A better version of The Blue Rain Coat

[tmtomh]

49 minutes ago, semente said:

Can you reduce the effect? It's too overpowering in my opinion.

 

It is a very interesting effect, and it appears to be the converse (if not exactly the opposite) of an ambience filter: it sounds like it adds out-of-phase information from each channel to the opposite channel, but to reduce crosstalk rather than to add reverb or the feeling of a larger hall/room.

 

It's an interesting effect, and because it doesn't seem to add reverb per se, it doesn't sound nearly as artificial or processed as conventional ambience processing.

 

That said, I do tend to agree with semente - the effect is rather extreme and a bit artificial-sounding, and I would think a subtler/more modest application of this DSP would be better for longer-term listening, to more different kinds of material. I have plenty of music in my collection - and this is regular rock, not chamber music, acoustic jazz, or whiz-bang-soundstage audiophile recordings - where the L-R soundstage already is quite wide. This effect would make such music sound bizarre and totally unpleasant. 

 

But it is an interesting effect that I believe can benefit some musical sources. I would like to learn more about it - is anything like this available as an off-the-shelf DSP plugin?

[tmtomh]

27 minutes ago, STC said:

 

It is not adding anything. And the original signal was unaltered in anyway except in this case it was upsampled for recording purpose. You are supposed to hear the sound as close to what you hear with headphones but a direct comparison cannot be made unless your headphones have your pinna correction filter. 

 

Thanks! I am confused, though: how can the DSP cancel/reduce L-R crosstalk without adding (or subtracting via the addition of out-of-phase info) something to the original signal?

[tmtomh]

54 minutes ago, STC said:

This is going to be hard to explain.   I will do my best....

 

Unplug your right speaker and play your favorite track at a very low volume in a quiet room. The level must be very low so that when you plug your right ear with your finger you should not be able to hear the left speaker‘s sound. You will notice change in timbre ( and also a shift in the position but that’s not relevant to the question you ask). 
 

This is the original of the left channel and received by your left ear. Now you put your headphones by muting the right channel and you are hearing the same as what your left ear heard with the loudspeaker setup with the right speaker muted. NOTE: There will be difference in the tonality anyway due to pinna function where you may find the sound of headphones a little less bright compared to the sound of the loudspeakers. The pinna modify the frequencies from 2 to 5KHz where else when the same sound played through the headphones it hits directly to the ear canal without pinna shaping effect. The final tonal balance can vary depending on the type of the headphone and how much bigger it is covering the ears. Earbuds completely eliminate any role of the pinna function so again all these will sound different from one another even if the loudspeaker and the headphones got identical frequency response measurements. 
 

The only way to find out is to take the measurement inside the ear canal just before the ear drum and apply the correction filter to the headphones. But once you do that you will sort of externalize the headphone sound and you can feel the sound to be coming from the front. ( This is what Ambiophonics about but then it also involves moving the speakers very close AND it is not related to your question but it is the basis for the DSP development. This is just the gist to explain why there is no DSP involved in the original sound so do not read too much into it because this itself is very complicated subject. 
 

Now going back to your question, my DSP ( actually there is no program yet and this is done by using DAW  with just delays and phase), presents the original signal untouched from the speaker to the ear. The ