We’ve a beta of 1.1.4 ready for testing. The main changes are as follows below. We’re also going to work through some more bugs before releasing the final version 1.1.4. This is a fresh beta with lots of internal changes
Change-log since 1.1.3
Added new settings in the sampler so you can disable cross-fading when a new note is played.[New in build 06/25/2020]
Added new setting to reset the ADSR to zero when a new note is played. [New in build 06/25/2020]
Added line to settings.xml so you can disable X-Sub note wrapping. [New in build 06/25/2020]
We’ve added a standalone version of SubLab. [New in build 06/05/2020]
We’ve improved the clarity of the audio engine by changing out the up/down sampling algorithms we use. [New in build 06/05/2020]
Fixed issues with phase of the sampler, x-sub and synth engine, they should all now reset on a new note if you set “main menu -> note settings -> Rest Phase on new note: ON”. [New in build 06/05/2020]
Reworked a lot of the algorithms in-order to reduce down any clicks/pops when changing presets. [New in build 06/05/2020]
We’ve updated the sampler so that it now works with stereo samples. This required a lot of changes to the audio path, both to run an extra channel and so that we could keep the processor usage the same as before. Please check your projects to see if your tracks sound the same as with the previous version of SubLab.
Improved the security of the activation code to 2048 bit. You shouldn’t notice any difference, but please let us know if you run into problems when first loading or activating SubLab.
Added the ability to remove the main blue ADSR from affecting the sampler.
Added the ability to import a folder of samples or to drag in multiple samples at once.
Added a “Recent” sub-section to the sample file menu, so you can find recently loaded samples easier.
Added macOS runtime hardening. Again, you shouldn’t notice anything different.
While re-working the sampler we fixed some issues with clicks/resetting of the sampler (@Baphometrix this is what you highlighted in the other thread).
You can now right click on the main master meter and set it so that it displays peak metering.
Anyone who wants to beta test, just send me a mail (support@futureaudioworkshop.com) and I’ll forward on the download links. If sending a mail, make sure to use the same email address as your FAW account.
Thanks again and looking forward to getting this version checked and out!
Also, every time we post a fresh beta, we’ll delete all the older comments related to previous versions.
Install on Windows 10 in Bitwig Studio looking good so far. No problem with the new “bonus samples” menu option finding the target folder. All the menu options (new and old) look good and seem to be working. All the preset packs previously installed seem to be okay and still correctly mapped in the UI. Some of my own user presets that use specific samples seem to be loading up correctly and sounding good from prelim quick tests. I’ll test the “clicking” think in the Sample osc later tonight or tomorrow morning North America time.
One question for @Gavin_FAW… I see a new menu option about “Note Settings > Velocity on new note”. What exactly does this do?
Gavin, I’m sad to report that even with “Reset Phase on New Notes” set to ON, the Sampler will still start the actual waveform at various random phases. It will even still start on the downswing, exactly as I reported in the beta for 1.1.3 Here’s a screenshot showing the problem.
May I suggest y’all use Bitwig Studio to test your phase handling of the Sampler, so that you can see the problem right there in an oscilloscope while you’re debugging what’s going wrong?
Again, here is what I expect to see. Phase starts on the upswing, EXACTLY like the waveform’s original Sample Start position.
This is SUPER important! If I can’t trust the SubLab sampler to play in perfect phase every time with every single NOTE ON event, then it’s useless to me. I can’t risk unpredictable phasing issues with my kick or other subby sounds.
And lest you think it’s just some magic about Bitwig’s own sampler, here’s the Sampler in Phase Plant, Pigments, and a random small-time Sampler VST also rendering the phase of the same sample perfectly, each and every time…
Going to install Bitwig here @Baphometrix and try and see what you’re seeing. Currently I render out audio using a DAW and then open up in Adobe Audition and zoom in or else write out the sample values to a text file and graph them. On the scope front, I haven’t been able to find one that triggers and then allows me to zoom.
I’ve set it to re-trigg on a new note. When I rapidly play notes one after another, I see that it is resetting the phase, but it cross fades from where it is in the tail of the sample back to the start, the phase will be the same but due to the crossfade (64 samples in length), you’ll see a slightly different start waveform ( it will always be the same after the 64 samples are up). I’ve taken a video, you can see that as I trigger the notes, it resets, with a slight difference every time due to the cross-fade.
