David Baay

Usually a MIDI track is dedicated to driving a single instrument/sound. Although it is possible to have events with different channels driving different instruments of a multi-timbral synth from the same track, it makes it very difficult to keep track of things and compose/edit efficiently. For a single instrument (or a drum synth that has a different kit piece on every note number), you will sometimes have different harmonic parts/lines/voices that you'd like to isolate for editing/mixing/arrangement purposes which you can do by using multiple 'Lanes' of a track (Shift+T to show lanes). The 'Inline' PRV that your screenshot is showing is for the 'parent' track using only one lane. If you use multiple lanes, you can display an inline PRV for each. The separate 'main' PRV that you get by double-clicking that icon in the upper right corner of the clip (or Alt+3 or Views > Piano Roll View) lets you see all lanes of one or more tracks, but will not let you easily differentiate parts/lanes within a single track, except by enabling 'Hide Muted Clips' in the PRV's View menu and temporarily muting clips in lanes that you don't want to see. Nevertheless, that's where most of us do our detailed MIDI editing because it has more editing capabilities and display options. There's a lot more to understand, depending on how you ultimately choose to work, but that should help get you started, or at least know what question to ask next or what part of TFM to R. ;^)

If files are missing, Cakewalk will pop up a dialog prompting you to locate them. If you don't get any warnings, then audio clips referencing those files were deleted from the project (or never existed) before it was saved. Deleting a clip does not delete the file, so those files it should still be available to re-import unless you *did* in fact whack them via Windows Explorer. I suggest you run the Cakewalk Audio Finder utility (Utilities > CWAF Tool) and let it search your all you drives. It will show 'Missing' files that projects are referencing but were not found, 'Orphaned' files that are no longer referenced by any project, and 'Referenced' files that are used by one or more project files. After executing 'Find', you can click on a'Referenced' or 'Missing' file to see what projects are referencing it it, or click on a project to see what files it references. If clips were inadvertently deleted from one or more projects, the previously associated audio files should show as Orphaned. Depending on what else you have on your PC, and how you set up CWAF, a lot of the 'Orphaned' files may be .WAV files that are unrelated to Cakewalk. See post #5 in the below thread for a 'primer' I wrote on how I use it: http://forum.cakewalk.com/Clean-Audio-Folder-still-broken-I-see-m3023320.aspx#3023681

Sounds like the problem from back in Platinum era with 'Remove DC Offset During Record' in Preferences > Audio Playback and Recording. As I recall, it was fixed at some point, but maybe still manifests in some cases. That, or did you maybe enable the experimental 'Aggressive' ThreadSchedulingModel=3 in AUD.INI (Preferences > Audio > Configuration File)?

I usually have Follow Snap Settings enabled, but just tried fixing at 1/8, and it persisted after save/close/re-open as expected. My guess is you're using a Workspace, and need to resave it. I currently have Workspaces set to 'None' so per-project settings reign.

WASAPI Exclusive mode is probably the best place to start for onboard sound users as it's full supported by the Bakers. That said, I was initially suspicious about the possible involvement of ASIO4ALL in this issue as well, but the faulting module is Sampletank, and I would not expect that loading samples would affect the audio driver state.

I should, unless maybe you have Transport Seeking Stops Playback enabled in that same options menu.

Have not browsed extensively, and I've mostly been playing with Syntronik inserted directly rather than loading via SampleTank. Will try using ST4 some more. I only installed the VST3. As I recall, you have to unhide VST2s with the checkbox, then also un-Exclude them individually, and then you can Exclude the VST3 if you don't want to see it.

New one on me. Hard to imagine it would affect one some project unless those affected projects all shared a common ancestor, with lurking corruption. Can't really think of anything that would cause that otherwise or any procedure that might 'break the spell' other than things you have probalby already tried like archive/unarchive, delete/undo, solo/unsolo the whole project from the mix module or by quick-grouping, etc. Maybe save as a bundle, and open it from a new location...? If the problem was really precipitated by installing Studio One, I would wonder if maybe it overwrote the a version of the C++ runtime library that CbB was using or something like that You might try re-installing Cakewalk. But many of us have Studio One as well, and no problems with CbB.

As noted earlier, I finally got to the bottom of my bad-template issue by using the hardware port monitor, MIDI-OX . You could do the same by routing a single misbehaving MIDI track out to a hardware MIDI loopback path and back in to another MIDI track that echoes to the VSTi. Then monitor the MIDI IN using MIDI-OX. One way or another, I think you will find some unexpected MIDI controllers in the stream, and will need to do further investigation to figure out where they're coming from.

Your issue is pretty clearly related to MIDI controllers in some way. The OP indicated his issue affects recorded audio tracks a well, so not likely to be related.

- Assign the MIDI track output to the hardware synth port. - Connect audio output(s) from synth to audio interface/soundcard input(s). - Assign the input of an audio track to the interface input (left, right or stereo as appropriate). - Arm the audio track to record. - Start recording (R). If you want to monitor the synth's output through Cakewalk (usually the case), enable the Input Echo button the audio track.

Very unusual. Try recording a short bit of audio in Cakewalk (flatline with no input is fine, mono or stereo), drag and drop to your desktop, and then drag back in and drop it into an open area of the clips pane.

Cool. Glad to help. I got the 700NX a few years ago, replacing an RD-300s that I bought new in 1988! It was still going strong; I gave it to my son-in-law. I just had the itch for something new with more state-of-the-art touch and acoustic piano sounds. Apart from using it as my main controller all those years, I mostly used the E.P. on the 300s; the acoustic piano was pretty sad by modern standards, but the E.P was great, and I haven't found a plugin to equal the built-in chorus on that board.

Soloed something with Dim Solo enabled? Are the levels low on the master and/or hardware output bus meters?

The next thing that's going make your head explode is that it doesn't work in lanes. 😮 Just make your cuts in areas of low average amplitude, and you shouldn't have too much trouble. Splitting at zero crossings is no guarantee you won't get a pop. If the surrounding amplitude is high, stopping the driver dead at zero as it comes back to center from a high peak is a more radical discontinuity than dropping it to 0 by splitting it where the amplitude is at say -30dB.

Okay, that clarifies the scenario, and I can reproduce that specific case; - Open an existing project with it enabled. - Start a new project from a template with it disabled - Close the new project. - Re-save the existing project. It seems to only happen when a new project is started from a template. Opening some other existing project with it disables does ot have the 'override' effect that starting new from a template file does. The workaround will be to enable it in all your templates and re-save them one at a time, and you should be good after that.

Yes, it's normal. I brought it up with the Bakers early on when real-time synth recording was first introduced, and they pointed out that they can't differentially compensate audio generated by existing MIDI vs. live input MIDI when they're in a common stream from the synth. Since the intent of synth recording was to allow realtime recording from a live MIDI performance the compensation is based on that scenario.

Several options: - Set the intermedate size of the multidock larger by dragging the divider up. This will be the default size whenever it's opened. - Shift+D to maximize the multidock. - Don't ever close the PRV tab. Use D to open and close the maximized multidock, and use the PRV tracks pane to change tracks. - Undock and 'unfloat' the PRV, maximize it, and Ctrl+Tab between it and the Track View. - Define a screenset with an undocked, floated and maximized PRV (full-screen), and switch screenset with number keys. - Various combinations of the above are possible, also.

Yes, the issue with virtual ports getting mixed up may have been be resolved already, but at this time, Cakewalk will still re-assign tracks that use hardware MIDI ports to the first virtual port it finds in the project if the hardware isn't online.

By default, Cakewalk uses the ASIO buffer size when processing audio offline. Too small or too large can cause problems for some plugins. Try a moderate setting like 256-512 samples, and if you get more consistent results, you can set a corresponding non-zero value for BounceBufSizeMsec= parameter in the AUD.INI (Preferences > Audio > Configuration File) that will alway be used for offline rendering, regardless of your setting for realtime playback (BounceBufSizeMsec=0 tells it to use the realtime buffer size). Note that it's in milliseconds, not samples so 256-512 samples is on the order of 5-10ms, depending on your sample rate. I run at 48kHz, and have had BounceBufSizeMsec=20 (960 samples) set for years without any problems.

It's complicated, and I have not revisited the issue for quite a while so my recall is fuzzy, but... the 16PART+PERF mode in the RD-700 (NX in my case) allows for an external sequencer/DAW to use the sound module as a 16-part multitimbral instrument at the same time that a performer has access to an additional 5 parts by Local Control (i.e. live 'PERFormance' parts), and the patches assigned to those parts can be different from the other 16 MIDI-driven parts. To ensure the sound from playback of recorded MIDI is the same as what you heard while recording, you want to make sure your playing the MIDI driven parts from the keyboard. And there are some additional complications related to selecting patches from the DAW vs. selecting them from the front panel which get simplified by disabling the PERF mode.

First, on the RD-700, go to Menu > Utility > Rec Setting , and set 'Local Switch' to OFF. This stops the Roland's sound module from responding to 'Local' input coming directly from the keyboard. Now you'll only hear the Roland responding ot your live input if the MIDI track controlling it has Input Echo enabled. turning of Local Control is standard operating procedure for using any keyboard synth synth (i.e. combined MIDI controller keyboard and sound module) with a DAW. Another setting you'll want to make it is to go to Menu > System, page right a couple times, and change 'Part Mode' at the bottom of that page from 16PART+PERF to just 16PART . This is needed to get the piano to play back using the same sound you have selected in the front panel (or via patch change commands from Cakewalk if you get to that point) as you hear when playing under Local Control.

It's assumed that when recording a soft synth in real time, it will be driven by live MIDI input from outside the DAW, and the performer will be listening to a delayed audio metronome, so the audio is compensated accordingly. But playback of existing MIDI isn't delayed by audio output latency, so the audio resulting from that is going to end up over-compensated. Normally existing MIDI would be rendered offline by Freezing or Fast Bounce which will result in correct timing regardless of the realtime buffer setting. If you actually have a need to record soft synth audio from existing MIDI in real time for some reasone, you just need to use the lowest possible buffer size, which should always be the case in any real-time recording situation when audio is input monitored through - or originates from within - the box.

I just double-checked and cannot reproduce a problem getting the desired behavior saved in a template or saved and restored in a project created from that template or changed and resaved. One thing I noticed that could be misleading is that if you have two projects open at the same time with opposing settings, the display of the setting in Preferences may not update properly as you switch back and forth between projects, but the actual snap to zero behavior is always consistent with the state of the setting in each project when it was last saved.

- To snap to zero crossings without also snapping to a musical grid, set the snap resolution to a single sample. - To have it enabled in new projects, save a project template with it enabled, and create new projects from that template. - I don't really use the loop construction view, but 'editing' consists primarily of moving slice markers around. Your screenshot appears to be showing the waveform before the first marker. If you zoom in on a marker, you can drag it to a zero-crossing (grab the slice line inside the waveform, not the flag above it), but it doesn't appear to have a snap function or honor global snap.