Topic by revp01

  • Avatar
  • revp01
  • Member
  • Offline
  • Registered: 29-12-2016
  • Posts: 11

Topic: Chord Display Accuracy - Request

I'm loving this app and have been a user since Pianoteq 5. I only have a few complaints here & there with this tech. Here's one:
Is there any way that the Pianoteq Team can give us a firmware update so Pianoteq can display chords accurately. Many of the chords I play don't even register. This has been an issue in all versions of the app that I've owned. GarageBand & MainStage seem to be doing a better job than Pianoteq. I feel this would enhance the user experience in the stand alone app, which uses less CPU resources.  Would be great and even better for the coming iPad app. 

Anyone else would love to see this update?

Is there a reason or limitation on why the chord display works the way it does?

Reply by LouisFlute

  • Avatar
  • LouisFlute
  • Member
  • Offline
  • Registered: 14-12-2022
  • Posts: 3

Re: Chord Display Accuracy - Request

Do you please have an example of a chord that does not register accurately as I would be interested to see what the problem is.

I have noticed that if you play a chord consisting of three notes, no chord will be displayed if any of the three notes are actually the same note but just played in a different octave.This is not always immediately obvious if you are playing an arrangement for both hands.

Last edited by LouisFlute (14-12-2022 23:45)

Reply by davidizquierdo82

  • Avatar
  • davidizquierdo82
  • Member
  • Offline
  • Registered: 11-02-2015
  • Posts: 310

Re: Chord Display Accuracy - Request

revp01 wrote:

Many of the chords I play don't even register

Same here, I've so many times though about posting this very same "complaint", I'm glad you did.

P85>Kawai CA97>Numa XGT>FP90X>LX706
Pianoteq 8 Pro (all instruments) + Organteq 2
i7 4790K W11 64bits + UMC1820 + MTM + DT770 pro X
http://youtube.com/DavidIzquierdoAzzouz

davidizquierdo82's Website

Reply by revp01

  • Avatar
  • revp01
  • Member
  • Offline
  • Registered: 29-12-2016
  • Posts: 11

Re: Chord Display Accuracy - Request

I will post a pic as soon as I reconnect o my keyboard controller. It happens a lot when I'm playing gospel or jazz chords.  Lot's of 9, 11, 13, & diminish chords.

LouisFlute wrote:

Do you please have an example of a chord that does not register accurately as I would be interested to see what the problem is.

I have noticed that if you play a chord consisting of three notes, no chord will be displayed if any of the three notes are actually the same note but just played in a different octave.This is not always immediately obvious if you are playing an arrangement for both hands.

Reply by revp01

  • Avatar
  • revp01
  • Member
  • Offline
  • Registered: 29-12-2016
  • Posts: 11

Re: Chord Display Accuracy - Request

Look here.  https://imgur.com/a/1bVEXIu

pic 1: Pianoteq no chord display
pic 2: MIDIculous (with Pianoteq as plugin) shows chord with accuracy
pic 3: Logic Pro (with Pianoteq as plugin). shows chord but not accurately

MIDIculous is a software I picked up recently because it does a fabulous job of helping me write chord charts for others who perform with me at times. This was one of the features I was hoping Pianoteq could get figured out. Mainly because of its simple interface and its lightweight use of the CPU.

Clearly Logic Pro tries but it's lacking the correct info to follow chord identification by the bass lines.

Let me know your thoughts.

Reply by julien

  • Avatar
  • julien
  • Admin
  • Offline
  • Registered: 19-04-2007
  • Posts: 1,047

Re: Chord Display Accuracy - Request

Please post more chords that are not properly identified, we will look at it.

julien's Website

Reply by miiindbullets

  • Avatar
  • miiindbullets
  • Member
  • Offline
  • From: Idaho, USA
  • Registered: 31-03-2022
  • Posts: 251

Re: Chord Display Accuracy - Request

I've noticed it won't register any extensions beyond the 9th. Here's a Cm11, for example, which registers in Chordie but not Pianoteq:

https://i.imgur.com/te8j3MF.png

YouTube

miiindbullets's Website

Reply by davidizquierdo82

  • Avatar
  • davidizquierdo82
  • Member
  • Offline
  • Registered: 11-02-2015
  • Posts: 310

Re: Chord Display Accuracy - Request

Here's a quick one I've just come accross, I know this might not be a listed chord by itself (although it's still a chord) but PTQ could display the closest/strongest option at least..

https://thumbs2.imgbox.com/ff/f9/tuIqZMRm_t.png

P85>Kawai CA97>Numa XGT>FP90X>LX706
Pianoteq 8 Pro (all instruments) + Organteq 2
i7 4790K W11 64bits + UMC1820 + MTM + DT770 pro X
http://youtube.com/DavidIzquierdoAzzouz

davidizquierdo82's Website

Reply by budo

  • Avatar
  • budo
  • Member
  • Offline
  • Registered: 30-11-2016
  • Posts: 1,394

Re: Chord Display Accuracy - Request

davidizquierdo82 wrote:

Here's a quick one I've just come accross, I know this might not be a listed chord by itself (although it's still a chord) but PTQ could display the closest/strongest option at least..

the other examples are totally standard (min11ths) that PTQ should certainly be able to detect.  but what are you expecting for this one?  i can think a few different ways to spell it, but none of them are very useful and they all imply rather different contexts.

