David Baay

I'm still seeing the issue that audio is laid down one buffer late when Punch In starts at time zero. Punch in a measure later, and it's good. I have not yet tested to see what happens close to - but not at - time zero.

Thanks for all the info, Mark. That's very helpful. I see now that based on this, if you have redundant tempo entries and 'Set' a later point, the redundant entries get deleted which shouldn't be of any consequence in most cases. I'll have to play around and see if there's any real-world scenario where it's actually problematic.

Hi Mark, can you expand on this? How exactly do they differ? In particular, I would expect the lists to be the same, and the envelope to be the same as the old 'chart' view - at least so long as the changes are all jumps.

Thanks for the clarification. i see the problem now. I had not noticed there was a tempo at 15:04 before you set 18:01. I have not seen this issue - possibly because I've only been setting whole measures recently. I'll have to try setting some beats around measures and see if I can reproduce it.

After seeing both you and Tom report that punches are early vs. my experience that they're late, I did another test with punch-in starting later in the timeline instead of at 1:01:000 as I had done initially, and got the same 'early' punch result as you. Will need to verify both cases work correctly when Noel gives us another update.

You do this now by inserting a time signature change, and then another one to change it back after the measure(s) with added beats.

I know SM/BAN inside and out, but do not understand what either the documentation or you mean by "create" or "add" new bars or beats. The sentence after the one you highlighted more accurately describes what it does. Nothing is created except tempo changes, and those changes result in existing bar and beat lines moving relative the absolute timing of audio transients or MIDI events to align the timeline to the performance. The same dialog does let you change the time signature of a measure, which can effectively 'add' beats to a measure, but that's a separate operation from 'Setting' a measure or beat to Now. And so far as i know, that also works as it always has. If this isn't just a misunderstanding of the awkwardly-worded documentation, I'll need a specific example of what you're trying to accomplish in order to understand whether something is not working as it always has.

I don't really understand what you're saying here, but what you showed in your original post is working as designed, and as it always has. Cakewalk modifies the most recent tempo to make the specified beat land on the current Now cursor position, and inserts a matching tempo at the point to serve as a reference point for the next Set.

Loop back a send from your metronome bus via physical patch from an output to input (preferably separate channels from your main/monitor out and instrument inputs). If your manual offset is set correctly, the recorded clicks will land exactly on bars/beats to the sample when compensation is working.

For me, compensation is now correct without punch enabled, but audio is still late by the ASIO buffer size when punch is enabled. EDIT: When loop-recording with punch enabled, compensation of all takes is correct, and loop recording alone is also good. Only punch alone still has an issue.

A Fermata is usually a pause of all instruments, and only requires a single, lower tempo to the achieve the desired pause. Any changes to individual notes within that pause can and should be done with duration. Excessively dense tempo changes can cause problems with FX processing, and other playback anomalies. Give me any sample project and I can reduce the tempo map to have no more than one, fixed tempo per note start that will be indistinguishable on playback from the one with curves and nodes between note starts. I generally perform tempo variations in real time rather than drawing them, but I did enough editing in the old tempo view to know that it was quite awkward because you couldn't drag or 'sculpt' existing tempo changes; you either had to re-draw them completely, or painstakingly click above or below a tempo with snap enabled at the correct resolution to prevent inserting a new one, or manually edit values in the list. The new envelope implementation should be much more user-friendly for this purpose once you get used to it.

You had me going there for a minute, @Noel Borthwick, but I tested and then checked the flow chart to confirm: Hardware Inputs come in after Input Gain. Only inputs from audio clips and soft synths are affected by Gain.

Late to the party. For future reference: Select the clip. Edit > Select by Filter (uses the same filter dialog as Find) Shift+Drag the selected controllers back to the original track with Drag and Drop mode set to 'Blend'.

M:B:T time 1:01:000 is locked to SMTPE 00:00:00:00 in Cakewalk, and cannot be changed. If the song starts with a partial measure of "pick-up" notes, then the first downbeat will have to start on 2:01:000. For this and other reasons, I do not find automatic tempo detection to be very useful. I've posted versions if the following manual procedure using Set measure/Beat At Now (Shift+M) many times, and still find it give the most predictable and precise results. Easier to do than to describe: Note that this was originally aimed at aligning the timeline to a MIDI clip, but tabbing to transients in an audio track works the same as tabbing to MIDI events, and it's not necessary to show transient markers in order for the tab function to work; just read the word 'event' as 'transient'. In some cases, you might want to tweak the Now time position if the transient marker is not accurately placed. The main thing to understand is that Set Measure/Beat At Now is basically telling SONAR to set the previous tempo to make the specified Measure and Beat of the timeline fall on the absolute time at the current Now cursor location: the playback of audio is not affected, and MIDI event start times and durations are adjusted automatically so that they also play back with the same timing rather than following the tempo changes. You're effectively stretching/compressing the timeline to fit the existing playback timing. 1. Trim and drag the clip to align the first event wherever it should be in the timeline. 2. If that time is not 1:01:000, use Set Measure/Beat At Now (Shift+M) to "pin" that first beat so that it becomes the reference point for determining tempo. 3. Count out several measures listening to the clip (or go to the last event if the clip is short) set the Now time at the beginning of that event by tabbing to it, and use Shift+M again to set the correct measure and beat in the timeline to the absolute time of that event (i.e. "Now".) 4. Cakewalk will alter the tempo at 1:01:000 (or the first point you "pinned") to make the timeline match the clip without altering the playback timing of the clip (or any other clip in the project), and add a matching tempo value at the beat you set to serve as a reference for setting subsequent beats. 5. If the clip was recorded to a click, and/or was quantized, that may be all you need to do; if not, you can set additional beats every few measures or every measure, or even within measures or beats (note that fractional beats are entered as decimal values not ticks, so 02:480 is 2.5). 6. If the clip didn't start at 1:01:000, go to the tempo track, and change the tempo at 1:01:000 to match the first tempo SONAR inserted (this gets trickier in a project with mixed MIDI and audio) That's about it. For a very long clip, I recommend setting the first point at 8 measures or so to see what beat the last event should fall on. Then undo the 8-measure 'Set' , and set that last event to establish an overall average tempo, and roughly align the whole clip so that it's easy to see what beats other events should fall on in the middle of the clip, and decided how many you need to snap to tighten up the timeline match. Setting the first 8 measures is just an interim step to figure out how long the clip is without listening and counting 100+ measures or whatever it is.

As I have posted many times, it is not musically necessary to have tempo changes in between notes. Only one, fixed tempo (i.e. a flat line) is needed to perfectly determine the time between the start of one note, and the start of the next, and the type of music has no bearing on this. Classical music is no more complex or sophisticated than any other type of music in this respect.

Probably too late to make it into the paper, but: My number one favorite feature of Cakewalk is Set Measure/Beat At Now, which allows me to freely record MIDI without a click, and with as much rubato as I want, and then tell Cakewalk precisely where the bars and beats fall after the fact. This is useful for: - Eliminating the tendency to perform poorly and make un-natural tempo corrections when trying to follow a click. - 'Rescuing' an improvised performance that has great overall feel with a few glitches, hesitations, restarts, etc. - Tightening the timing of a piece that has a rubato intro/outro with a steady tempo in the middle. - Quantizing within measures to get a tight rhythm while still allowing different sections to have different or gradually changing tempos. - Syncing overdubbed parts to a freely played performance. - Selectively softening/flattening the tempo variation in a rubato performance after the fact. - Converting a freely played performance to notation.

Track Gain only affects input to the track from recorded/imported audio clips. Live input level has be set at the interface, either by a hardware control or a software mixer/console app that comes with it.

Sounds like Cakewalk is sending MIDI Master Volume (CC7) , lowering the output level from the Clavinova that resets when you power cycle it. By default, the Volume control in a MIDI track will be disabled as indicated by parens around the default value of (101). Moving the volume widget will immediately enable it and the new value will be sent every time the project is launched, and every time playback is started. I usually set Volumes to 127 on all MIDI tracks driving external hardware to start, and only lower it if the resulting input level is clipping.

So long as output from Cakewalk is not going to the same destination as output from the hardware mixer, latency should not be a problem. My understanding is that OBS is a live-streaming app, presumably being used to stream the performance to remote listeners via the Web.

Input gain does not affect live signals, only recorded/imported audio. Input level to CbB has to be set at the interface or in a software mixer that sits between the hardware and CbB - many hardware interfaces have a software console that does this. In your case where a hardware mixer is the interface, managing levels to the ASIO channels may be a hardware function only. That said, if input Gain is left at unity in CbB, the sound of live input monitoring should be identical to what you get when playing back a recording - with or without FX plugins operating on the output. Of course, It will definitely sound different from monitoring the mixed output from the mixer to PA/headphones, but the output from Cakewalk should sound the same whether live or recorded, using I/O level controls in the FX chain and track output Volume to mix

If you really need to split out the MIDI parts, you can use Process > Run CAL > Split Notes To Tracks. Just be sure to set the starting track to the next number after the last existing track in the project. A better option if you just want separate audio tracks, and are using a multi-out drum synth is to create a separate audio track for each output of the drum synth, and freeze/bounce those. If your drum synth doesn't have separate outputs for each kit piece or doesn't have as many as you need (e.g. to break out percussion pieces), you can apply a drum map to the MIDI track output, and use it to solo kit pieces and bounce them to tracks one at a time. I do often like to have the MIDI parts separated to facilitate editing, but I usually put them in separate lanes of a single MIDI track.

I have no familiarity with Opus, but it sounds like Opus is echoing MIDI input to a virtual Out, and your track Input is set to All Inputs. Set the Input of the track to the specific port and channel your keyboard is sending on, and disable MIDI Out from Opus if you're not using it.

I did a quick test, and could not repro the 'soundless note' problem with either input-monitored hardware or a soft synth with an arp and input quantizing enabled. A couple other things I thought of that could have a bearing; - Are there any MIDI FX plugins in the project? - Do you have a non-zero 'Timing Offset' entered under Audio > Sync and Caching?

The purpose of punch recording is to ignore input before and after the punch points, so that's expected. Personally, I would just record the MIDI in a new lane in sound-on-sound mode without punch in and edit as necessary at the overlap point. As for a note played after the punch point not sounding.... - Does it happen with either or both the arp and input quantizing disabled? - What project tempo, time signature and quantize setting? - Is the sound source a soft synth, or input-monitored hardware? - If a soft synth, can you reproduce it with some Cakewalk-bundled synth? - Does it matter how close the the first note is to the punch-in point? - How long is the punch range? - What interface, driver mode, and buffer setting?

The only crash I have ever had with CWAF was related to having a project folder nested within another project folder. It was a long time ago so I don't remember the exact situation, but IIRC, I found it by watching the status bar at the bottom of CWAF to see what folder it was scanning when it crashed. It may also help to exclude as many non-project-related folders as possible. Here's a post about how I set it up and use it: