Topic: Harpsichord quality

Hi,

Is there any way that future versions of Pianoteq can approach an authentic harpsichord sound? For example, compare https://www.youtube.com/watch?v=w0jStHeiH8M of a real Grimaldi vs the sound in Pianoteq. The same can be said for the Blanchet, Ruckers II and the Giusti from the Karsten Collection.

I readily admit that their playability is absolutely fantastic compared to samples but speaking as a harpsichordist (sadly without a real instrument to play at present) the sound is nowhere near good enough to record a solo album, for example.

Just taking a couple of random harpsichord videos from YouTube: https://www.youtube.com/watch?v=uczsnta4Hcg and https://youtu.be/kzYt3A_Y2SE. It would be absolutely amazing if we could approach this level of authentic sound including the soundboard resonance (and characteristic tonal release sound for example at end of the phrase here https://youtu.be/kzYt3A_Y2SE&t=205). To my mind, a truly authentic sound is only possible via something like pianoteq and not samples given the amount the denoising that destroys the sound. What's also missing in samples is that delicious onset as the plectra hit the strings.

Put another way, can we have harpsichords that sound as authentic as your pianos (which are stunning) and are certainly capable of solo album recording. I'm very happy to be involved in any discussions about how to improve them and more than happy to dedicate time to testing.

Thanks!

Last edited by bachstudies (15-03-2024 15:56)

Re: Harpsichord quality

Funny to hear that, knowing that some of my piano friends say that Pianoteq's pianos sounds not at all like a piano, but the harpsichord and the clavichord are much more realistic...

I say that not to start an argument, just to express my thought that maybe this is just a case of the grass being greener somewhere else...

Re: Harpsichord quality

The keyboard you use also plays a major role in "realism": playing a harpsichord on my usual "piano-like" keyboard is a far cry from what it feels like on a real harpsichord, let alone a clavichord!

Re: Harpsichord quality

Luc Henrion wrote:

The keyboard you use also plays a major role in "realism": playing a harpsichord on my usual "piano-like" keyboard is a far cry from what it feels like on a real harpsichord, let alone a clavichord!

You have presented a problem but not any kind of answer.

What is your suggestion here?

A custom controller?

Preference for a type of light synth action or weighted spring?

Re: Harpsichord quality

It's a matter of taste. I didn't say I had an answer either! I was just trying to explain where the problem might lie. Yes, try a lighter keyboard, why not? But personally, my Juno DS61 for example is (much) too light, so I'm going back to a weighted keyboard. Having said that, I'm not a harpsichordist but a pianist.
Just my usual 2 cents ;-)

Re: Harpsichord quality

bachstudies wrote:

Hi,

Is there any way that future versions of Pianoteq can approach an authentic harpsichord sound? For example, compare https://www.youtube.com/watch?v=w0jStHeiH8M of a real Grimaldi vs the sound in Pianoteq. The same can be said for the Blanchet, Ruckers II and the Giusti from the Karsten Collection.

I readily admit that their playability is absolutely fantastic compared to samples but speaking as a harpsichordist (sadly without a real instrument to play at present) the sound is nowhere near good enough to record a solo album, for example.

Just taking a couple of random harpsichord videos from YouTube: https://www.youtube.com/watch?v=uczsnta4Hcg and https://youtu.be/kzYt3A_Y2SE. It would be absolutely amazing if we could approach this level of authentic sound including the soundboard resonance (and characteristic tonal release sound for example at end of the phrase here https://youtu.be/kzYt3A_Y2SE&t=205). To my mind, a truly authentic sound is only possible via something like pianoteq and not samples given the amount the denoising that destroys the sound. What's also missing in samples is that delicious onset as the plectra hit the strings.

Put another way, can we have harpsichords that sound as authentic as your pianos (which are stunning) and are certainly capable of solo album recording. I'm very happy to be involved in any discussions about how to improve them and more than happy to dedicate time to testing.

Thanks!


