David Baay

22 minutes ago, Blindeddie said:

Re-opened Cakewalk and tried loading the synth again

How did you start the project after re-launching Cakewalk? From template chosen in the New Project view of the Start Screen or by File > New or by Ctrl+N or... something else?

And what is the value of Sampling Rate under 'Default Setting for New Projects' in Preferences >  Audio > Driver Settings?

On 7/19/2019 at 5:37 AM, Sidney Earl Goodroe said:

There is no velocity control for the strings so you must record mod wheel or edit it in a controller lane. I can give you a lot of tips if you decide to get GPO 5.

Here's my tip for that:

- Record from your keyboard, playing with variable velocity as you normally would though the patch doesn't respond to it (applies to most Garritan  instruments).

- Select the whole clip/track.

- Show the Controller lane in the PRV, and add  a CC1, Modulation lane.

- Right-click the CC1 button, and choose  'Copy Selected Values to this Lane'.

The velocity values will be copied to the CC1 lane, giving you basic starting volumes for each note that you can leave as-is or use as a reference for manually drawing in more controllers for swells, etc. The same can be done of CC11, Expression as needed.

It may be helpful to layer some other instrument that does respond to velocity so you have an audible response to guide you while recording.

On 7/19/2019 at 2:29 PM, Variorum said:

I've fixed a couple errors with one of the violin patches in the Orchestral Strings block

I don't use a lot of strings in my music, but hate to think there's a flaw lurking out there.  Would you care to share the corrected SFZ?

A few additional thoughts/clarifications: 

- My main point was that distinguishing by channels makes toggling input echos unnecessary. But if you're changing channels on the controller, there's no need to have a forced  Output channel to the synth; what's important is to restrict the Input channel, so that only one is recorded and echoed at a time (in the case of non-multi-timbral synths)

- If TTS-1 or any other multi-timbral synth is using all 16 channels of one port, the other synth tracks would need to use a different port in order not to end up being layered with a TTS-1 sound. Unfortunately, most controllers do not offer multiple output ports. At that point you do have to start disabling Input Echos or muting tracks.

- Most decent controllers can send on more than one channel at a time to address the need for layering  but not all make it easy to quickly select output channels from the front panel without going into menus, and many can only send on two channels at once.  

Confirmed, and I can't imagine it was intended.  Probably a side-effect of the customizability enhancements to the Smart Tool which were primarily aimed at the Clips view. 

7 hours ago, scraf said:

Had assumed that Input Echo wouldn't have affected the signal chain, just routed the sound, but many ways to skin the cat here.

Okay. So it was more about being surprised that it didn't work than that you really need it to work. Normally it would, but, as you say, there other ways to to achieve the same end using mute/solo

7 hours ago, MusicMan11712 (aka Dr. Steve) said:

Specifically, playing/recording while using a single usb keyboard channelizing midi data going to softsynths (1) by changing the sending channel on the keyboard and (2) by changing input echo tracks on the fly.

If I'm understanding the goal, I would  suggest you just set a unique Input channel on each track, and enable Input Echo on all of them. Each track will only echo  events arriving with the matching channel from the controller. 

In general I try not to question the validity of a user's methodology when something isn't working as expected, but I am curious...  what's the use case for being able to toggle Input Echo on after the transport is running? I don't think I've ever had much occasion to do this myself, except maybe in the case that I just set up a new track and forgot to enable it, in which case I would probably be stopping and rewinding before enabling it anyway. 

On 8/14/2019 at 6:38 PM, chris.r said:

if you then decide to cancel before drop... ...  after clicking ESC key

D'oh!  ln all my decades of computer use, I never realized this was possible. I have always 'escaped' by finding a place in the UI where it's disallowed to drop the object. 

Thanks for the 'pro tip'.  😎

7 hours ago, scraf said:

via an iConnectivity Audio4+

As I've never heard of this occurring with any external audio source connected via a conventional analog or digital input, I have to suspect it's something to do with the  way this interface is routing audio data. My first guess would be that it's not honoring Cakewalk's request to open the input port when the transport is running.

You might try enabling 'Always Open All Devices' in Audio preferences if it isn't already.  And check the iConnectivity documentation or online sources for relevant information.

Is Patterning 2 being driven by MIDI from Cakewalk, or is it playing an internal pattern and synced/slaved by MIDI to Cakewalk or...? 

Some clarification of the setup might help find a workaround if not a fix.

2 hours ago, Danimal said:

I must have accidentally switched insert soft synth options (fourth icon from the "+" within the Synth Rack) from "Simple Instrument Track" to "MIDI Source".

Hence why I suggested splitting the 'Instrument' track to check functionality of both parts.

Glad you got it figured in any case. Thanks for letting us know.

+1 to just treat the Mac like a hardware synth.  Good write-up, Kenny.

Pro tip for those not used to working with hardware synths:

On the audio track that hosts the input from the Mac, Shift+click the solo button to enable Solo Override so you can solo the MIDI track(s) driving MainStage without losing input monitoring of the incoming audio.

You'll need to convert audio to MIDI first. Melodyne Essential can do a monophonic line, but Editor or Studio are needed to get polyphonic detection. 

New machine or just a new O/S on the old machine? If a new machine, I would start by checking DPC latency:

https://www.resplendence.com/latencymon

And what do CPU and disk usage look like on the Performance tab of the Task Manager (or Resource Monitor linked at the bottom of that view)?