@youtube
@facebook

Reply by davidizquierdo82

  • Avatar
  • davidizquierdo82
  • Member
  • Offline
  • Registered: 11-02-2015
  • Posts: 310

Re: Chord Display Accuracy - Request

budo wrote:
davidizquierdo82 wrote:

Here's a quick one I've just come accross, I know this might not be a listed chord by itself (although it's still a chord) but PTQ could display the closest/strongest option at least..

the other examples are totally standard (min11ths) that PTQ should certainly be able to detect.  but what are you expecting for this one?  i can think a few different ways to spell it, but none of them are very useful and they all imply rather different contexts.

If I'm right, this chord is mostly a Cmin9, there's an additional C# which makes no other combinations anyway, so it can be omitted for the case, I'd be happy to see Cmin9 rather than nothing at all..

Alternatively, it could show something like +Cmin9 inferring that there's some useless note/s added to the chord, just an idea

P85>Kawai CA97>Numa XGT>FP90X>LX706
Pianoteq 8 Pro (all instruments) + Organteq 2
i7 4790K W11 64bits + UMC1820 + MTM + DT770 pro X
http://youtube.com/DavidIzquierdoAzzouz

davidizquierdo82's Website

Reply by budo

  • Avatar
  • budo
  • Member
  • Offline
  • Registered: 30-11-2016
  • Posts: 1,394

Re: Chord Display Accuracy - Request

davidizquierdo82 wrote:

If I'm right, this chord is mostly a Cmin9, there's an additional C# which makes no other combinations anyway, so it can be omitted for the case, I'd be happy to see Cmin9 rather than nothing at all..

Alternatively, it could show something like +Cmin9 inferring that there's some useless note/s added to the chord, just an idea

the problem is it's "mostly" a lot of things because it's so dissonant.  not that this is a bad thing, au contraire it's quite a nice chord.  there are many ways to spell it, especially if you consider other roots not in the chord.  it could be read over C, over Eb, over A, over B, ...

to see more examples of stuff like this, just take almost any chord out of a piece like Catalogue d'oiseaux by Messiaen or something similar. 

personally i would rather not have PTQ try to pin it down by eliminating notes.  i would rather it just say nothing.  but if it's a more typical chord like the min11ths it should be able to say something.

@youtube
@facebook

Reply by TheodorN

  • Avatar
  • TheodorN
  • Member
  • Offline
  • Registered: 19-11-2022
  • Posts: 60

Re: Chord Display Accuracy - Request

Could this be Cmin9b chord? In jazz, sometimes chordal tones are raised, and sometimes lowered, though I'm not at all sure what rules apply, when to play a flat ninth, sharp eleventh, flat thirteenth, and so on.

Regarding chord identifying in Pianoteq, is it possible that a solid chord must be played for it to work, that is, not arpeggiated? Come to think of it, how can a software recognise notes played one after the other, as a chord? We know music is dealt into phrases, and there are rests between the phrases, that's how they're dealt up.

I mean, how can some program know when a chord ends, or begins? If more than a quarter rest is between the notes, and less than half a rest (the rest equal in time to a half note)? Just wondering.

Reply by groovy

  • Avatar
  • groovy
  • Member
  • Offline
  • Registered: 12-06-2009
  • Posts: 724

Re: Chord Display Accuracy - Request

miiindbullets wrote:

I've noticed it won't register any extensions beyond the 9th. Here's a Cm11, for example, which registers in Chordie but not Pianoteq:

Have you seen, that your example is working now in Pianoteq 8.0.4?

https://i.postimg.cc/J085XypQ/Cm11-chord-detection-2.png

Reply by budo

  • Avatar
  • budo
  • Member
  • Offline
  • Registered: 30-11-2016
  • Posts: 1,394

Re: Chord Display Accuracy - Request

groovy wrote:
miiindbullets wrote:

I've noticed it won't register any extensions beyond the 9th. Here's a Cm11, for example, which registers in Chordie but not Pianoteq:

Have you seen, that your example is working now in Pianoteq 8.0.4?

https://i.postimg.cc/J085XypQ/Cm11-chord-detection-2.png

nice.  how about a 13th

@youtube
@facebook

Reply by davidizquierdo82

  • Avatar
  • davidizquierdo82
  • Member
  • Offline
  • Registered: 11-02-2015
  • Posts: 310

Re: Chord Display Accuracy - Request

@budo @THeodorN

Thanks for the suggestions, I wasn't clear on the chord either but felt like it should be shown by PTQ anyway, that's the advantage of IA after all I guess..