Hello bachstudies,

Echo and reverberation are some of the most common problems when recording audio. …
It is always impossible to get the same acoustic sound as in a video because you have not been there and analyzed the acoustics. But nice sound in the videos, yes. But we can’t create the acustics there.

I'm not a pianist, although I have made over 600 recordings Pianoteq, Organteq, Harpsichord a s o. But I tried anyway to get the microphones at the same distance from my Ptq harpsichord as the microphones are in the video bwv 813 Hantai. To show that we can get a bit closer to the sound you like, hopefully.

I wonder what you say about this recording, BWV 927 ? I immediately got nine thumbs up on youtube even before I changed the acoustics.
BUT, it is of course always a matter of personal preferences in terms of listening  and musical taste. Thank you for your post, interesting thoughts.

https://forum.modartt.com/uploads.php?f...927%20.mp3

Best wishes,

Stig

Last edited by Pianoteqenthusiast (17-03-2024 15:00)

Re: Harpsichord quality

hi!

IMHO harpsichord sound in PT has not changed significantly since v4.x and it shows. If the sound is not good when using zero effect, there is only so much you can do to hide the defects later in the FX chain. I personnally use this sound only for practicing precision for my organ repertoire, in fact featuring its unrealistic qualities. The bottom line is: I don't enjoy playing with any of the harpsichord sound in PT contrary to the pianos.

Re: Harpsichord quality

I wonder if we will need a fundraising to get all client's requests.

Think about the amount of instruments Modartt have to try to keep-up the quality with the latest pianoteq main version.
Each new model they release is one more to future updated need to keep-up the quality.
So far as I know, they are the only VST company doing this.

Harpsichords are plucked instruments. Maybe Modartt made some algoritim refining for plucking strings, since they created a high quality classic guitar. If it's the case, maybe the harpsichords can be updated with the improved algoritim.
But all takes time, and time is money.

Last edited by Beto-Music (21-03-2024 17:16)

Re: Harpsichord quality

I disagree, harpsichords have improved significantly since version 4. When the Ruckers was introduced and in later versions, they received quite the overhaul improving both the typical mechanical noises as well as the pluck and the voicing between the different registers. They can certainly be improved further but they did show some progress compared to their introduction. Also I would challenge anyone to demonstrate here with comprehensive audio examples how the samples (which should still be available for free courtesy of Mr Bigwood) of the Blanchet harpsichord sound better than the current "lifeless" model on Pianoteq.

"And live to be the show and gaze o' the time."  (William Shakespeare)

Re: Harpsichord quality

Pianoteqenthusiast wrote:

I wonder what you say about this recording, BWV 927 ? I immediately got nine thumbs up on youtube even before I changed the acoustics.
BUT, it is of course always a matter of personal preferences in terms of listening  and musical taste. Thank you for your post, interesting thoughts.

https://forum.modartt.com/uploads.php?f...927%20.mp3

Best wishes,

Stig

My original post is far more about the harpsichord sound itself versus setting appropriate microphone distance or adding realistic reverb and certainly nothing to do with not using a lighter keyboard. Your example of BWV927, while well-played, still sounds a far cry from a real harpsichord (not your fault!).

Re: Harpsichord quality

bachstudies wrote:
Pianoteqenthusiast wrote:

I wonder what you say about this recording, BWV 927 ? I immediately got nine thumbs up on youtube even before I changed the acoustics.
BUT, it is of course always a matter of personal preferences in terms of listening  and musical taste. Thank you for your post, interesting thoughts.

https://forum.modartt.com/uploads.php?f...927%20.mp3

Best wishes,

Stig

Your example of BWV927, while well-played, still sounds a far cry from a real harpsichord (not your fault!).

Thank you bachstudies, for your interesting answer, and your solid arguments that explain your position.
When I read your comment, I thought no, is the sound so bad….”a far cry from a real harpsichord”, (completely different from) so I had to try again.

Let's say that this harpsichord Grimaldi now, is only a little bit far from a real harpsichord. My Grimaldi 1697 from Vasa (Vaasa) Finland. Anyway, I like this one. I have always been an experimenter, interested in experimenting with Modartt products. I think it sounds better now.
You may have to turn up the volume a bit.

https://forum.modartt.com/uploads.php?f...927%20.mp3

Best wishes,

Stig

Re: Harpsichord quality

Pianoteqenthusiast wrote:

Thank you bachstudies, for your interesting answer, and your solid arguments that explain your position.
When I read your comment, I thought no, is the sound so bad….”a far cry from a real harpsichord”, (completely different from) so I had to try again.

Let's say that this harpsichord Grimaldi now, is only a little bit far from a real harpsichord. My Grimaldi 1697 from Vasa (Vaasa) Finland. Anyway, I like this one. I have always been an experimenter, interested in experimenting with Modartt products. I think it sounds better now.
You may have to turn up the volume a bit.

https://forum.modartt.com/uploads.php?f...927%20.mp3

Best wishes,

Stig

I'm not sure this is worth going back and forth about too much longer but it simply doesn't sound close to a real harpsichord. I not trying to be snobby. It sounds "fake" digital or something a few steps up from general MIDI. But, please understand me - this is an issue with the original sounds and not anything you are trying to do with it. I just don't get how anyone, playability aside, could ever say these are close to the real deal. Maybe I'm going crazy. I know several sample-based instruments that do come very close but the online demos are generally non-flattering. Ones that come close are the Sonus Paradisi Ruckers and Mietke and the Spitfire harpsichord (after correcting the horrendous slicing/round-robin/crossfade job they did). The Orchestral Tools duo are very respectable (soundboard resonance is great!) but have a little too much "hollywood" sound to be used for solo baroque recording duties.

Re: Harpsichord quality

bachstudies wrote:
Pianoteqenthusiast wrote:

Thank you bachstudies, for your interesting answer, and your solid arguments that explain your position.
When I read your comment, I thought no, is the sound so bad….”a far cry from a real harpsichord”, (completely different from) so I had to try again.

Let's say that this harpsichord Grimaldi now, is only a little bit far from a real harpsichord. My Grimaldi 1697 from Vasa (Vaasa) Finland. Anyway, I like this one. I have always been an experimenter, interested in experimenting with Modartt products. I think it sounds better now.
You may have to turn up the volume a bit.

https://forum.modartt.com/uploads.php?f...927%20.mp3

Best wishes,

Stig

I'm not sure this is worth going back and forth about too much longer but it simply doesn't sound close to a real harpsichord. I not trying to be snobby. It sounds "fake" digital or something a few steps up from general MIDI. But, please understand me - this is an issue with the original sounds and not anything you are trying to do with it. I just don't get how anyone, playability aside, could ever say these are close to the real deal. Maybe I'm going crazy. I know several sample-based instruments that do come very close but the online demos are generally non-flattering. Ones that come close are the Sonus Paradisi Ruckers and Mietke and the Spitfire harpsichord (after correcting the horrendous slicing/round-robin/crossfade job they did). The Orchestral Tools duo are very respectable (soundboard resonance is great!) but have a little too much "hollywood" sound to be used for solo baroque recording duties.


Thank you bachstudies, for the insightful discussion. I appreciate your input on this matter, but I think it's time for me to end our discussion now.

All the best,

Stig

Re: Harpsichord quality

Needless to say, PIANOTEQ 8.2 is certainly an improvement(necessary) over version 7.0(even 8.0!) Far from muddiness as some my think
-effectively addressed a certain annoying shrilling in higher octaves. Especially, Steinways is vastly improved since version 5
Ofcourse,one should take great care in readjusting(always) the velocity curve(s). Unfortunately happens almost every time
a major version upgrade occurs...Why?
Now, about the harpsichords!
I suspect also a different algorithm is involved for harpsichord. Obviously, the piano engine is the one refined so far. And not the hpsi.
Even the release samples dont sound totally convincing. Its not just experimeting with microphone positions. In my opinion-since i also play the harpsichord,
not authentic enough(an elusive electronic sound i could describeit)
(I agree with both salvadorl&bachstudies)

