Organ sound correction, request for advice

[ivvos]

Hi all sound experts,
I'd appreciate your advice in handling this wav file located at: https://cloud.mail.ru/public/oUdg/RbAzbpnaV. It's an organ sound that obviously changes in the way it shouldn't from about 6.5 sec. The spectral pan and spectral phase pictures indicate an abrupt blue line displacement (please see the screenshots), and that seems to be the reason for the sound change. Is there any way to fix that issue (by repanning or somehow)? Thanks in advance.

[Ah_Dziz]

did you test it in mono? What does the blue line represent?

[ivvos]

Hi Ah_Dziz,
the issue is in the right channel, so I guess it should persist when it's converted into mono. Anyway, I took the sample of a higher note, pitched it down and substituted it for the original sample. That's a quick fix I'm happy with so far, unless a better solution appears.

[CHOOS]

What about simply adjusting the phase of the right channel at that point.

Cut it into the the first part and where it goes wrong.

Get a phase adjustment plugging and resample and join together

[ivvos]

A great idea, CHOOS, but I'm challenged when it comes to phase editing. Could you please advise a decent plugin that does the job? Should I capture a phase snapshot on the good part and then apply it to the other part, or is there another rule-of-thumb to follow?

[CHOOS]

I know phasebug works. It's a tiny simple plugin
There a stereo and mono version so if you leave it stereo you could edit just one side within it.

But there's bound to be way more

[ivvos]

Thanks CHOOS, I gave it a try and the Phasebug adjusted the phase of the 'blue' frequency. Unfortunately, the phases of all the other frequencies changed, too, so the complete picture looked quite distorted. Looks like I need a phase changer for just one frequency. Thanks for help anyway

[CHOOS]

OK I just had a look.
The left channel that doesn't change the phase is actually somewhat off but not terribly.
The right channel where it changes half way through the phase is actually fine.
What happens is that at the change it's like the fundamental is boosted either by an added fundamental sine wave or simply boosted in a very precise way.
So you could either EQ it out as soon as it starts with automation.
Or use a very fine Q on a dynamic EQ like TDR Nova (FREE) and compress exactly that frequency (like a multiband)
or simply chop it up and replace the "broken" part

[CHOOS]

Video shows one (attempted 2) way to fix it
https://rumble.com/vti38o-a-fix.html

[ivvos]

Thank you so much CHOOS, I appreciate your smart solution and will try to reproduce it. Meanwhile, I just use another sample of the same pitch, which is not broken.