The question is where exactly in the MIDI/Audio data chain is the break; is it in the MIDI or audio portion of the Instrument track or is it the synth itself not generating audio output? I'd start by splitting the instrument track, and looking at the MIDI and Audio track setups separately. Are you echoing live MIDI input or playing back existing MIDI clips? If you're playing live, do you depend on Always Echo Current MIDI track, or do you manually enable input echo? 

Adding via the Synth Rack continues to work fine here. Are you inserting as a Simple Instrument or separate audio and MIDI tracks? In any case, check the MIDI/Audio I/O routing of the track(s) being created when you insert the synth, and all the usual stuff (input echo is enabled on the MIDI track, keyboard/audition function in the synth UI shows output signal, etc.).

Double-clicking on a track header maximizes the track height, effectively pushing other tracks off screen. Double-click again to return it to the original height, or hit F to fit all tracks vertically. If they really got hidden somehow, hit Shift+H to show all tracks.

46 minutes ago, Corrado Prizzi said:

Yes, cheating like hell, but what isn't when sequencing. I'd say this is the most fun I've ever had playing keys.

Yes, I can see how that might be both helpful and fun. I'm a pretty decent self-taught keyboard player, but I still can't solo fast without hitting lots of 'wrong' notes.

+1   to just bringing a copy of the project over the new machine.

But to answer the question... the effect of FX inserted directly in the FX bin of each track should be applied to the exported track so long as the Track FX and Automation boxes are checked in the export dialogue.

But FX on buses to which the track is sending or outputting won't be included. To get those, you would need to solo the track, export with Source = Buses, and select the Master or some upstream bus that collects all the output from that track and its sends. But keep in mind that many FX (especially dynamic FX like compression)  won't behave the same when only one track is active.

Typically you would either export tracks without bus FX or export buses as 'stems' that include all the tracks and send FX that route that bus. Trying to do anything in between will not generally give the desired end result of having the sum of the exports sound like the Master bus.

On 8/17/2019 at 7:34 PM, craigr68 said:

I insert it into the master track, typically set ceiling to -1.0 and threshold around -10.

I've always thought the way 'Threshold' controls typically work is counter-intuitive, and not usually helped by the way the functionality is described.  A simple way to think about it is that the Threshold is simply an input gain control, and by lowering it, you're amplifying the signal until the peaks start to run into the Ceiling.  So by setting the threshold to -10dB, you're essentially boosting the whole signal by 10dB right off the bat.

Here's how the Adaptive Limiter's help file puts it:

"Only the signal above the Threshold is limited; all signal below the Threshold has a constant gain change that is controlled by the difference between the Threshold and Ceiling."

So if you just want to limit the peaks  without boosting,  leave the Threshold at 0, and lower the ceiling into the existing peaks. If you want a little boost (typically the goal of Limiting), start with a less aggressive threshold.

The way I typically use a limiter for a final Master is to set the Ceiling it at -0.3dB (as suggested earlier to void overs with MP3 compression or problems with D/A converters in consumer devices that behave badly at the limit), and push the Threshold down until the reduction meter shows the peaks are getting knocked down by maybe 3-5dB, and not continuously. Depending on the source material, and your loudness goals, you can push it harder, of course, but I usually just want to be able to bring the level up as high as possible without noticeably losing dynamic range or having the peaks sound 'crunchy'.

If  I were sending a mix to a mastering engineer, I might take the same approach, but set the Ceiling at -6dB, and only lower the Threshold to where the very highest loudest peaks are getting knocked down a couple of dB. 

Hey Chuck, FWIW, any time you run into sync problems like that, it can help to have 'Set Now Time with Full Restart' enabled in Track View Options.  The trade-off is that moving the Now time during playback will sound a little glitchy (i.e. not 'gapless').   Basically, it's the equivalent of quickly stopping and restarting  the transport when you move the Now time.

54 minutes ago, markrounds said:

The issue is only when you have a count-in measure enabled.

Yes, I understood that. Just tried it, and the issue is readily reproduced here as well. I'm guessing the Bakers will get on this pronto. 

Try increasing your ASIO buffer size. By default, Cakewalk uses this buffer size for offline processing (i.e. non-realtime), and some synths don't render properly with very small buffer sizes. If it works to increase the real-time buffer, you can set a non-zero value for BounceBufSizeMsec=0  in AUD.INI, and Cakewalk will use that value for all offline processing. I keep mine at 20  (960 samples at 48kHz).

Interesting that Melodyne are able to reproduce it. I would have thought there would be many reports of this on the forum if it affects all installations.  I use Melodyne mainly for MIDI extraction, and have probably never tried to record with pitch-correction enabled; I'll have to give it a try.

A similar thing has been known to happen with both audio and MIDI recording...  new clips are laid down a measure late in the timeline when count-in is enabled. I've never experienced it myself so have always suspected it was related to interoperability of certain audio interface drivers with Cakewalk. If you're using WDM driver mode, maybe try switching to ASIO or vice versa.

BTW, what sample rate are you running? If it's 88.2 or 96kHz, possibly some subroutine is not accounting for the doubled number of samples in a measure vs. more common 44.1 and 48...? Just a thought.