Re: Harpsichord quality

Here’s just a general and quite obvious statement:

We are the most sensitive to (acoustic) instruments we know best and play daily. For example I have two acoustic vibraphones (Musser and Malletech) and I’d say I know Rhodes e-piano quite well. To my ears these are sounds and PTQ models that I could easily distinguish from real ones and to my ears there’s much for improvement. In fact I must admit that PTQ Rhodes is the model I would never use. To my ears it’s just so far from real one. Vibraphone is ok, but there’s still lot to improve. On the other hand Vibes is typically a sound which all these companies do not pay much attention, so PTQ vibes is (although not perfect) one of the best.

I can’t say much about historical pianos or hapsichords. To my amateur ears those sound nice. Obviously I do not hear the nuances like harpsichordist. I think same goes with every instrument but also goes with Modartt. They themselves have most ”ears” for pianos and thus they have put the most effort to them as well which is of course wise. With specialities like Vibes they haven’t got same ”ears” and probably ambition level is also bit lower.

This is not critic, just a thought how things go naturally…

Re: Harpsichord quality

Ecaroh wrote:

Here’s just a general and quite obvious statement:

We are the most sensitive to (acoustic) instruments we know best and play daily. For example I have two acoustic vibraphones (Musser and Malletech) and I’d say I know Rhodes e-piano quite well. To my ears these are sounds and PTQ models that I could easily distinguish from real ones and to my ears there’s much for improvement. In fact I must admit that PTQ Rhodes is the model I would never use. To my ears it’s just so far from real one. Vibraphone is ok, but there’s still lot to improve. On the other hand Vibes is typically a sound which all these companies do not pay much attention, so PTQ vibes is (although not perfect) one of the best.

I can’t say much about historical pianos or hapsichords. To my amateur ears those sound nice. Obviously I do not hear the nuances like harpsichordist. I think same goes with every instrument but also goes with Modartt. They themselves have most ”ears” for pianos and thus they have put the most effort to them as well which is of course wise. With specialities like Vibes they haven’t got same ”ears” and probably ambition level is also bit lower.

This is not critic, just a thought how things go naturally…

Probably true.
I actually really like the Rhodes models and now the Wurly and Clavs but I don't have the real deal to compare like you so I'm not so discerning - I have the luxury of ignorance!

They just sounds great comparing well with Rhodes on recordings when the presets are edited and layered with FX. Not like experiencing the full mechanical nature of the real deal like you've had.

Do try mono out to stereo FX though - if you haven't already. That makes them sound more authentic to me.

Re: Harpsichord quality

Key Fumbler wrote:
Ecaroh wrote:

Here’s just a general and quite obvious statement:

We are the most sensitive to (acoustic) instruments we know best and play daily. For example I have two acoustic vibraphones (Musser and Malletech) and I’d say I know Rhodes e-piano quite well. To my ears these are sounds and PTQ models that I could easily distinguish from real ones and to my ears there’s much for improvement. In fact I must admit that PTQ Rhodes is the model I would never use. To my ears it’s just so far from real one. Vibraphone is ok, but there’s still lot to improve. On the other hand Vibes is typically a sound which all these companies do not pay much attention, so PTQ vibes is (although not perfect) one of the best.

I can’t say much about historical pianos or hapsichords. To my amateur ears those sound nice. Obviously I do not hear the nuances like harpsichordist. I think same goes with every instrument but also goes with Modartt. They themselves have most ”ears” for pianos and thus they have put the most effort to them as well which is of course wise. With specialities like Vibes they haven’t got same ”ears” and probably ambition level is also bit lower.

This is not critic, just a thought how things go naturally…

Probably true.
I actually really like the Rhodes models and now the Wurly and Clavs but I don't have the real deal to compare like you so I'm not so discerning - I have the luxury of ignorance!

They just sounds great comparing well with Rhodes on recordings when the presets are edited and layered with FX. Not like experiencing the full mechanical nature of the real deal like you've had.

Do try mono out to stereo FX though - if you haven't already. That makes them sound more authentic to me.

