There are a few small weird things that are annoying, no doubt, but I must say that I'm impressed that I spent two days building Sampler Instruments from things I liked in the Spitfire PianoBook collection and I had not a single crash or other work-stopping or Instrument-wrecking issue, other than the small teething pains and my unfamiliarity with the new features.
I did that mostly as a way to get to know Sampler without touching my existing EXS library since I'm not 100% confident that I can go back to v10.4.8 and EXS and load Instruments that have been edited in Sampler with no issues. It appears to work, but how the heck can EXS load Sampler Instruments and completely ignore all the new parameters without exploding? Fingers crossed that it really is as safe as it seems.
I did notice a few things:
- EXS would let you have Zones with overlapping velocity / key ranges within the same Group, but Sampler does not. No way, no how. None of the toggles in the Zones menu have settings which allow this.
I did some testing, and what happens is that upon importing an EXS Instrument that's set up like that, Sampler will create new Groups containing the Zones that formerly overlapped.
I understand that preventing overlapping Zones within the same Group will make things less confusing for new users, but what effect this will have on Instruments that have both overlapping Zones and round-robin Group arrays will possibly be confusing even to more experienced users. So far, everything is sounding correct, but what appears to happen is:
When you have an EXS Instrument that uses overlapping zones and round-robin arrays, upon import into Sampler all overlapping Zones are separated into new Groups, and everything sounds correct. But now your round-robin array has doubled in size, so you effectively have two arrays working simultaneously. Everything is fine... until you encounter a situation where one array advances to the next RR but another array does not. This can happen in a variety of ways (usually while editing), and when it does what was formerly samples 1+2 layered, RR-ing with samples 3+4 layered, will now play as samples 1+4 layered, RR-ing with samples 2+3 layered. It's not the biggest disaster in history, but it may cause results that sound slightly different in Sampler than they did in EXS - however it seems this is only while you're editing for the first time and messing around with Group RR array assignments. Once you save and re-load the Instrument the arrays all reset and samples play in the correct order.
To test this, I made an Instrument in EXS with four samples of me saying the numbers 1 through 4, with one RR Group containing sample 1 at velocities 1-127 and sample 2 at velocities 64-127, then another RR Group containing sample 3 at velocities 1-127 and sample 4 at velocities 64-127 - all of them on the same key range. Play softly and you hear samples 1 and 3 alternating; play hard and you hear samples 1+2 layered, alternating with samples 3+4 layered. It all works as expected, using only two Groups in a single 2-way RR array.
When that Instrument is imported into Sampler, two new Groups are created to hold the two highest-velocity samples (whose velocity ranges overlap with the two lower-velocity samples) and at first it sounds as expected, but when you inspect the Round Robin assignments in Groups View, you see that the lower velocity samples are in the two original Groups, and their values for "RR Cycle" are "B.1 (rr1)" and "B.2 (rr2)" as expected. But in that column the values for "RR Cycle" for the two newly-created Groups are somehow set to "Move to new Cycle" - which is not supposed to be an actual value that can be selected, it's supposed to act as a command to move Groups around the RR arrays. So, somehow a command is being stored as a parameter value! This seems.... not right. Everything works as expected until you need to start editing the RR arrays, and then you'll have to manipulate the assignments so you have the Groups for the two low-velocity Zones set to "A.1 (rr1)" and "A.2 (rr2)" and the Groups for the two high-velocity Zones set to "D.1 (rr1)" and "D.2 (rr2)" or similar. Then things do work as expected.
But during the process it is entirely possible (quite easy, actually) to scramble things up so that on hard velocities you now hear samples 1+4 layered, alternating with samples 2+3 layered, as described above. Upon changing to a different Instrument and re-loading the original, the RR array resets itself and behaves as expected, but dang if it won't be confusing during the initial editing process.
Not loving this change at all... but as long as it doesn't break any of my old EXS Instruments I guess I'll live. Of course, once you're forced into putting overlapping Zones into separate Groups, you do get the additional control offered by separate Groups.
- The ancient problem of RR arrays not advancing to the next Group until all sustaining notes have been released is still there - at least it's no worse than before but dang it would have been nice if this had changed so that RR Groups would advance on note-on instead of waiting for all notes to be released, so the behavior would be as it is in Kontakt and other samplers.
- I wish there was a way to have "Optimize Sample Gain" in the Sample Editor pane default to OFF. This feature analyzes the level of the selected Sample and edits the Zone Level parameter to achieve -12dbLUFS. If you have created an Instrument without using the "Optimized" drag-n-drop method then the Zone Levels will default to +6.00db (instead of 0db, which is a little annoying), but when you select a Zone and view it in the waveform view at the bottom, and look at the menu, "Optimize Sample Gain" is switched ON, but hasn't done anything yet. If you toggle it off and then back on again, Sampler will edit the Zone Level parameter after analyzing the sample. So it's safe... ish. It won't change things unless you expressly tell it to, but it would be more clear if the toggle defaulted to the off state when it has not been used yet, since it currently appears that it HAS changed the Zone Level when it hasn't.... yet.
Sure, it's really great to have this option, but often I really REALLY want precise manual control over Zone Levels, with no chance of Sampler trying to help - and the fact that the menu toggle appears to be ON when it hasn't done anything to the Zone Levels is confusing.
- AutoMap seems to work amazingly well, even its Pitch Analysis mode seems much more accurate than that in Redmatica KeyMap (even setting fine-tune parameter values in the Zone List to accurately re-tune every sample, which is amazing), but it always centers the root key in the middle of the auto-created Zones. For me, best practice is to have the root key at or near the upper limit of each Zone, to avoid having any samples transposed upwards by a large amount. I really wish there was some user control over this behavior - like a preference called "Root Key Placement in Auto-Mapped Zones" that would let you select how many semitones from the upper limit of each Zone the root key would be located, perhaps with a checkbox for "Centered" that overrides the numerical parameter.
This way, you could leave the checkbox for "Centered" turned on and the behavior would remain as it is currently, but turning that checkbox off would enable a numerical field called "Set Upper Zone Limit XXX semitones from root key when Auto-Mapping", with the upper extreme value being "Upper Zone Limit equals Root Key" and lower values would effectively move the Upper Zone Limit up the keyboard by semitones. As it currently stands, if you want root keys to not be centered, you've got to do some serious manual editing.
(continued next post)