Looking out on the internet I've found a simple website that showed all kind of shords, entered the notes of mine above and it showed this:

https://www.scales-chords.com/chord/pia...3%2Badd(7)

https://i.ibb.co/xzdYDy9/Chord.jpg

So my chord does exist even it's not the most common one (and many others I guess), food for thought Moddart in case you'd like to improve your chord engine

I think that's the point of showing chords within PTQ, to identify the more complex ones easily and not for every single chord of a typical progression IMHO

Last edited by davidizquierdo82 (23-12-2022 10:30)
P85>Kawai CA97>Numa XGT>FP90X>LX706
Pianoteq 8 Pro (all instruments) + Organteq 2
i7 4790K W11 64bits + UMC1820 + MTM + DT770 pro X
http://youtube.com/DavidIzquierdoAzzouz

davidizquierdo82's Website

Reply by Fannon

  • Avatar
  • Fannon
  • Member
  • Offline
  • Registered: 11-09-2022
  • Posts: 14

Re: Chord Display Accuracy - Request

Looks like version 8.0.4 includes some improvements to chord detection? Very nice that you improve on such details!

https://www.modartt.com/pianoteq#changes

Just another minor suggestion: Is it possible to display the chord quicker? I'm not sure if this is caused by the animation or by the detection itself,. But other chord detectors display the played chord practically instantly.

Last edited by Fannon (29-12-2022 11:50)

Reply by davidizquierdo82

  • Avatar
  • davidizquierdo82
  • Member
  • Offline
  • Registered: 11-02-2015
  • Posts: 310

Re: Chord Display Accuracy - Request

Fannon wrote:

Is it possible to display the chord quicker? I'm not sure if this is caused by the animation or by the detection itself,. But other chord detectors display the played chord practically instantly.

That's another minor issue I'm also dealing with, I've since ever wondered why is it so sluggish at showing chords.. hopefully we can see some more improvements in this field

Also, version 8.0.4 is still not showing my chords (and I guess many other chords alike)

P85>Kawai CA97>Numa XGT>FP90X>LX706
Pianoteq 8 Pro (all instruments) + Organteq 2
i7 4790K W11 64bits + UMC1820 + MTM + DT770 pro X
http://youtube.com/DavidIzquierdoAzzouz

davidizquierdo82's Website

Reply by miiindbullets

  • Avatar
  • miiindbullets
  • Member
  • Offline
  • From: Idaho, USA
  • Registered: 31-03-2022
  • Posts: 251

Re: Chord Display Accuracy - Request

groovy wrote:
miiindbullets wrote:

I've noticed it won't register any extensions beyond the 9th. Here's a Cm11, for example, which registers in Chordie but not Pianoteq:

Have you seen, that your example is working now in Pianoteq 8.0.4?

https://i.postimg.cc/J085XypQ/Cm11-chord-detection-2.png

Awesome!

YouTube

miiindbullets's Website

Reply by budo

  • Avatar
  • budo
  • Member
  • Offline
  • Registered: 30-11-2016
  • Posts: 1,394

Re: Chord Display Accuracy - Request

davidizquierdo82 wrote:
Fannon wrote:

Is it possible to display the chord quicker? I'm not sure if this is caused by the animation or by the detection itself,. But other chord detectors display the played chord practically instantly.

That's another minor issue I'm also dealing with, I've since ever wondered why is it so sluggish at showing chords.. hopefully we can see some more improvements in this field

Also, version 8.0.4 is still not showing my chords (and I guess many other chords alike)

it's probably sluggish compared to other programs because it's spending a lot of CPU to produce sound and isn't budgeting as much for the detection.  that's fine with me.  it should be throwing as much CPU as possible at the model.  chord detection programs don't have to deal with this.

@youtube
@facebook

Reply by Fannon

  • Avatar
  • Fannon
  • Member
  • Offline
  • Registered: 11-09-2022
  • Posts: 14

Re: Chord Display Accuracy - Request

budo wrote:
davidizquierdo82 wrote:
Fannon wrote:

Is it possible to display the chord quicker? I'm not sure if this is caused by the animation or by the detection itself,. But other chord detectors display the played chord practically instantly.

That's another minor issue I'm also dealing with, I've since ever wondered why is it so sluggish at showing chords.. hopefully we can see some more improvements in this field

Also, version 8.0.4 is still not showing my chords (and I guess many other chords alike)

it's probably sluggish compared to other programs because it's spending a lot of CPU to produce sound and isn't budgeting as much for the detection.  that's fine with me.  it should be throwing as much CPU as possible at the model.  chord detection programs don't have to deal with this.

Actually, chord detection should not need much CPU at all and it's done already. After all, there's no sound processing involved, just taking MIDI data in and calculating the chord(s), maybe with a bit of time interpolation? The challenge is probably more with writing the code that does a good job at it. And maybe the chord detection is already instant and it's just a UI animation which makes it appear slow? In that case, it'd be an easy fix.