To be precise I DO NOT have a real Rhodes but I am quite familiar with that instrument. I have played different Rhodes's and listened my whole life to it. In practical reasons I use nowadays Clavia Nords E-piano. That's a company which have put their soul to Rhodes sound. If you doubt that just listen to newest Rhodes of their collection (https://www.nordkeyboards.com/sound-lib...-stockholm)

But sorry, this was not meant to take this thread to discussing Rhodes or Vibes. It was just my idea how we are different in our hearing in nuances. If someone is harpsichordist and plays acoustic H. often I would trust his/her opinion about its virtual replicant.

Re: Harpsichord quality

Ecaroh wrote:
Key Fumbler wrote:
Ecaroh wrote:

Here’s just a general and quite obvious statement:

We are the most sensitive to (acoustic) instruments we know best and play daily. For example I have two acoustic vibraphones (Musser and Malletech) and I’d say I know Rhodes e-piano quite well. To my ears these are sounds and PTQ models that I could easily distinguish from real ones and to my ears there’s much for improvement. In fact I must admit that PTQ Rhodes is the model I would never use. To my ears it’s just so far from real one. Vibraphone is ok, but there’s still lot to improve. On the other hand Vibes is typically a sound which all these companies do not pay much attention, so PTQ vibes is (although not perfect) one of the best.

I can’t say much about historical pianos or hapsichords. To my amateur ears those sound nice. Obviously I do not hear the nuances like harpsichordist. I think same goes with every instrument but also goes with Modartt. They themselves have most ”ears” for pianos and thus they have put the most effort to them as well which is of course wise. With specialities like Vibes they haven’t got same ”ears” and probably ambition level is also bit lower.

This is not critic, just a thought how things go naturally…

Probably true.
I actually really like the Rhodes models and now the Wurly and Clavs but I don't have the real deal to compare like you so I'm not so discerning - I have the luxury of ignorance!

They just sounds great comparing well with Rhodes on recordings when the presets are edited and layered with FX. Not like experiencing the full mechanical nature of the real deal like you've had.

Do try mono out to stereo FX though - if you haven't already. That makes them sound more authentic to me.

To be precise I DO NOT have a real Rhodes but I am quite familiar with that instrument. I have played different Rhodes's and listened my whole life to it. In practical reasons I use nowadays Clavia Nords E-piano. That's a company which have put their soul to Rhodes sound. If you doubt that just listen to newest Rhodes of their collection (https://www.nordkeyboards.com/sound-lib...-stockholm)

But sorry, this was not meant to take this thread to discussing Rhodes or Vibes. It was just my idea how we are different in our hearing in nuances. If someone is harpsichordist and plays acoustic H. often I would trust his/her opinion about its virtual replicant.

Maybe it's partly a question of individual instruments, room acoustics and psychological bias?
A player who is used to a particular vintage of the real instruments, when heard in a particular room compare against the models played with the right velocity curve for the instrument..
So many variables. Of course it can all still get better.

Like the harpsichords two Rhodes can be tuned to sound very different. Though of course there has to be a decent amount of fundamental accuracy to the model for that to be relevant.
I don't really have your direct experience with the real thing. I have been listening to Rhodes recordings for decades - albeit mostly dripping with tape and digital delays of various forms, chorus, flange and reverb FX from plates to springs to 80s digital FX, eq and compression. Quite often listening to the instrument on its own, but that's not like experiencing in the flesh, under your own fingertips.

EDIT: Can the Rhodes get better sure. I'm not convinced  Modartt's efforts are actually inferior to the Nord examples, just different, because different Rhodes sound different.
Nord's music demos for the Rhodes Mk1 sounds in your example are probably much better sales demos to Modartt's chosen efforts in this regard be fair (that's not to knock the performers or their production, just that Nord's music examples are probably more suited to most listeners expectations)  - and that's an amazingly capable Mk1 Nord have recreated in the samples of your chosen example there, with a round sound and crystal crispy bell like highs and also the barky sound too. Hopefully the Rhodes models will get more music demos with more carefully crafted presets when the models are further adjusted/improved in the future.

