in the serial buffer to the corresponding button or analogue pin. In case there is an error or to avoid that they fall out of sync
Here is what I have so far. I have not tested this code yet, partly because I’m still learning how to program the android part of this project, partly because I want some feedback as to whether I’m being silly or this is actually an effective way to go about doing this:
//Initialization
/*
This Program has three functions:
1) Scan 8 digital pins and compile their state in a single byte ( 8 bits)
2) Scans 6 Analogue inputs corresponding to the joysticks and triggers
3) Send this info as packets through seril --> Bluetooth
*/
#include <SoftwareSerial.h>// import the serial software library
#define A = 2
#define B = 3
#define X = 4
#define Y = 5
#define LB = 6
#define RB = 7
#define LJ = 8
#define RJ = 9
#define RX = 10
#define TX = 11
//Pin 12 and 13 are unused for now
SoftwareSerial Bluetooth(RX,TX); //Setup software serial using the defined constants
// declare other variables
byte state = B00000000
void setup() {
//Setup code here, to run once:
//Setup all digital pin inputs
pinMode(A,INPUT);
pinMode(B,INPUT);
pinMode(X,INPUT);
pinMode(Y,INPUT);
pinMode(LB,INPUT);
pinMode(RB,INPUT);
pinMode(LJ,INPUT);
pinMode(RJ,INPUT);
//Setup all analogue pin inputs
//setup serial bus and send validation message
Bluetooth.begin(9600);
Bluetooth.println("The controller has successfuly connected to the phone")
}
void loop() {
//Main code here, to run repeatedly:
//Loop to sum digital inputs in a byte, left shift the byte every time by 1 and add that if the pin is high
for(byte pin = 2, b = B00000001; pin < 10; pin++, b = b << 1){ if (digitalRead(pin)== HIGH) state += b; }
//Send digital state byte to serial
Bluetooth.write(state);
//Loop to read analgue pin 0 to 5 and send it to serial
for( int pin = 0, testByte = 0x8000; pin < 6 ; pin++, testByte = testByte >> 1) { Bluetooth.write(analogRead(pin)+testByte); }
}
//Adding some validation would be wise. How would I go about doing that?
// Could add a binary value of 1000_0000_0000_0000, 0100_0000_0000_0000, 0010_0000_0000_0000 ... so on and then use a simple AND statement at the other end to veryfy what analogue reading the info came from
// so say the value for analgue is 1023 and came from analgue pin 1 I would have 0100_0011_1111_1111 now using an bitwise && I could check it against 0100_0000_0000_0000 if the result is 0100_0000_0000_0000 then I know this is the analogu reading from pin 1
//could easily implement a test loop with a shiting bit on the android side
Is it pointless for me to be doing bit shifts like this? Ultimately, if I’m not mistaken, all data is sent as bytes ( 8 bits packets) so will the Bluetooth.write(analogRead(pin)+testByte) actually send two bytes or will it truncate the int data? how will it be broken down and how can I recuperate it on the android end?
How would you go about implementing this? Any insights or words of advice?
It's great that you are learning this! Some suggestions:
Your code will be easier to read, understand and maintain if you space it out a bit, particularly by adding newlines...
void loop() {
//Main code here, to run repeatedly:
//Loop to sum digital inputs in a byte,
//left shift the byte every time by 1 and add that if the pin is high
for( byte pin = 2, b = B00000001; pin < 10; pin++, b = b << 1) {
if (digitalRead(pin)== HIGH)
state += b;
}
//Send digital state byte to serial
Bluetooth.write(state);
//Loop to read analgue pin 0 to 5 and send it to serial
for( int pin = 0, testByte = 0x8000; pin < 6 ; pin++, testByte = testByte >> 1) {
Bluetooth.write(analogRead(pin)+testByte);
}
}
Also, you can use the shift operators with values other than one. so instead of keeping a shift mask that you then add in, you just use the a simple variable. A more classic way to do what you expressing in the first looop would be something like this:
#define PIN_OFFSET 2
for ( int i=0; i < 8; i++) {
if (digitalRead( i+PIN_OFFSET)== HIGH)
state += (1 << i);
}
The second loop could be done similarly:
for( int pin = 0; pin < 6; pin++) {
short testByte = 0x8000 >> pin;
short result = analogRead(pin) + testByte;
Bluetooth.write( result >> 8);
Bluetooth.write( result & 0xFF);
}
Note that the Bluetooth.write() call sends a byte, so this code sends first the most significant byte, then the least.
Lastly, you probably want to zero your state variable at be beginning of loop() -- otherwise, once a bit is set it will never get cleared.
You may want to think about what you are sending to the phone. It will be binary data, and that can be difficult to deal with -- how do you know the start and end, and often a value will be misinterpretted as a control character messing you up big time. Consider changing it into a formatted, human readable string, with a newline at the end.
I hope that helps.
The Arduino end of this is more than fast enough to do what you want to do and you need not worry in the slightest about minimising the number of bytes you send out from the micro to the phone when you're talking about such a tiny amount of data. I'm not quite clear what you're sending to the phone; are you wanting a different value for each button so when a button is pressed it sends out something like:
Button Was Pressed, Button Number (Two bytes)
And if it's an analogue value:
Analogue Value Read, Analogue Value MSB, Analogue Value LSB (assuming the analogue value is greater than eight bits
I don't know the Arduino code you're using, but my suspicion is that this line:
Bluetooth.write(analogRead(pin)+testByte)
Would send one byte which is the addition of the analogue value and whatever testByte is.
Would you be able to post a link to the documentation for this API?
Any delay with all this will come at the phone end. Please don't concern yourself in the slightest about trying to save a byte or two at the Arduino end! (I'm currently doing a project with a Cortex M3 sending data via a Bluetooth module to Android phones and we're shipping several k of data around. I know from this experience the difference between two and twenty bytes is not of consequence in terms of delay.)
It may be good for you to opt for ready available bluetooth terminal apps for your android handset. On Google play, few apps are:
1) https://play.google.com/store/apps/details?id=arduino.bluetooth.terminal&hl=en
2) https://play.google.com/store/apps/details?id=com.sena.bterm&hl=en
(and few more are availble on Google play).This will help you to focus only on controller side development.

