Initial support for OSX 10.2, select output device from dropdown list in config panel, select microphone input device from dropdown list, multiple output devices supported, multiple input devices supported, multichannel devices supported, smoother chat interface offers list of chattable users and selects the other user automatically in the 2-user case. New filter 'fill', bender filters have better default settings. http://ffem.org/gdam/macosx.html
Fixed bug where beatcalc locked up machine. http://ffem.org/gdam/macosx.html
Microphone input now working. http://ffem.org/gdam/macosx.html
GDAM-0.942 is available. New features include: alsa 0.9 support (courtesy of Takashi Iwai), many sequencer improvements, new beatmatching features and synchronization fixes, new filters (stall, djrewind, morph, beat-aware respeed, seek) new features and skin improvements for old filters (bounce, spin, bend) (full changelog)
Working MIDI control, including the ability to learn bindings on the fly. New dj-rewind filter. http://ffem.org/gdam/macosx.html
Greatly reduced size of download and installed app. Various new features (song loop, temporary rate nudging in turntable skin) and filters (slide, stall). http://ffem.org/gdam/macosx.html
Better support for irregular filenames, improved performance (should mean fewer dropouts), gdam-launcher options saved between sessions, default is to run with unstable features disabled. http://ffem.org/gdam/macosx.html
UPDATED 12 FEB: 0.942-2 was totally broken, 0.943-3 is fixed
A new version of the binary release for Mac OS 10.1 this new version uses the standard OS X installer. we have also added support for .aiff files, and the help files are now included. http://ffem.org/gdam/macosx.html
A new version of the binary release for Mac OS 10.1 this fixes a bug with ID3 headers which caused GDAM to complain that many mp3 files were invalid. http://ffem.org/gdam/macosx.html
GDAM-0.941 is available. GDAM now compiles and runs on MacOS X. A binary distribution is also available. New features include: turntable cue points (courtesy of Florian Eyben), improved ladspa support with new skins, many sequencer improvements, bender keep-in-beat mode and on-the-fly position markers for improved scratching.
0S X users who have unix development tools and/or fink installed can build gdam from source tarball or cvs. Most OS X users will want to use the binary distribution, which includes all requisite libraries and an x server. http://ffem.org/gdam/macosx.html
GDAM-0.939 is now available for download. Skin files accidentally left out of the 0.938 release have been replaced. The device and event polling has been rewritten, which should solve everyone's problems with unresponsive clients and frozen time labels.
GDAM-0.938 is now available for download.
A popular request, mono cue mixes down each stereo stream to a mono channel. This allows use of the left channel for a mono room mix, and the right channel to cue new tracks. Run gdam-launcher with "--define mono_cue" to activate.
Recent versions of GDAM have included 2-d "panners" to simultaneously control the cutoff frequency and filter depth of equalization filters with mouse motion within a rectangle. Now the bluetone 'wheels' skin contains a panner which drives two eq filters on two different turntables. By setting one filter to high-pass and the other to low-pass, this central panner can be used to crossfade in eq-space.
The long stretch of time since the last news item does not mean that development has stalled. GDAM has made quiet but significant progress.
Changes include: resolved FreeBSD compile problems (see user manual for instructions), support for usb control devices (keyboard, trackball, etc), more flexible input binding (logarithmic and inverted response curves, bind to skin handlers, modify rather than set arg value), various skin improvements, and breaking GDAM into seperate packages to relieve dependency problems.
Currently development is focused on: a sampler (roughly, recycle+cubase), and a groovebox-style skin for step sequencing. (run gdam-launcher with "--define unstable" to see what is in the works)
It is no longer necessary to edit the skin xml files to change key bindings for the turntable and wheels interfaces. The bindings are stored in files 'wheelhotkeys.xml' and 'wheelshotkeys.xml' which are installed with other skin files. Users can copy those files into ~/.gdam/skins/ and edit them. Those copies will be used whenever a turntable or wheels is created.
The cvs version now contains rough, working support for XMMS Visualization plugins. gdancer, blurscope, sanalyzer, and jakdaw all work.
A few user requests have been implemented. They are currently in the cvs tree, and will be included in 0.930 and future versions. We like feedback and suggestions.
In the `bluetone' turntable, it is possible to change the time display to show time left until the end of track, or to count down to/up from the first beat location. The button to do this is to the right of the time label, marked `...' for the moment. A button may be added to the default skin as well.
A notebook page of visibility controls has been added to the song selector. There are several visibility options, which are applied one after the other. You can choose to hide or show songs by default, show or hide songs with unknown bpms, and show or hide songs which also exist in a different gdam song database.
We've begun collecting and expanding documentation into a users guide, available online in the docs section.
Alsa allows me to open multiple subdevices on my hoontech sound card (4dwave nx, uses the trident 4dwave chipset.) As each subdevice is opened by the gdam server, volume and pan controls for that subdevice become available in alsa-enabled mixers. In gamix, the `pcm playback' and `pcm1 playback' panels pops up, each with two vertical sliders to control the front and back volumes and a horizontal pan control. I turn down the back volume on pcm and the front volume on pcm1, leaving me with two discrete stereo output streams. There are several benefits of this setup. The two subdevices are fed in sync, and sources continue to play in perfect time as they are switched back and forth between subdevices. In addition, if i'm using one stream for cueing (say, plugging headphones into the rear speaker jack) i can choose to increase the volume of the other stream (playing from the front speaker jack to the stereo) on the rear output, making it easier to monitor what is happening with the room mix.
Finally we support ALSA (tested with version 0.5.7). We look forward to trying out some of its powerful features. Uncomment and review the alsa device entries in the server config file.
Several visualization filters have been added recently. The oscilloscpe provides snapshots of the waveform. The spectrum filter shows the power of different frequencies at any moment, while the scrolling spectrum shows the power spectrum over time (also called a `voiceprint').
The contiguous queueing and playlist abilities formerly available through the `queue' component have been merged into to normal turntable. One benefit of this is that when playing continuous-mix albums, it is possible to mix from one track to the next (although it's not yet possible to nudge back and forth across the track boundary.) The intelligent file and directory sourcing features from gdam123 (automatic ordering of tracks according to playlist or cddb file) are now available in the turntable component.
We've recently gained access to a linuxppc machine, and have verified that gdam compiles and functions. The only known problems are related to the readline substitute, gdam123,gdamcli are affected.
We've been expanding and improving the skinnability of gdam's interface. The new skin (`bluetone' in the cvs source code) is mostly complete. Once finished, it will serve as an example for those who'd like to design new skins. The skins can be launched by clicking on the small icon buttons next to `turntable' and 'wheels' in the launcher.
We have abandoned the readline library completely in favor of getline, a lightweight replacement used by ncftp. Readline was the most common cause of configuration problems, and this change resolved the problems some people were having.
Gdam now does all its internal processing in 24 instead of 16 bits. Full volume is represented by 24 bits, leaving eight for overflow. This means that audio data will no longer clip if it isn't compressed into 16 bits at each step. A bunch of endianness cleanups were done at the same time.
Gdam now compiles on FreeBSD (with gnu-make). (Versions of gcc from <10.1999 were giving internal compiler errors -- just upgrade if you need it.)
Gdam now supports LADSPA plugins, an open format for simple audio plugins. Plugins can be loaded, and mininetwork configurations can be compiled into LADSPA plugins.
The cacheable source now has the ability to save descriptions of loops to the song database, and load them. This makes it easy to quickly recreate your favorite loops.
A few user-requested features have made their way into recent releases. Right-clicking on the `bpm' column header of the song selector will hide/show songs for which the tempo is unknown. Control-clicking on the song selector launch button will launch an empty song selector, which can be filled with a short playlist.
Now the server is automatically run if it isn't available upon startup.
we are creating debians and redhat packages automatically from cvs nightly now.
[update: due to underwhelming interest and confusion, we've stopped making these...]
this is the first release using automake, and for the first time debian and redhat packages are also available. These have not been very well tested; we would very much appreciate hearing from anyone who installs 0.924.
Among other benefits, this will allow us to make deb and rpm packages. Some files have been renamed (most notably, the gdam server from `gdam' to 'gdam-server') and there may be install problems for the moment.
the walker and gong sources. Walker performs a random walk back and forth across a buffer. The mean length for which sound is played forward or backward can be adjusted. Progress will eventually be made, but the position within the buffer can also be set by hand. Gong simulates a gong being struck. Two models currently exist, for a free-hanging gong and a clamped plate. The volume, decay rate and primary tone can be adjusted.
An exciting new addition to the cvs repository is the filter mininetwork. The mini-network provides a graphical representation of a collection of filters, and allows users to connect them together by drawing lines between them. Complex feedback loops and filters can be constructed by hand. Current collection of one-to-one filters available within the mini-network includes gain, delay, variable delay, modulator, mix, sine triangle and square waves, noise source, constant source, level meter, frequency estimator, balance, and gate. Once the kinks are worked out, it'll be included in a release.
this version might be considered unstable as several vital sections have been rewritten. Most notably, the mix source has been rewritten to prevent it from adding any latency. Realtime mode is more fully supported, and is automatically activated when the server is run as root. This might be rewritten in a safer fashion, especially if there are reports of realtime mode causing lockup. However, in spite of these changes all seems stable.
Thanks to everyone who notified us of install problems on their systems. We haved fixed a slew of minor problems with permissions and paths, and more people are reporting successful installs. Recent debian and redhat systems should have no problems. Some success has been reported with mandrake and slackware. If you've had problems or success with these or other distributions, let us know.