CaptureIR progress

[The Robman]

Does CaptureIR still require dual-processor PCs?

[johnsfine]

Does CaptureIR still require dual-processor PCs?

No. Some results, especially computing the frequency, are better on a dual-processor PC. But it works just fine on a single processor system.

DecodeIR.dll rarely uses the frequency to deduce the decode. In most cases the structure of the signal is all the info you need for a correct decode.

Some IR sensors (such as the ones in a VCR) take out the frequency anyway, so it can't be used in decode regardless of the software or number of CPU's. Those sensors are much less picky about distance. One of those might work from 1 inch to 40 feet, on a signal where a sensor capturing frequency works from 2 inches to 8 inches.

[binky123]

I thought I'd give captureir.jar a whirl so I opened up the old vcr and hooked up DB25 pin 9 to VCR IR sensor pin 1(Vout) and DB25 pin 25 to VCR IR sensor pin 2(GND) or clip to VCR case.

I clicked start capture, pressed a Sony Remote TV Power button and then clicked stop capture. It captured it but decodeir.dll couldn't decode it. It looks like the first lead-in was not captured. If I hand decode it, it matches up with Sony12 1Ø1Ø1ØØ 1ØØØØ. It also looks like the phase notation is inverted but I'm not really sure of the output for captureir.

This is on WinXP Home with Core2 Duo E6300 processor and 1GB DRAM in case any of this is important.

Sony12 Power is

Code: Select all

+4-1     +2-1 +1-1 +2-1 +1-1 +2-1 +1-1 +1-1 +2-1 +1-1 +1-1 +1-1 +1-44
lead-in  1    0    1    0    1    0    0    1    0    0    0    0

The file from captureir:

Code: Select all

#	Protocol	Device	Sub Device	OBC	Hex Cmd	EFC	Misc	Frequency	Raw Signal
1	1	0	(Unknown)							0	496 -1295 496 -697 496 -1287 502 -691 501 -1268 522 -670 523 -670 522 -1267 523 -669 522 -670 523 -670 522 -671 25614 -2463 522 -1268 520 -671 522 -1268 523 -669 524 -1266 523 -669 522 -669 524 -1267 522 -671 522 -670 523 -668 524 -669 25614 -2463 522 -1267 523 -669 522 -1267 523 -670 522 -1266 524 -668 523 -670 524 -1266 524 -668 523 -669 523 -669 523 -663

[johnsfine]

Which captureIR did you use?

The last one I posted (forget where, but I'll check later) has an option to invert the polarity of incomming signal.

I didn't look carefully at your timing numbers but what you said indicates that polarity is backwards, so you can fix it easily with that option in CaptureIR.

[binky123]

I used the CaptureIR from here as listed in the Sticky in the Software forum.

I can't get the Setup button highlighted. I'm using JRE 1.5.0_10 if that matters.

[johnsfine]

Oops!

Why is the Oct 2005 version there, not the Oct 2006 version.

I need to look around a bit.

OK, I looked around and found this one
https://www.hifi-remote.com/forums/dload ... le_id=3760

It seems to be a few days newer than the copy I'm using myself. I need to check a few computers and figure out what I might have left in the wrong place. Meanwhile that should be a good copy to use.

There is a Preferences choice in the File menu. There you can choose the polarity of the IR sensor: Active High or Active Low.

[binky123]

Thanks. It works well with the correct DecodeIR.DLL from here.

It was able to decode various signals in the 38-42KHz range but not the 58KHz range due to the IR sensor in this Quasar VCR.

[johnsfine]

I assume the polarity did turn out to be active low. Most modulated sensors are active low.

The modulated sensors I have tested accept a wider range of frequency (I would expect 58Khz to work with a 38Khz sensor). But I haven't yet tried one built into a VCR, just similar one bought as components.

I don't know when I'll have time to open up that VCR. But when I do, do you have a suggestion for identifying the pins of the IR sensor? If I test with a volt meter with the VCR powered, both the power pin and the output pin of the sensor should be high.

[binky123]

Yes, I had to change it to Active Low.

I was able to use this same method using an IR sensor on a RCA Sat receiver to record RCA VCR signals @ 58KHz.

You should be able to hook up a voltmeter and when sending an IR signal, the voltage reading should drop. I just tried it on the Quasar as I still had it open and the voltage dropped to about 3.4v and then down to 1.5v and then back up to 4.8v steady state. Even if you hook up the probe to the wrong pin, you just won't get any sampling data.