Welcome to the Piano World Piano Forums Over 3 million posts about pianos, digital pianos, and all types of keyboard instruments. Over 100,000 members from around the world.
Join the World's Largest Community of Piano Lovers
It's Fun to Play the Piano ... Please Pass It On!
I often see people in this forum who say that they record something using, say, Pianoteq, and then they use that recorded midi to render that using another VST, say, Garritan CFX or some other sampled piano.
What are your experiences using this method? I would think that it is downright impossible to get a 100% equivalent recording with another VST because the touch response settings are always different across the different VSTs that you use. I can't believe people get good results on a consistent basis using this method. I'd be interested to hear about your thoughts and the methods that you use, and the experiences you've had while doing this.
You are absolutely right on the money, and I experienced this when I recently made a MIDI recording of the famous Gluck-Sgambati Melody in Dm using PianoTeq. If one could have create equivalent touch responses (using an industry standard VSTi interface), using a standard MIDI file might be a more effective method of comparing VSTIs.
When I recorded the Gluck-Sgambati using PianoTeq v6.2.2 and the Steingraeber E-272 VSTi, I used a specific velocity "touch response" preset that was developed to best match the advanced WNG carbon-fiber / nylon composite action on my Mason & Hamlin BB acoustic instrument. This was actually the first time I ever "muted" the action for a MIDI recording made by my acoustic using ProRecord - i.e. I used the QuietTime Mute Rail to stop the hammers just shy from striking the strings... so the feedback I was hearing was of the virtual Steingraeber instrument as I recorded the piece. Usually I play the acoustic and the MIDI recording is produced as a result of this dual acoustic performance. I feel the recording came out MUCH better using the virtual Steingraeber E-272 and custom velocity preset. I think this validates your point that even on the same acoustic instrument (or using the same virtual keyboard controller), the feedback the performer hears as they play differs from instrument to instrument... in effect making that MIDI file somewhat unique to the instrument it was recorded on. Since I also have some experience creating joint acoustical/MIDI recordings, this phenomenon is amplified when you consider differences in pedal travel, key travel, key length, etc.
Jason Solomonides Mason & Hamlin 7' BB 93623 Yamaha 6'1" C3 (w/WNG) D3010008 My Piano Recordings:
I would try recording the acoustic performance (with a live mic setup) while simultaneously capturing the MIDI output, since you have a MIDI retrofitted acoustic instrument. Open both the live mic WAV file and the corresponding MIDI sequence in Reaper for instance and get them in sync. Then in post production, with the VST pianos running from Reaper, you can tweak their velocity curve settings and parameters until VST playback behaviours like playing dynamics and note release balances, pedal effects, etc .are as close a realistic sounding match to your benchmark acoustic live mic recording, as you can get, then you've probably optimised the VST settings I would say, if that's the keyboard you'll always be using. You'll probably be able to do the same again for each different VST you own, and then you'll be able to switch samples and VST's with minimum discomfort in the playing experience.
If you play the acoustic on silent while listening only to the VST, you're making subconscious compromises to your playing technique to accommodate the weaknesses and shortfalls of the VST. However, if you play uncompromisingly from the acoustic, and capture whatever MIDI results, then afterwards that MIDI is how you "want to play it" and you'll try and force the VST into submission editing and tweaking those settings until it is reacting how it should to THAT performance, not to a compromised performance, which might otherwise become a bad habit you get into, like exaggerating your staccato because the releases decays are too long, etc.
wolfgangmeister, it would make sense to me that, if you heard the piano VST while playing, that the recorded MIDI would fit much better to that VST when later editing it. This would circumvent the problem that I was describing in the first place.
propianist, that sounds like what I had imagined the process to be like. Sounds like a whole lot of work, but if you do it once, you are likely set as long as you don't change equipment. It becomes clear that it's not possible to get a 1:1 equivalent. There will be differences in the new render. Great point regarding the 'compromised performance'. I think I can relate to what you say, because when I play VSTs, I find myself often making compromises because deep down I know of the limitations and issues of the VST. That's some good advice to disregard (or rather, circumvent) all problematic points of a VST.