Dave Smith Instruments and Roger Linn have released an OS update for the Tempest drum machine, responding to a recent user petition.
The release is now available as a beta release via the DSI forum.
“I wish to thank you all for your patience, with a special thanks to John the Savage for his time and effort in helping me reproduce the bugs and test our fixes,” says Roger Linn in a post at the forum. “I am also personally grateful to Chris Hector for his hard work in implementing these fixes, as well to the guys at DSI and especially Dave, who personally stepped in to help make this happen.”
“We haven’t fixed everything on the petition, but overall I think it’s very good news and we felt that enough was done to ask for your kind help with testing, which would be much appreciated,” adds Linn.
The update addresses a variety of bugs and user requests. Here’s Linn’s summary of the changes:
1. Tempest not being able to discern between note ‘velocity’ and pad ‘pressure’…
Description: When playing back a recorded sound which has pad pressure assigned as a mod source, the sound is modulated by velocity instead.
2. Note sustain without a duration limit…
Description: This title is not quite accurate. The problem was that sometimes ADSR sounds in Beats would terminate before their full duration expired, or a performed ADSR sound would terminate before you released the pad. This was because Tempest has only 6 voices, so if all voices are playing, new sounds need to steal voices from playing sounds. Tempest does this by stealing the playing voice with the lowest VCA level, but this can result in recorded ADSR sounds with long durations being terminated before their recorded duration ends.
Status: Mostly fixed. Now held ADSR sounds—whether playing from beats or played live—should never terminate before release.
However, with only 6 voices, there will still be cases where a sound will be terminated. To help, the third post in this thread contains a PDF containing suggestions on how to set parameters in order to best optimize the voice-stealing for specific use cases.
Note: a related bug exists that haven’t yet fixed: If while recording, you hold a sound pad for more than the maximum 255 ticks (about 2-1/2 bars), the resulting duration will be (correct duration – 255).
3. “MIDI sequencer sound” prevents the corresponding internal sound from triggering, yet the ‘OFF’ setting remains broken…
Description: If “System > MIDI Polyphonic Keyboard Play > MIDI: sequencer sound” is set to a pad, that pad’s sequence playback no longer plays the internal sound, though the sound is heard if the pad is played live. This occurs even if “MIDI: Sequencer Out Channel” is set to Off.
4. Adding swing to the ‘ROLL/ARP’ function…
Description: If Swing is set to a value other than 50%, Quantize = 16th, Roll = On, and a sound pad is held in order to repeat a sound on 16th notes, the 16th notes always play at 50% (no swing), though they play back on the next loop with the correct swing setting.
Status: Fixed for Note Roll but not yet for Beat Roll.
5. The missing “Env Shape” parameter as seen on page 16. of Operation Manual 1.0…
Description: In the 1.0 manual, page 16 shows a screenshot of the rightmost Pitch Envelope screen containing a “Soft Knob 4” parameter called “Env Shape:0”. However, this parameter was never implemented.
Status: I’m afraid we goofed by putting a wrong screenshot in that 1.0 manual, and we apologize for that. Page 27 of that same 1.0 manual contains a section entitled “A Note About Envelope Shape” that describes in detail how to change the envelope shape by setting the envelope to modulate itself in the Mod Paths.
6. Basic file management…
The flash memory file system has two annoying problems:
1) The contents of each folder are not alphabetized.
2) After deleting a file, an invisible “hole” exists in the file list so that subsequent saves to that folder are placed in the position within the list of the formerly deleted file.
Status: Both problems are fixed. Now all files within each folder are automatically alphabetized, and newly saved or imported files always appear in the correct alphabetized position.
Note: this alphabetization will slightly slow the display of the files, and this alphabetizing has not yet been applied to the Load Sound soft knob in the 16 Sounds screen.
Note 2: If you use the Delete File function and delete a file, a blank space will appear in the file list within the Delete File screen where the deleted file existed, through re-entering the screen will remove the blank and display it correctly. We are aware of this and will be fixing it.
