X-Git-Url: http://git.johnwright.org/?p=fmit.git;a=blobdiff_plain;f=TODO;h=32bf762946d2fd010cbe6f602f6d33b3abd3bfd7;hp=2cf53df2a4028a84eade41b55e493c84d2a166e2;hb=HEAD;hpb=4b957a098b637ebe4d07827a80aeae157791a148 diff --git a/TODO b/TODO index 2cf53df..32bf762 100644 --- a/TODO +++ b/TODO @@ -1,7 +1,54 @@ -look for translations +automatically build qm files from ts: +http://www.cmake.org/Wiki/CMake:How_To_Build_Qt4_Software + +remove ConfigForm.png +update french translation + +automatiser le seuil de réaction +automatiser d'autres choses ? + +Add low-pass filter to avoid problems with high sampling rate + +Fix bug (voir mail) + + 1. Would it be possible to have a celeste offset i.e. where you could add a + constant value (Hz) to every note: +5Hz to -5Hz , this would be usefull for + tuning the celeste on the organ! + + 2. Would it be possible to have the note names only using sharps i.e. A#, C#, + D#, F#, G# this is what I have always encountered in the organ building + world. + - pouvoir passer d'un naming scheme � l'autre. + + 3. I would really like to use fmit. But it doesn't compile with the newest + version of Jack. It complains about "CaptureThread.cpp:(.text+0x7276): + undefined reference to `jack_error_callback'". It does compile with an + older version of Jack, but that version of Jack won't work with the + version of FFADO that works with my audio interface. Any chance you + could update fmit to compile with the newest version of Jack? + + 4. (voir mail de Jean-Jacques Sacr�): probl�me de chargement d'un + fichier de traduction. en effet, si PACKAGE_NAME correspond pas au path + d'installation, �a marche pas. nouveaut� dans la lib qt ? + + 5. (voir mail de wayne): implementer le low-pass filter + +Two remarks (of details): + + - as also can be seen on your welcome page, to my knowledge, 329.6Hz +is E3 rather than E4: since I remarked this for the whole frequency +range, it's probably due to the lack of a minus one somewhere in an +integer part operation. + + - I found that the default sound level for incoming audio is very +low so that setting is first required before using fmit; possibly, it +would be set higher in the next release. + + +unification of scales system volume à partir de la somme des harmoniques pas de la valeur max - perceptive volume from weighted spectrum ? + perceptive volume from weighted spectrum ? loudness ? de Bark ? high pass with FIR band pass ranges with low pass tests corrects des working transport possibles @@ -12,17 +59,17 @@ Doc Do not use a too big sampling rate. 22050Hz should be largely enough in most cases. promotion - contact AGNULA project - http://apps.linuxaudio.org/apps/screenshots + look for translations algo sep en bandes avec resampling ? Quantizer Viterbi + Detection automatique de la présence d'un son interface statistics - in a preselected range of notes + average on last n secondes, throw too old notes keep the history of the note (not only an average over the note) be able to drop too short notes average of volume @@ -32,7 +79,9 @@ interface add inharmonicity values in cents phase DFT - envelope ? + enveloppe LPC + adapt win length to f0, proportional to nb_periods + Spectrogram microtonal view autodetect octave ? stretch scale view to the complet scale length (not always 12 !) @@ -43,7 +92,10 @@ interface code cleaning use AutoQSetting under Views -autotools - maintainer-clean - doc directories of libs - curse autotools authors families up to five generations +COMMANDS ======================================================================= +lupdate ../ui ../src -ts fmit_fr.ts +lupdate ../ui ../src -ts fmit_de.ts +linguist tr/fmit_fr.ts +lrelease tr/fmit_fr.ts + +cmake -DCMAKE_BUILD_TYPE:STRING=Debug