I have been starting a song too quick (practicing against a live version) and was asked by the band to start slower. So I decided to go back to a studio version of the song and get the tempo via tapping. I tapped in the intro of the song in 7/8 time and got around 188BPM. Then tried to play back against the tapped tempo and the metronome was too fast. I then repeated the test with ALL the different tempo algorithms, and pretty much got the same answer for each one.
I then set the tempo to what seemed to play about the right speed at 172BPM, every time I tapped in against the same song same result around 188BPM, metronome was too fast. Song was definitely not 188BPM, so I got the guitarist to tap the tempo in on his looper and he came back with 174BPM. So I manually set the global temp to 174BPM and it feels pretty much spot on.
I then retested with an empty gig file and tapping the tempo in from the drop down and got the same result tap tempo was too fast.
Has anyone else noticed this or have I missed something?
GP 4.8.2
Old Mac Pro (10.11.6)
TIA
A bit more info. I have repeated the test on a faster computer (only an i5 Win 10 Laptop). This time I launched a second instance of GP set the tempo on the first instance and tapped on the second instance and the results are much better, but with variation wider that I would expect. But certainly reduced from the Mac Pro.
The speed of the CPU should not matter — but there are multiple algorithms available for use with tap tempo. Which one are you using?
Also, HOW are you tapping?
I tired them all on the Mac Pro and a couple of different ones on the dual GP test. They all give about the same result. I did wonder if tapping to a regular beat like the metronome would actually yield better results than a complex intro at 7/8 time with multiple parts.
In the first test I used a button on a control surface, via rig manager, to a global widget, into System Actions. In the second I used a mouse and clicked on the Tap Tempo button from the GP dropdown that opens if you click in the top bar.
Although the results won’t be too bad, it’s never going to be as accurate as using a MIDI message from a controller, which is handled in real-time.
Huh? The system is just measuring the difference between taps and (depending on the algorithm) possibly averaging them in various different ways – it does not know anything about complex rhythms.
But my brain does and my taps my be not truly on the beat but on the sounds of the original recording. Against a regular rhythm and accent used in the second tests.
Yes it would appear so, but I would have expected some averaging to get me to something near the right BPM. I play against it okay, I just can’t tap it. The second test I did on a faster machine with two instances of GP showed it wan’t a GP issue allowing for a mouse click.