7. The envelope/LFO bug that causes the digital oscillators to quit or otherwise not trigger when modulated to, from, or through a relative ‘level’ of zero…
Description: If the modulation destination “Osc 3 Level” or “Osc 4 Level” is modulated such that the computed modulation value is zero or negative, audio output from the oscillator stops even if the computed modulation subsequently exceeds zero, fixed only when a new note is started.
8. Tempest not remembering the folder-of-origin when saving…
Description: When saving a file, the Destination Folder always defaults to the first folder of the file type, which is cumbersome. Sound files always default to Kicks folder, Beat files always default to Factory folder and Project files always default to the Factory folder.
Status: Fixed. Now the Destination Folder in Save screens always defaults to the last folder of that type (sound, beat or project) that was last saved to or loaded from.
9. ‘Key Follow OFF’ not working for the digital osc’s…
Description: Oscillator pitch is always modulated by the “16 tunings” pad pitch, even if the “Key Follow” parameter in the third Sounds page of OSC3/4 is set to Off.
10. Glide not working for the digital osc’s…
Description: Glide (second OSC screen) does not work for OSC3 or OSC4.
Status: Unfortunately, this is a capability that we originally thought could be implemented in Tempest’s internal sample playback chip but subsequently learned that it cannot. We apologize for having goofed by adding these unimplemented parameters to the UI in the first place, and have now removed them from the OSC3 & 4 screens. However, Glide does work for the analog oscillators.
11. The ever-present but still not functional Glide/Legato modes…
Description: In the Glide Mode parameter of the OSC screens, the FixRateA and FixTimeA options (legato mode) don’t do anything.
Status: Unfortunately we feel that legato glide modes on a drum machine are less important than many other priorities, and are sorry to tell you that these features will not be implemented in order for us to prioritize issues of more importance to Tempest users. We apologize for having added these unimplemented features to the UI in the first place and have removed the FixRateA and FixTimeA options from the Glide Mode field.
12. The ‘Playlist’ crash bug…
1) While synced to USB MIDI clock, if you change the number of bars in a Playlist step, Tempest freezes temporarily then unfreezes after a few seconds.
2) Editing a Playlist while playing causes timing glitches and randomly added notes.
Status: These problems were caused by Tempest needing to instantly recompute playlist data while playing. We have fixed the problem by locking out edits to the Playlist while it is playing, but edits can still be made while not playing.
And those introduced with OS 1.4:
13. Using the ‘ROLL’ function to record a part causes other pads to randomly trigger – even pads with parts already recorded on them – and these unintended triggers also get recorded…
Status: Unfortunately we are unable to reproduce this and believe it may have been fixed by fixing bug #15. If anyone can still reproduce this bug, please let us know.
14. Having ‘ROLL’ engaged also causes other pads to randomly trigger when trying to ‘Erase’ parts with the sequencer running.
Status: As with #13, unfortunately we haven’t been able to reproduce this and believe it may have been fixed by fixing bug #15. If anyone can still reproduce this bug, please let us know.
15. Using the ‘ROLL’ function to record in ’16 Levels’ mode fails to record the correct velocities: i.e. it only repeats the velocity of the lowest pad pressed…
Status: This is fixed.
16. Recording a ‘ROLL’ with a sound assigned to a voice results in the pad lighting up as if it is being triggered, yet there is no sound…
Status: Unfortunately we haven’t been able to reproduce this. It may have been fixed by fixing bug #3 above. If anyone can still reproduce this bug, please let us know.
17. There is an advanced bug that causes the pad names to randomly change when auditioning sounds; it happens most predictably when you have auditioned a sound from one pad and then try to audition a sound from another pad. When this happens the available sound folder also changes, not just the pad name: i.e. the ‘Kick’ pad, for instance, will suddenly become a ‘Snare’ pad – *which instantly loads a snare sound by the way (devastating!) – after which you can no longer access the kick folder from that pad at all, unless you “revert” the sound…
Status: Fixed. Now soft knob 3 is always “Sound Folder” and (Shift) soft knob 3 is always “Sound Type”, and there is no interaction between them so they can be independently changed at any time.
Note: There still exist some cases in which changing the Sound Folder changes the Sound Type. We’re aware of this and will be fixing it.
18. If the Tempest receives a RAM sound dump via USB, while also receiving note-on/off data on the MIDI DIN input, then USB MIDI sys-ex is not received correctly half of the time: i.e. it just stands there waiting for MIDI data in those cases.
Status: This is technically difficult to fix so we kindly ask that while receiving a RAM sound dump via USB, please do not play a MIDI keyboard that is connected to the DIN MIDI input.
19. There are still issues with loading older projects: i.e. voices dropping-out…
Status: Unfortunately we can’t reproduce any such problems. It may be that such files were created with beta versions that may contain undiscovered bugs. If you have a file saved from a released OS version that demonstrates such a bug when loaded into this beta version or 1.4.0, please let us know.
Other miscellaneous bugs:
20. When switching beats on-the-fly, the first step of the sequencer-out is consistently late; so if you’re triggering a sample loop on the one, every time you switch beats on the Tempest, it will be out of sync for the first pass…
Description: If System button > Polyphonic Keyboard Play screen > MIDI Sequencer Out and MIDI Sequencer Sound parameters are set to send a MIDI Note On message in response to an event playing at the start of the sequence, then when you change to this beat from another while playing, the resulting Note On message will be sent 28 mS late and will immediately be followed by a Note Off message.
Note that this problem does not occur for internal sounds or for beat transitions within a playlist.
Status: This is not yet fixed.
21. Pad names need to be added for all the sound folders, and vice versa. And many of the existing pad names need to be properly linked to their respective parent folders, so “Tamb” doesn’t default to “Shaker”, etc. when auditioning sounds (not to be confused with point 17. which is a completely different bug)…
Status: We feel that it makes sense for there to be more Sound Types than Sound Folders. For example, there are Sound Types for Low Tom, Mid Tom and High Tom because these are different sound pads, but they may all use the same Tom sound file at 3 different tunings so there should only be one “Toms” sound folder. Same for percussion— it makes sense to have Sound Types for Tamb, Shaker, Conga Low, Conga High, Conga Slap, Block High and Low, Cowbell and Clap because these identify the sound pad, but it is more convenient to have a single “Percussion” sound folder.
22. When triggered via MIDI, sustained ADSR sounds hang whenever the bank those sounds are from is not the currently selected bank…
Description: If an ADSR sound is triggered remotely from MIDI using System button > MIDI Remote Pad Play screen, the sound will stick on if:
1) if the sound exists in Bank B, or
2) the sound exists in Sound Bank A but the Sound Bank B button is on.
Status: This is fixed.
23. The Tempest’s timing drifts in ‘Slave’ mode when using the MIDI DIN input, gradually losing sync with the ‘Master’ clock source; and this behavior is further exacerbated with use of the ‘ROLL’ function…
Status: From our tests and feedback from user tests on the forum, there is no drift when Tempest slaves to external MIDI clock. This may have been reported when using earlier software versions than 1.4.0, where sync was significantly improved.
24. Pitch Bend not working for the digital oscillators…
Description: When using ‘System button > MIDI Polyphonic Keyboard Play screen’ in order to play Tempest’s sounds from an external MIDI keyboard, pitch bend would affect the analog oscillators but not the digital oscillators (3 & 4).
Status: This is fixed.
Also, the following bugs are fixed that were not on the petition:
Description: If Note Roll was on and you held a Bank B sound pad containing an ADSR sound, then released the pad, the sound would usually stick on. This is fixed.
Description: With Note Roll on and playing or recording a Beat, hold Erase while also holding and releasing a sound pad containing an ADSR sound. The sound will latch on. This is fixed.
Description: If “<<” or “>>” is pressed while a beat other than 1 is selected, Tempest switches to beat 1. This is fixed.
Description: In the Beat Events screen, if Shift is on, the “Insert” or ”Delete” soft key was ignored. This is fixed.
Description: If the Beat was stopped while an ADSR sound was playing, the sound would stick on. This is fixed.
Description: If Roll was on, sequence is playing and Erase is held, pressing and releasing a sound pad containing an ADSR sound would cause it to stick on. This is fixed.
Pricing and Availability
The update is available now as a beta release. See the DSI forum for details.
If you’ve used the update, let us know how it is working for you. And if you’ve got feedback on it, make sure you let DSI know via the forum.
via Chris in Berlin