To all, if I misquote anything please comment and correct. That said, Over the last 4 to 5 years I have been reading the forum on this site and it has given me much information. About 6 months ago I purchased a TSU9600 and just started as a member.
There are many notes on the site about Clean and Dirty codes. From what I understand a "Clean" code matches the manufactures specifications, and was "most likely" generated by a computer. When a code is "Learned" my a remote it "resembles" or not in some cases, the RS pulses from another remote. There is documents on this site giving hints about the best environment to do the learning in, dark room, etc. These codes are "usually" not the same as clean codes.
For example, if the Sony Clean Code was (in pronto format) ….0030 0018 0018 0018…. A dirty code could be …0031 0019 0018 0019… The difference in the dirty to clean codes Is not very much and the dirty codes should still work. The more then numbers are off from the clean code the more difficult the reception could be. I have learned quite a bit about the Sony code format and feel fairly confident about "Cleaning" up a Sony code. There also some utilitys on this site that can generate clean codes, but you must know what codes you want. (read, learn, repeat.)
The Sony codes on Philips Database (while hidden from view) are not always in a Pronto format, some are a long string of hex characters. I think these are the pep Raw Format, but I am not sure. These strings are much longer than a string of Sony codes in the Pronto format. Leading me to question which is a better code. Have you read the document about recommendations on "Learning" codes. My objected is always have the best codes you can get and the remote will work the best. Look at the codes you have learned with your remote. Pep will show you them in the "My Database". If the codes look to have a dissimilar format (length, dissimilar codes, etc.) it could be a bad learn.
Try learning the same button several times (different names) and compare the learned code. See how close the codes look. It could give you a start.
