|
Korg Forums A forum for Korg product users and musicians around the world. Moderated Independently. Owned by Irish Acts Recording Studio & hosted by KORG USA
|
View previous topic :: View next topic |
Author |
Message |
Liviou2004 Platinum Member
Joined: 20 Feb 2017 Posts: 1150 Location: France
|
Posted: Wed Apr 29, 2020 12:50 pm Post subject: COMBI : non-wanted phase shifting in Piano sound |
|
|
Hello,
I've noticed a little issue in Combi mode.
Starting from an Init combi, Kronos set all tracks on Int A000 Berlin Grand.
As I'm currently studying the Fx routing possibilities, I need the same sound on two tracks playing at the same time. So, I set tracks 1 and 2 on the same 01G midi channel.
Theorically, we should hear the same sound. But it's not the case. The Piano sound is a bit altered with a kind of phase shifting, as if there were a phase conflict between the two tracks.
(I hope I'm clear enough).
However, I haven't noticed the same issue with an electric piano sound ! It seems this issue appears with a piano sound only (to be confirmed).
Could you do the same test and tell me if you get the same result ?
Thanks |
|
Back to top |
|
|
enigmahack Approved Merchant
Joined: 04 Aug 2009 Posts: 727 Location: Moncton, NB, CANADA (Eh?)
|
Posted: Wed Apr 29, 2020 2:15 pm Post subject: |
|
|
This frequently happens with complex waveforms. Technically you can hear it with things like identical saw waveforms, but the effect is far less subtle.
Some sounds like EP's might have a simpler waveform or may contain harmonic content that doesn't phase so easily.
I've definitely encountered what you're talking about when having the exact same sound on multiple timbres. It also happens when I have MIDI in/out loopback enabled.
If you get an EP waveform in the HD-1 engine and enable dual oscillators, you might see this effect.
You can experiment with a variety of waveforms and you'll notice some phase much more than others. This is expected behavior. _________________ Korg Kronos 88 2, Korg Kronos 73, Kurzweil K2600S
Sound developer, custom sound designer and trainer/Kronos support - www.audora.ca for details! |
|
Back to top |
|
|
GregC Platinum Member
Joined: 15 May 2002 Posts: 9451 Location: Discovery Bay (San Francisco Bay Area)
|
Posted: Wed Apr 29, 2020 3:32 pm Post subject: |
|
|
This might be relevant
Many AP's are from the SGX-2 sound engine. Some [few ?]. are HD-1
EP has its own sound engine
I wouldn't expect perfect or compare performance from instruments when they are derived from different sound engines. _________________ Kronos 88. MODX8
Achieve your musical dreams
https://soundcloud.com/user-898236994 |
|
Back to top |
|
|
kronoSphere Platinum Member
Joined: 04 Jan 2012 Posts: 698
|
Posted: Wed Apr 29, 2020 4:41 pm Post subject: |
|
|
Sometmes in a mixdown some instruments can alter or even eat the sustain or theresonnance of others instruments. In some case one must "artificially" add some sustain, add some reverb to re-equilibrate the tracks involved.
Concerning the phase I totally agree with the previous post. May I suggest to pan indivually the 2 piano tracks, one a little lefter and the other a litlle righte. _________________ trees are going fast.
https://www.lairdeparis.fr
Current Gear : Kronos 88 / Seaboard Rise / Triton Extreme / Sequoia / Motif Rack XS / TC Helicon voicelive rack /Awave 11 / Audio & VSTi plug-ins connected /wide touchscreen / iPad Pro 512. |
|
Back to top |
|
|
voip Platinum Member
Joined: 27 Nov 2014 Posts: 3777
|
Posted: Wed Apr 29, 2020 6:51 pm Post subject: |
|
|
Something along these lines was discussed some time ago. I believe it may have been Dan Phillips that provided a solution, but haven't been able to find the relevant thread.
It's all to do with timing. The note and velocity will be the same for both timbres, resulting in, I think, exactly the same sample, or sample mix, being triggered for Timbres 1 and 2, but the note for timbre 2 gets allocated another slot in the polyphonic SGX-2 engine. This process takes a few processor cycles, so the sound for Timbre 2 is launched a fraction of a millisecond or so later. Since the audio produced by both Timbres mixes to the audio output without any further delay, various notes and overtones/harmonics mix constructively and destructively leading to the phasing effect. The lowest notes are the least affected, and simply sound louder with both Timbres in the mix, since the phase angle is shifted less in the lower registers. It is possible to verify this by changing the Delay time for one of the Timbres in the Timbre Parameter / Delay tab, and observing how the notes most affected by the phasing effect move up and down the keyboard as the delay is altered.