Looking at your video, the only thing I can think of is that the windows installer didn’t put the correct version on your system, I’m looking at doing a new build right now and just verifying that everything is correct. I’ll send you over the link when I have it done and thanks again for the help
@Gavin_FAW Still the same problem. Yes, it’s true that if the full release envelope gets to play out, then the phase always starts on the upswing (but is quite often weirdly folded on the first upswing). But when playing notes manually, the phase is all over the map if the Note OFF event happens BEFORE the full length of the release envelope gets to play out completely!
All of this is wonky AF and unreliable and makes the sampler effectively unusable for subbass sounds, because you can never trust the attack of the sample to be consistent and play in the same phase each and every time a new note is triggered.
Compare this with the video I’ll post immediately next, which shows Bitwig’s Sampler handling the retriggering perfectly in every single way, regardless of whether the release envelope has gotten a chance to fully play out. I’m using Bitwig’s Sampler only for convenience–it’s exactly the same in literally every other Sampler I own. No other sampler I own has trouble like SubLab has had pretty much since forever in this regard.
Here’s Bitwig’s Sampler retriggering the sample perfectly in phase each and every time regardless of whether any given trigger is cutting the release time short somehow. This is the ONLY way you can reliably use a subby/bassy sample in real productions. Otherwise you risk serious phasing issues with your kicks.
Again, this is the same for EVERY sampler I own, per earlier screenshots higher in this beta thread. Only SubLab has a problem with this. I respectfully submit that you might need to rethink your voice handling strategy and whatever envelope/retriggering interaction you’ve been using so far.
@Gavin_FAW One concrete suggestion I can make is to attempt something similar to the “Digi Mono” voice handling mode that Bitwig has. In their “True Mono” voice handling, they spawn only a single voice and re-use the same envelope, similar to what a lot of classic hardware synths do. In other words, if a new gate event is still traveling down the release portion of the envelope and you trigger a new gate, the attack envelope for the new note starts at the same amplitude as was on the release envelope at the time the gate retriggers.
But in Digi Mono mode, they instead spawn a 2nd voice immediately which always starts from the very bottom of the attack envelope. Then they quickly crossfade out the current voice and then kill it off once the crossfade is complete. So even though you’re playing in a MONO voice mode, there are temporarily 2 voices overlapping for a brief moment every time a new gate happens.
IMO this type of “Digi Mono” voice handling approach would entirely solve the problem you seem to be having with your current Sampler behavior. Not sure if JUCE supports that approach (I think you use JUCE?), but if so…
You can sit through a 34-minute deep dive video I did on Bitwig’s voice handling (or scrub through it) if you want to see some detailed examples of how True Mono vs Digi Mono works there.
There are a whole set of other problems when adding in an extra voice, it not that is overly complex to do, but from a processor usage perspective, one way around it might be to only have a two voices pre the distortion.
What I can do right now is add under the ADSR setting an extra “reset ADSR on new note” or something to that effect. This means you always start from zero when a new note on comes in.
What you describe as “digi mono” is kind of like what happens currently, just that we cross fade in the start of the sample and out the end of the sample, over 64 samples. What I can try is to only cross fade out the sample, this means you’ll have an unchanged start of the sample coming through and a faded out end. It should avoid clicks due to the start of a sample always being zero due to the “snap to zero crossing” being on.
Thanks, @Gavin_FAW. For what it’s worth, if you wonder why I keep harping on the need for a precise, repeatable phase, spend just a few minutes on this video AT THE TIMESTAMP. You’ll see a real-world example of a kick that hitting over the duration of a long 808-ish sub note, and slotting perfectly in phase with the sub note. This means no phasing and both will always sound consistent and loud and clear.
And you can skip ahead to this timestamp to see the problem when the sub and kick are out of phase alignment. So the problem is that if the Sampler itself is constantly changing the phase of the sub, then I can never keep the sub and kick aligned, and you’ll start having phasing issues like shown at this timestamp.
