Sanderxpander wrote:

This should be fairly easy to test definitively, by playing a single note and recording both midi and audio, then playing back the midi and recording audio again. Compare the two waveforms.

OK, I did this test, sending Reaper's output into Audacity.  With a single note, I can see that the levels are the same both ways, and it also sounds the same both ways.  I guess there must be something funny going on with my multitrack project inside Reaper.  (And yes, I'm sure Reaper is recording only midi, no audio, so that's not where the doubling up is coming from.)

Same thing here, multitrack Reaper project and clearly the playback is a lot louder when I play previous takes until it's "my turn" to play.

This is maybe related to Reaper

No sophisticated tools required, just a DAW.
The test as done by hanysz seems to show no difference. Which, tbh, is as expected.

Right.  But unexpected things are still happening to some of us some of the time.  The challenge is to figure out how to reproduce it...

hanysz wrote:

Sanderxpander wrote:

This should be fairly easy to test definitively, by playing a single note and recording both midi and audio, then playing back the midi and recording audio again. Compare the two waveforms.

OK, I did this test, sending Reaper's output into Audacity.  With a single note, I can see that the levels are the same both ways, and it also sounds the same both ways.  I guess there must be something funny going on with my multitrack project inside Reaper.  (And yes, I'm sure Reaper is recording only midi, no audio, so that's not where the doubling up is coming from.)

I don't have any sophisticated tools for settling this question, but the sound level meter on my iphone confirms that the replay is indeed louder than the original play.

No sophisticated tools required, just a DAW.
The test as done by hanysz seems to show no difference. Which, tbh, is as expected.

Sanderxpander wrote:

This should be fairly easy to test definitively, by playing a single note and recording both midi and audio, then playing back the midi and recording audio again. Compare the two waveforms.

OK, I did this test, sending Reaper's output into Audacity.  With a single note, I can see that the levels are the same both ways, and it also sounds the same both ways.  I guess there must be something funny going on with my multitrack project inside Reaper.  (And yes, I'm sure Reaper is recording only midi, no audio, so that's not where the doubling up is coming from.)

I don't have any sophisticated tools for settling this question, but the sound level meter on my iphone confirms that the replay is indeed louder than the original play.

This should be fairly easy to test definitively, by playing a single note and recording both midi and audio, then playing back the midi and recording audio again. Compare the two waveforms.

OK, I did this test, sending Reaper's output into Audacity.  With a single note, I can see that the levels are the same both ways, and it also sounds the same both ways.  I guess there must be something funny going on with my multitrack project inside Reaper.  (And yes, I'm sure Reaper is recording only midi, no audio, so that's not where the doubling up is coming from.)

Nick

This should be fairly easy to test definitively, by playing a single note and recording both midi and audio, then playing back the midi and recording audio again.

Have done a bit of that reverse engineering.
Testsignal: A dry Steinway B Prelude with Limiter off. (a 1kHz sine would have been easier, but not available inside Pianoteq)

Recorded two notes c3 with different fixed velocities 127 and 64 inside Pianoteq . With audacity I measured the peak-levels of the wavs. For each stereo channel separately:

At velocity 127:
L: -6.8 dB  R: -9.9 dB

At velocity  64:
L: -18.6 dB R: -22.1 dB

Delta:
L: 11.8 dB  R: 12.2 dB

Result_1: At half velocity 64 the output level is ~12 dB less.

Pianoteq's VU meter doesn't have a dB-scale, but I found that this 12 dB volume drop was equivalent to 4 marks on the green bar.

Result_2: One grid mark on the VU meter is probably equivalent to 3 dB.

No, it's not an illusion.  I'm trying to do some multitrack recording now, and recorded-me is consistently drowning out live-me.  Playing with levels in the DAW to try to get a reasonable balance, it seems like playback is about 8db louder than live playing.

Any chance you're recording both MIDI and plugin audio output in your DAW, and hearing doubled output on playback?

Starting with an empty track does the real-time output level differ depending on whether the tranport is running or not or the track is armed/unarmed for recording?

I'll do a quick double-check but have not noticed a level difference between MIDI recording and playback; most likely a DAW configuration issue.

EDIT: Got distracted from my mission and kept playing for a while (as usual), but did not find any difference in played vs. recorded velocity or audio output level with the Pianoteq 8.05 VST3 in Cakewalk by Bandlab. I checked using both track meters in the DAW and 3rd-party plugins as well as my ears.

My understanding is that a single midi note played live or played back has the same level in the VU meter.

Had access to Pianoteq a few minutes ago and verified it. Pianoteq records the incoming midi notes with the incoming velocity. Both - playback and liveplay - are remapped by the velocity curve and then send to audio out. Both audio streams are having an equal volume level (VU meter).

This had been Pianoteq Standard v8.0.4 standalone and not the DAW plugin.

hanysz wrote:

..., it seems like playback is about 8db louder than live playing.

Unfortunately Pianoteq's VU-meter has no dBFS scale (decibel Full Scale) to compare the levels of playback and live playing. - Or is it a 3 dBFS or 6 dBFS "grid"?

My understanding is that a single midi note played live or played back has the same level in the VU meter.

This should be fairly easy to test definitively, by playing a single note and recording both midi and audio, then playing back the midi and recording audio again. Compare the two waveforms.

..., it seems like playback is about 8db louder than live playing.

Unfortunately Pianoteq's VU-meter has no dBFS scale (decibel Full Scale) to compare the levels of playback and live playing. - Or is it a 3 dBFS or 6 dBFS "grid"?

My understanding is that a single midi note played live or played back has the same level in the VU meter.

Sorry if this has been discussed before, but I haven't been able to find anything related.  From first purchasing PTQ I noticed that playback was always at a higher volume. Is there any way of altering playback volume so that it matches recording volume?

This might actually be an illusion. I used to think this was the case too, but after some experimentation found the extra mechanical noise from my Kawai keyboard (which is substantial) was causing me to crank the volume when playing. Then when listening to the playback without all that mechanical thudding, the recording sounded significantly louder than before. If using well isolated headphones the effect should be a bit less.

Basically the effect of listening to music in a noisy environment vs. a quiet environment. I find this also effects my preferred level of "key noise" and such within Pianoteq.