One solution might be to combine two pianos that rely on different samples. Another might be to force the SGX-2 engine to use a different sample for each Timbre, by transposing one Timbre and then using the Pitch adjustment to bring it back into tune. The Transposed and re-pitched Timbre sounds quite different from the original piano sound, however, but the phasing effect is more or less gone.
If the two pianos are routed to different effects, then the end result might not sound quite as phasey, rendering the above tricks redundant.
. |
|
Back to top |
|
|
KK Platinum Member
Joined: 13 Oct 2016 Posts: 1422
|
Posted: Wed Apr 29, 2020 8:21 pm Post subject: |
|
|
On a similar subject, you probably all heard at least once a song in which there would be a short characteristic stereo phasing - where the same signal for the song would be duplicated at almost exactly at the same time - that was particularly popular in the 80s disco era.
The OOPS (out of phase stereo) is another technique that could sometimes help decipher what is going on songs which are particularly busy. It is a common feature in software now, but in my time with a standard amplifier, this could be also obtained by connecting both speakers' ground terminals in series with a low resistor to the amp ground (to avoid causing damage to the amplifier). This way, anything that is happening at the same time on both channels becomes silent or very low in amplitude, while you can hear other instruments, etc.
In the Kronos, a similar simple experiment can be done very quickly :
1) Select any EXi program, then initialize it (don't save it later, so nothing to worry about). By default, you should have an AL-1 program with two saw/pulse oscillators running at identical amplitude levels.
2) Go to EXi 1 / Mixer tab.
3) Check the Phase Inv box only for OSC2.
4) Play a key. Oh no ! No sound anymore. What the *** happened ?
It's exactly what would happen with an OOPS example with both signals identical. Except that in the Kronos, the signals perfectly canceling each other out with silence are two oscillators mixed in the AL-1 as if they were stereo signals in a standard amplifier.
If you want to experiment more for fun, with both OSCs still canceling each other with silence, now increase only OSC1 initial phase to 1 and then gradually up to max and listen to what happens when you hit keys.
This is just one tiny fun thing you can do with the Kronos. On top of being the greatest synth, you can use it as the most incredible synth lab machine to keep you busy for the next 20 years. |
|
Back to top |
|
|
Liviou2004 Platinum Member
Joined: 20 Feb 2017 Posts: 1150 Location: France
|
Posted: Thu Apr 30, 2020 6:39 am Post subject: |
|
|
voip wrote: | Something along these lines was discussed some time ago. I believe it may have been Dan Phillips that provided a solution, but haven't been able to find the relevant thread.
It's all to do with timing. The note and velocity will be the same for both timbres, resulting in, I think, exactly the same sample, or sample mix, being triggered for Timbres 1 and 2, but the note for timbre 2 gets allocated another slot in the polyphonic SGX-2 engine. This process takes a few processor cycles, so the sound for Timbre 2 is launched a fraction of a millisecond or so later. Since the audio produced by both Timbres mixes to the audio output without any further delay, various notes and overtones/harmonics mix constructively and destructively leading to the phasing effect. The lowest notes are the least affected, and simply sound louder with both Timbres in the mix, since the phase angle is shifted less in the lower registers. It is possible to verify this by changing the Delay time for one of the Timbres in the Timbre Parameter / Delay tab, and observing how the notes most affected by the phasing effect move up and down the keyboard as the delay is altered.
One solution might be to combine two pianos that rely on different samples. Another might be to force the SGX-2 engine to use a different sample for each Timbre, by transposing one Timbre and then using the Pitch adjustment to bring it back into tune. The Transposed and re-pitched Timbre sounds quite different from the original piano sound, however, but the phasing effect is more or less gone.
If the two pianos are routed to different effects, then the end result might not sound quite as phasey, rendering the above tricks redundant.
. |
Thank you, voip, for having think about this question.
I will try to get the right Dan Philips post.
And I will test what you suggest.
EDIT : Well, I've tried to transpose and back-transpose in Combi, but same result. I've tried to copy one program twice and refer to in Combi, same result. I've tried in HD1 pianos too : same result.
As concern Dan Phillips, I didn't find his posts here. Do you know what pseudo he is using ? |
|
Back to top |
|
|
|
|
You cannot post new topics in this forum You cannot reply to topics in this forum You cannot edit your posts in this forum You cannot delete your posts in this forum You cannot vote in polls in this forum
|
Powered by phpBB © 2001, 2005 phpBB Group
|