TAL-U-No-62

A broken Juno. That's what this is. Because this synth has a critical bug in its voice allocation system.

Basically, if you use TAL-U-No-62 polyphonically with long release times, there will be occurrences where previous notes are overwritten in an unpredictable, antagonistic manner.

TAL U-No-LX does not have this problem in either RR or N voice allocation modes. In fact, I've never found a synth that intentionally implements the strange behavior TAL-U-No-62 does. Surprisingly, the predecessor version TAL-U-No-60 doesn't suffer from this issue, so long as "Save CPU" is turned off (if on, it severely compromises polyphony in a completely different manner).

Unfortunately, this freebie is unlikely to ever be updated or fixed, as it's immensely old, and it's existence stands in competition with TAL U-No-LX, despite the fact that this old freebie is not as accurate or as feature-rich.

Again, this problem only appears to affect long release times on released keys, when subsequent notes play when a release tail is active. This should not affect chord or pad type sequences, only arpeggios that use long release times as a feature. If you are using gated polyphonic sounds or monophonic sounds, this is a non-issue. So I give it 2 stars for that, but it's still a lacerated synth, and it saddens me that it was left this way.

TAL-U-No-62 had potential. It had a solid foundation, and its DNA would live on to TAL-Elek7ro (which doesn't have this bug) and eventually TAL NoiseMaker, and likely TAL's commercial synths to some degree as well. But I can only recommend it fully if you restrict it to gated mode, or in monophonic mode, or otherwise rely very little on the release time parameter.

So, polyphonic leads with overlapping release tails? Don't bother. Expect buggy, randomly cut release tails. Just use gate mode with delay/reverb, or try TAL-U-No-60.