summaryrefslogtreecommitdiff
path: root/vq
Commit message (Expand)AuthorAgeFilesLines
* Incremental update. Nearing beta 2, four more new descriptions to go...monty_branch_20000724Monty2000-08-154-6/+589
* Incremental commit (new noise computation, LSP books to match)Monty2000-08-071-109/+13
* Incremental updateMonty2000-08-041-2/+105
* Incremental updateMonty2000-08-032-0/+1140
* This commit was manufactured by cvs2svn to create branchunknown2004-03-2021-4842/+0
* First merge of new psychoacoustics. Have some unused codebooks toMonty2000-05-0822-940/+1979
* Incremental update toward first cut. The full engine is in place.Monty2000-02-232-6/+7
* Incremental update toward first cutMonty2000-02-233-192/+92
* The huff builder is happy.Monty2000-02-211-0/+193
* Because residue codebooks use very coarse quantization, I needed toMonty2000-02-2110-173/+354
* Enhancements (added cell spacing metrics and minimum cell distances to trainer)Monty2000-02-1612-72/+389
* Fixed bug reading codebooks with very large auxiliary encoding structsMonty2000-02-132-22/+39
* Collect training data from the lib when compiled with -DTRAINMonty2000-02-121-0/+230
* ...and correct a typoMonty2000-02-071-3/+3
* Minor optimization to align p and q indexes in .vqh files.Monty2000-02-072-6/+7
* Incremental update. Reduced maximum codebook/mapping/mode/backendMonty2000-01-281-2/+2
* The lib and vq, at least, build again. Tackling the examples and xmmsMonty2000-01-283-9/+9
* Minor change to the codebook abstraction (static books)Monty2000-01-217-169/+131
* Incremental update toward VQ integration, vorbis_info rearrangementMonty2000-01-121-2/+1
* Added '-s' option to vqmetrics, vqcascade and vqpartitionMonty2000-01-107-42/+42
* Allow cascade and metric utilities to take more than one codebookMonty2000-01-074-96/+230
* first cut of complete VQ utilitiesMonty2000-01-069-75/+364
* stabilize the generic metric some, bugfixesMonty2000-01-058-26/+66
* More VQ utility work. Utils now include:Monty2000-01-0512-518/+743
* More shuffling of includes to ease codebook integration. This breaksMonty2000-01-053-26/+51
* Minor build fixes, integrate XMMS into the autoconfed stuffMonty1999-12-309-40/+34
* Incremental updateMonty1999-12-304-74/+472
* More optimizationMonty1999-12-281-11/+40
* bugfixes for new functionalityMonty1999-12-286-129/+236
* bugfixes to new functionalityMonty1999-12-281-1/+2
* missed a save. see previousMonty1999-12-281-8/+8
* Incremental update, seperated weighting from the metricMonty1999-12-287-58/+99
* Incremental updateMonty1999-12-247-159/+264
* alloca() 'frees' memory when the function returns, not when we exitMonty1999-12-181-1/+2
* Change builder to find center of gravity of entries, not pointsMonty1999-12-181-8/+44
* New syntax for more convenient use, fixed a bounds bug for gmaxwellMonty1999-12-181-126/+171
* The vq book builder is mostly running now, but does not produce output yet.Monty1999-12-177-539/+627
* Additional refinements, cleanups for training routineMonty1999-12-174-24/+136
* refinements. More accurate distance error, better fudge.Monty1999-12-172-12/+23
* Incremental updateMonty1999-12-165-87/+198
* quant fixMonty1999-12-152-1/+232
* more quant workMonty1999-12-151-13/+11
* quant fixMonty1999-12-151-2/+2
* split out the includeMonty1999-12-153-60/+111
* add train.c, command line util for training/retraining codebooks.Monty1999-12-114-877/+333
* Incremental; the book is half builtMonty1999-12-101-48/+116
* New split code.Monty1999-12-091-494/+524
* Incremental update. This is the foam we're going with for now.Monty1999-12-081-127/+70
* incremental updateMonty1999-12-071-53/+71
* Silly, forgot to move to said faster machine.Monty1999-12-061-0/+9