The 7000 indicates a specific encoding system.

The first 0000 is where the modulation wavelength goes. If this works I guess modulation wavelength is ignored here and gotten from the ID.

The 0000 000A together indicate the length of the remainder as being 10 (decimal) pairs of numbers.
The next 0000 is the ID which does a table lookup for information not encoded here. 0000 seems to be the ID for RC5 timing characteristics (without forcing you to use RC5 frame structure).

The 0011 I have no clue.

The 0010 I think is a flag that the whole remainder is a repeating signal.

All the 0002's represent single '1' bits in the IR signal. All the 0001's represent single '0' bits in the IR signal. The 001B is the toggle bit ('0' or '1' depending on whether it is an even or odd press), but I don't know why 001B means that.

The first bit (the first 0002) is always '1' in Streamzap. The second bit is '1' when the command is <64 and '0' when the command is 64..127. The third bit is the toggle.

The next 6 bits "0002 0001 0001 0001 0002 0002" are binary 100011 which is decimal 35 which is the system number and ought to be the same for all commands.

The next 6 bits similarly encode the low six bits of the command number.

The 0000 after that represents the "lead out" burst.

The final 0000 is padding since the number of PAIRS of values is counted.

Evgeny,

The CCF you sent me works wonderfully, I tested the full command set and am sorry to report that the Streamzap Win32 drivers do not pass commands that are not physical buttons on the remote up to the application level. What this means is that only the labeled buttons (Power, VolDown, etc...) work. The others must be intercepted by the Win32 driver and not passed on.

So, based on that, I would say clean up the CCF removing the numbered entries and upload it to this site.

As others begin to work with Media Servers and would like to use Streamzap, yours are the cleanest codes that can be found. Also, I wanted to report that these 7000 code actually honor the rules of repeat better than the learned 0000 codes. With the learned codes, the repeat was immediate. For example, press 2 to enter a track and you'd get a string of 20 or so '2's immediately. The way I worked around this was to set the learned repeat pairs in the 0000 IR data to be non-repeat pairs. That way, you had to remove your finger from the button to get them to work.

Great Job guys. Kudos to both Evgeny and John. You don't know how much I appreciate this.

Thanks again,
Lyndel

This message was edited by Lyndel McGee on 04/15/03 20:50.

aCode="0010 0002 0002 001B 0002 0001 0001 0001 0002 0002 0001 0001 0001 0001 0001 0001 0000", so nCodeSeq, as the length of aCode is "0011"


0010 is delimiter:
. if code has only "once" signal, 0010 must be at the end
. if code contains only "repeat" section, then 0010 will be the first
. if both "once" and "repeat" parts exist in code, 0010 inserts between them


zTemplate, corresponded for dID=0 is:

zTemplate="|1[01]{01}[01]%11R"
index= 0123456789abc.....

Toggle bit at Streamzap toggles between "0" and "1", so we need to point to "01".
This couple "01" lies at zTemplate from offsets 03h, 07h and 0Bh; and the codes for them are 03h+14h="0017", 07h+14h="001B" and 0Bh+14h="001F".
I have choose 001B only by tradition, others works too.