Change in the audio samples when reading directly from the file using RandomAcessFile?

Iam trying to draw a graph of a sound file. I am doing in such a way that i take the bytes of the sound file using Fileinputstream and change it to the shorts and take the samples of of the sound and draw the graph according to that note:first 45 bytes are header so i skip these bytes.And i succeeded in doing this.
File-->skip header(44)-->Bytes-->to shorts-->seek to point-->take samples in Shorts-->DRAW THE GRAPH
But the problem is, I cant take the bytes of large audio file(2GB) to memory. Memory crash occurs.
So tried reading the shorts directly from the file using RandomAcessfile. But when I do like this am not getting a correct graph. I hope there is some sort change in samples am reading.
File-->skip header(44)-->seek to point-->take samples in Shorts-->DRAW THE GRAPH
My doubt is, any change occurs to the short samples of an audio data when we read directly from the file??RandomAcessFile is a good method? Is there any way to get the samples of a 2GB audio file without any change in samples.
Note:I skip the header first 44 bytes.
double mPerwidth = (double) iPixel / 960;
int SampletoDraw;
long totalLen = mRfs.length() - 44;
SampletoDraw = (int) (totalLen * mPerwidth) * 2;
System.out.println(iPixel);
// val=mRfs.readShort()/2;
mRfs.seek(SampletoDraw);
bData[0] = mRfs.readByte();
mRfs.seek(SampletoDraw + 1);
bData[1] = mRfs.readByte();
val = (short) ((bData[1] & 0xff) << 8 | (bData
[0] & 0xff));

Categories

Resources