New MCE Remote Software

[popechild]

Oh, and for what it's worth, if I try to do a "Test Channel Change" using a couple of the numbers that do the "almost every other time" thing, (for instance, test channel 13, when 1 and 3 each work around every other time), nothing happens at all on the sat box.

[StephaneM]

Originally Posted by popechild

And it seems to be very picky as to exactly how/where I hold the remote in front of the receiver

You have to point your original remote directly (and straight) to the MCE receiver, and the remote must be very close to the receiver (like two times the width of your thumb). It's better to learn codes with both the receiver and original remote layed down on a strong surface. Also avoid any other IR source during learning (fluorescent bulbs, plasma tv, direct sunlight)

those learned commands seem to be getting passed to the sat box. I started with learning "0". Once I had learned twice (code 1 and code 2), I tested the button. First "test" worked great. Pressed "test" again, sat box didn't get anything. Pressed it a third time, worked great.

My software use debounce algorithm. That's why there is code 1 and code 2. Every other blast it use code 1, then code 2. So you have to be careful when you learn code 1 and code 2, they must be learned one after another. This is to ensure that if your remote also use debouncing codes, LM RKM will learn the two codes used.

You can also copy/past code 1 into code 2 (this way you'll definitely know if your remote is using debounce codes or not)

It seems like it's not learning the codes properly from my remote, or learning them properly but only occasionally sending them correctly.

Proper learning can requires a few tries:

* start with a quick press and release on your original remote's button
* test 4 or 6 times the learned button in a row
* if this is not great, re-learn, this time press a little longer on the original remote's button.

I don't suppose you have the Dish Network VIP-211 in your database do you?

There are some Dish Network devices defined (500 / 625 / 211 / 311), though I cannot guarantee that they will work.

Regards,
Stéphane.

[popechild]

I actually can't get it to learn anything at all if it's pointed straight at the receiver, no matter how close or far away it is. I have to point it at about a 45 degree angle before it'll recognize anything during the learning process. (I can learn fine from straight away with the replacement driver software, so I don't think it's a problem with the receiver.)

Is there a way you can provide the codes for a few of the digits from the 211 in your database? If it works for me, I'll be happy to pay for the donator version to get the full set of codes, but I'm hesitant to pay anything up front since I'm having so much trouble getting it to work. It would be well worth it if it would let me skip this whole learning hassle though.

[StephaneM]

Originally Posted by popechild

Is there a way you can provide the codes for a few of the digits from the 211 in your database?

The 211 profile is empty, so here is a sample code for the "0" button of a 625:

Code1

55530.8143654802|450 -6150 |400 -2800 450 -1650 400 -2800 450 -2800 400 -2800 450 -1650 450 -2800 400 -2800 450 -2800 400 -2800 450 -2800 400 -2800 450 -2800 400 -2800 450 -2800 400 -2800 450 -6150 |450 -2800 450 -1650 400 -2800 400 -2850 450 -2750 450 -1650 450 -2750 450

Code2

55179.6166625074|500 -6100 |450 -2800 400 -1650 450 -2800 400 -2800 450 -2800 400 -1650 450 -2800 400 -2800 450 -2800 450 -2800 400 -2800 450 -2800 400 -2800 450 -2800 400 -2800 450 -2800 400 -6150 |450 -2800 400 -1700 450 -2750 450 -2800 400 -2800 450 -1650 400 -2800 450

[popechild]

Originally Posted by StephaneM

The 211 profile is empty, so here is a sample code for the "0" button of a 625:

Code1

55530.8143654802|450 -6150 |400 -2800 450 -1650 400 -2800 450 -2800 400 -2800 450 -1650 450 -2800 400 -2800 450 -2800 400 -2800 450 -2800 400 -2800 450 -2800 400 -2800 450 -2800 400 -2800 450 -6150 |450 -2800 450 -1650 400 -2800 400 -2850 450 -2750 450 -1650 450 -2750 450

Code2

55179.6166625074|500 -6100 |450 -2800 400 -1650 450 -2800 400 -2800 450 -2800 400 -1650 450 -2800 400 -2800 450 -2800 450 -2800 400 -2800 450 -2800 400 -2800 450 -2800 400 -2800 450 -2800 400 -6150 |450 -2800 400 -1700 450 -2750 450 -2800 400 -2800 450 -1650 400 -2800 450

Thanks for posting that. More progress I suppose, but still not there yet. Those codes work, but in the same way that my learned codes work (when I can get it to learn) - every other time. I put those codes in for the "0" button, and here are the symptoms (All of these symptoms remain the same whether I use code 1 and 2 as you have them, or use either of the codes duplicated in code 1 and 2):

1. If I press "Test Button", it works exactly every other time. First time yes, second time no, third time yes, etc.

2. If I put "0" in the "Test Channel Change" field, it works correctly every other time as well.

3. If I put "000" in the "Test Channel Change" field, the first time it gives me "00", the second time it gives me "0", then repeats "00" and "0" every other time.

So obviously it's only getting every other signal. It doesn't matter what I put in the delay fields, it seems to do the same thing. I can try "000" (and it'll give me "00"), then wait 30 seconds to try again and it still just gives me "0" the second time.

It definitely sounds like it has something to do with the whole debounce thing, but I don't understand that enough to know if these symptoms mean it's debouncing when it shouldn't, not debouncing when it should, or what.

I'm hoping this will help give you a better idea of what's going on!

[StephaneM]

Originally Posted by popechild

It definitely sounds like it has something to do with the whole debounce thing, but I don't understand that enough to know if these symptoms mean it's debouncing when it shouldn't, not debouncing when it should, or what.

Debouncing is a technique to avoid repeated keypress when not wanted. It works like this:

* You press (steadily) on "0", the remote is sending the same IR code over and over
* The receiver acknowledge one code and drop all the other repeated codes
* You release and press again "0", the remote is sending another IR code
* The receiver will acknowledge this new code

LM RKM is aware of this, that's why it force you to learn two codes for the same button. Each time you ask LM RKM to blast for "0" it will use both codes alternatively (to verify this, remove the "Code 2" field content for the "0" button and test the IR code, you'll see that the IR emitter will only blinks every other time)

I know some IR devices let you turn debouncing on/off, so it is very possible that the codes I gave you are for a device with debouncing turned off.

To correctly learn debounced codes, you have to learn Code 1 and Code 2 each after another. For instance, find the best position for the receiver to be able to learn your remote and do this:

* Click on "Learn" for Code 1
* With your remote pointing away from the MCE Receiver, press and hold button "0"
* Move your remote toward the MCE Receiver until you touch it
* Move your remote away, release the "0" button
* At this point "Code 1" must have been learned
* Click on "Learn" for Code 2
* Do the same thing, remote away, press and hold "0", move remote to the receiver, move away, release the "0" button
* "Code 2" must have been learned

It is very important that you do not press again button "0" between learning of "Code 1" and "Code 2" otherwise you'll end up learning the same code for Code 1 and Code 2.

If I'm not wrong, as your remote is using debouncing you can press and hold the button a long time (usually I don't recommend holding the button too long)

Regards,
Stéphane.

[popechild]

Gosh, I wish you had one of these boxes to see if you could get it to work right.

I finally found a solid angle that seems to learn every time. I've done exactly what you described, making sure to learn twice in a row without any key presses in between, at least a dozen times, using multiple different numbers as test. Almost every time it does the exact same thing - when I go to test it, it recognizes exactly every other test. Never two in a row, ever.

I say almost every time, because once or twice it wouldn't do it every other time, it would do it exactly every 4th time. It would recognize it once, then nothing 3 times in a row, then it would recognize it again, then nothing 3 more times, etc. I then tried re-learning that button and after I relearned code 1 and code 2, it started doing it correctly every other time like all the other keys.

That got me thinking - is it possible that my unit is using 4 codes instead of 2, or something like that? Seems that maybe if it's using 4 codes, and I'm only sending it 2 of them, then it would only recognize it every other time. And if I didn't learn one of the codes correctly for a button, and I'm only sending it one correct code, it would only respond once out of every four tries, which is what I'm seeing.

Does this sound like it could be a possibility? Again, I don't understand debouncing enough to know if any units use 4 codes instead of 1 or 2. And if it IS doing this, is there anything that can be done to accommodate it and send it 4 codes instead of 2?

Thanks again for your help.

[popechild]

Sorry, but there's one other thing that's bugging me thinking about all this debounce stuff. Are you familiar with the MCE Remote Replacement Driver? It has a little command line app that lets you record an IR signal and then blast the recorded signal. When I use that driver and that app, I can record a single key press from my remote and then play it as many times as I want in a row and the satellite box recognizes it every single time. Doesn't this seem to indicate that my box isn't using debounce codes at all?

Any chance there's a way to take the learned IR code from the replacement remote app and copy it over into LM Remote Keymap?

[StephaneM]

Originally Posted by popechild

Doesn't this seem to indicate that my box isn't using debounce codes at all?

Yes indeed.

Any chance there's a way to take the learned IR code from the replacement remote app and copy it over into LM Remote Keymap?

No, this is not possible.

The only thing left to try is to change the duration of the blast (by clicking on the "More Options" button). By default this is 20ms, you can try 50, 100, 150, 200.

You can also try this code for the button 0:

54200|350 -9300 |399 -2699 399 -4350 399 -4300 399 -4350 399 -2699 399 -4350 399 -4350 350 -4350 399 -4350 399 -4350 350 -9300 |399 -2699 399 -4350 350 -4350 399 -4350 399 -2699 399 -4350 350 -4350 399 -4350 399 -4350 399

[popechild]

Changing the length didn't seem to do anything. And unfortunately that code didn't seem to work at all. If I put it for both Code 1 and Code 2 it never works. If I only put it in for Code 2, and use one of the codes you provided before for Code 1, it does the thing where it only works 1 out of every 4 times. Which seems to confirm that for some reason, if both Code 1 and 2 are valid, working codes, it will work every other time. If only one of the codes is valid, it will work every 4th time.

At this point, unless you have any other suggestions, I may just have to give up on getting this to work.

[StephaneM]

Originally Posted by popechild

At this point, unless you have any other suggestions, I may just have to give up on getting this to work.