Personally I find the electric pianos highly convincing and it's probably easier to replicate those than any acoustic instruments, and this is why for me they're more able to fool my ears than the (still excellent) acoustic models.

Several companies have also made pretty convincing Rhodes physical models in hardware and software, whereas Modartt are clearly head and shoulders above the competition for modeling acoustic pianos.

I would love to hear further improvements in the modelling of course.

Sorry to continue with the derailling of the Harpsichord thread.

Last edited by Key Fumbler (28-04-2024 23:24)

Re: Harpsichord quality

Key Fumbler wrote:
Ecaroh wrote:
Key Fumbler wrote:

Probably true.
I actually really like the Rhodes models and now the Wurly and Clavs but I don't have the real deal to compare like you so I'm not so discerning - I have the luxury of ignorance!

They just sounds great comparing well with Rhodes on recordings when the presets are edited and layered with FX. Not like experiencing the full mechanical nature of the real deal like you've had.

Do try mono out to stereo FX though - if you haven't already. That makes them sound more authentic to me.

To be precise I DO NOT have a real Rhodes but I am quite familiar with that instrument. I have played different Rhodes's and listened my whole life to it. In practical reasons I use nowadays Clavia Nords E-piano. That's a company which have put their soul to Rhodes sound. If you doubt that just listen to newest Rhodes of their collection (https://www.nordkeyboards.com/sound-lib...-stockholm)

But sorry, this was not meant to take this thread to discussing Rhodes or Vibes. It was just my idea how we are different in our hearing in nuances. If someone is harpsichordist and plays acoustic H. often I would trust his/her opinion about its virtual replicant.

Maybe it's partly a question of individual instruments, room acoustics and psychological bias?
A player who is used to a particular vintage of the real instruments, when heard in a particular room compare against the models played with the right velocity curve for the instrument..
So many variables. Of course it can all still get better.

Like the harpsichords two Rhodes can be tuned to sound very different. Though of course there has to be a decent amount of fundamental accuracy to the model for that to be relevant.
I don't really have your direct experience with the real thing. I have been listening to Rhodes recordings for decades - albeit mostly dripping with tape and digital delays of various forms, chorus, flange and reverb FX from plates to springs to 80s digital FX, eq and compression. Quite often listening to the instrument on its own, but that's not like experiencing in the flesh, under your own fingertips.

EDIT: Can the Rhodes get better sure. I'm not convinced  Modartt's efforts are actually inferior to the Nord examples, just different, because different Rhodes sound different.
Nord's music demos for the Rhodes Mk1 sounds in your example are probably much better sales demos to Modartt's chosen efforts in this regard be fair (that's not to knock the performers or their production, just that Nord's music examples are probably more suited to most listeners expectations)  - and that's an amazingly capable Mk1 Nord have recreated in the samples of your chosen example there, with a round sound and crystal crispy bell like highs and also the barky sound too. Hopefully the Rhodes models will get more music demos with more carefully crafted presets when the models are further adjusted/improved in the future.

Personally I find the electric pianos highly convincing and it's probably easier to replicate those than any acoustic instruments, and this is why for me they're more able to fool my ears than the (still excellent) acoustic models.

Several companies have also made pretty convincing Rhodes physical models in hardware and software, whereas Modartt are clearly head and shoulders above the competition for modeling acoustic pianos.

I would love to hear further improvements in the modelling of course.

Sorry to continue with the derailling of the Harpsichord thread.

sorry to “derail” too, but I had the opportunity to record an interesting CD recently, which combined a real piano (Steinway B) with a Rhodes sound, and after much effort and hesitation, it was the Pianoteq that was finally used, BUT via a “re-amplification”: amped on a real speaker, and re-recorded with an SM57. The difference in realism was obvious. Hard to imagine a harpsichord “re-amplification” - of course! - but that's to underline the extreme importance of the acoustics of the room in which the instrument is located; in other words, of the reverb impulses used. My 2 cents ;-)