Mercurial > hg > audiostuff
view spandsp-0.0.6pre17/ChangeLog @ 4:26cd8f1ef0b1
import spandsp-0.0.6pre17
author | Peter Meerwald <pmeerw@cosy.sbg.ac.at> |
---|---|
date | Fri, 25 Jun 2010 15:50:58 +0200 |
parents | |
children |
line wrap: on
line source
07.08.29 - 0.0.4 - Steve Underwood <steveu@coppice.org> - 06.05.23 - 0.0.3 - Steve Underwood <steveu@coppice.org> - T.38 now implemented, though it needs further polishing. - G.726 and G.722 now implemented. 04.08.29 - 0.0.2 - Steve Underwood <steveu@coppice.org> - T.4 no longer uses libtiff for compresion and decompression on the line side (it is still used to handle the TIFF files). Spandsp no longer depends on accessing the "internals" of libtiff. New 1D and 2D compression and decompression code now handles the line side. This should be more robust than using libtiff, and handles the fudging of bad scan lines rather better. - T.30 line turn-around timing corrected. - T.30 DCS header contents corrected. - T.30 provision for V.17 added, as an option. - T.30 now has a statistics interface, for applications to find details about image transfers. - T.30 processing now correctly closes TIFF files are they have been sent. - FAX header line insertion is now supported for transmitted FAXes. - asynchronous serial processing now has a V.14 compatible rate adaption option. - V.8 modem negotiation module added. - OKI ADPCM now supports Dialogic compatible operation at 6k samples/second, as well as 8k samples/second, using its own sample rate converter. - message queuing added to ease things like T.31 handling - incomplete skeleton (still rather nasty, to be honest) T.31 processing. - incomplete skeleton of V.22bis added. There is probably quite a bit to do to get this complete and robust. It probably requires the T.31 processing be expanded to something more like V.250, so it doesn't just handle FAX. - Incomplete V.17 added (there could be an IBM patent related to the TCM part of V.17, but I think it has expired. Don't distribute binaries of spandsp built with V.17 enabled until this is resolved - and the modem is completed, of course :-) ). The current version transmits OK. It receives OK with the long training sequence, but the symbol and carrier syncing isn't good enough for robust receiving with the short training sequence. - A basic BER tester has been added. - The faster modems now have proper shutdown procedures, whether one is defined in the spec. or not. - The modems (except simple FSK) now have APIs to access performance information. - A common GUI (using FLTK) added for the modem tests. - A bit more documentation added. - Some API inconsistencies cleaned up. - The usual bundle of assorted buglet fixes and polishing. 04.04.29 - 0.0.1k - Steve Underwood <steveu@coppice.org> - The first version which gives solid faxing for a large number of people.