I can add an option to repeat the codes, as for you it only work every other time it might do the tric. Send me an e-mail at lmgestion@hotmail.com and I'll send you a test build.

Regards,
Stéphane.

[popechild]

Email sent.

[StephaneM]

Originally Posted by popechild

Email sent.

Check your junk e-mail folder. I sent you the test build a few minutes before you posted this message.

[popechild]

Originally Posted by StephaneM

Check your junk e-mail folder. I sent you the test build a few minutes before you posted this message.

Got it. Will setup and test this evening.

[popechild]

Originally Posted by popechild

Got it. Will setup and test this evening.

Success!! (...at least initially - keeping my fingers crossed.)

As soon as I tried the new version, it was working better. The only strange thing is, it would never recognize the first key press on any number. So if I tested the "5" button, it wouldn't work. If I tried it again, it would work, and it would keep working every time for as long as I kept hitting "5". Then if I tried "4" it would do the same thing. No luck on the first press, then every one worked. If I went back to "5", it wouldn't recognize the first one again.

So it was 99% there as far as testing was concerned, but almost useless for actual usage, because every time a new digit was pressed, it wouldn't recognize it.

Finally I just started messing with every setting I could find to see if anything at all would help. What finally did the trick was changing the setting from "any emitter" and trying each specific emitter. 3 were listed, the first 2 didn't do anything, and the third to my surprise worked perfectly! It still requires leaving the new setting you added set to "2", but other than that it now seems to be working. No idea why this made a difference.

I'll do some testing over the next few days to see if it's actually going to hold up during regular usage, but so far so good!

[popechild]

Scratch that. Two days later and all of a sudden it's not working right again. The problem I was having before switching from "any emitters" to "emitter 2" is back, even though I'm still on emitter 2. Trying different settings doesn't seem to work either.

So here's how it behaves, one more time, in case there are any other suggestions before I just give in and purchase a USB-UIRT...

"Test Button"
This will ignore the first press of any new button being pressed, and then recognize it every time after that, so the following sequence of "Test Button" tests would give me the following results...

0 - no
0 - yes
0 - yes
1 - no
2 - no
1 - no
1 - yes
1 - yes
etc...

"Test Channel Change"
On this one, the first new # is always dropped, no matter what, so again, if I were to try this combination of tests in the following order, these are the results I would get (by the way, this is the way it works within Sage also...):

111 = 11
200 = 0
202 = [nothing]
185 = [nothing]
118855 = 185
999944 = 9994
445 = 4

I appreciate all your help so far. I don't suppose you've got any additional suggestions for this one? Maybe my system just doesn't like your software for some reason.

[StephaneM]

Originally Posted by popechild

I appreciate all your help so far. I don't suppose you've got any additional suggestions for this one?

Yes. As I explained the new "repeat" param is linked to the duration. If you switch back "duration" to the default of 20 ms, and repeat to 1 (without relearning IR codes), do yo see the same pattern ?

That is before adding repeat, blast would work every other time. I suppose this was true even when you changed the button. For instance "0" was working every other time, and if you switched to "1" it would work every other time either.

I could add a repeat interval settings. This way you would not have to change the duration.

Also from what I read, it seems to me that setting the repeat count to "3" would work...

[popechild]

Originally Posted by StephaneM

Yes. As I explained the new "repeat" param is linked to the duration. If you switch back "duration" to the default of 20 ms, and repeat to 1 (without relearning IR codes), do yo see the same pattern ?

That is before adding repeat, blast would work every other time. I suppose this was true even when you changed the button. For instance "0" was working every other time, and if you switched to "1" it would work every other time either.

I could add a repeat interval settings. This way you would not have to change the duration.

Also from what I read, it seems to me that setting the repeat count to "3" would work...

Okay, I started going back through and doing tests with various settings. I had changed the "blast duration" to 50, which had been working fine. When I changed it back to 100, it started working correctly again.

One somewhat unrelated question. When using the Sage plugin, the directions have you copy the .ir file that's created into the Sage common directory. At first I was assuming that when changes were made to the IR Device profile within Remote Keymap that Sage uses, that .ir file would need to be copied over to the Sage common directory, overwriting the old .ir file with the new one. But it appears from my testing that Sage recognizes the changes without needing the updated .ir copied over again. That's certainly simpler than it would be the other way, but wasn't what I was expecting. Is this correct?

[StephaneM]

Originally Posted by popechild

One somewhat unrelated question. When using the Sage plugin, the directions have you copy the .ir file that's created into the Sage common directory. At first I was assuming that when changes were made to the IR Device profile within Remote Keymap that Sage uses, that .ir file would need to be copied over to the Sage common directory, overwriting the old .ir file with the new one. But it appears from my testing that Sage recognizes the changes without needing the updated .ir copied over again.

You don't need to copy the .ir file when you make some changes in LM RKM. The file only contains the ID and name of your IR device. There is nothing in this file that tells SageTV how to blast IR at all.

[archcommus]

Hi, I'd like to use your software but I don't want to use the custom volume overlay, it appears though even with the box unchecked. Any suggestions? Thanks.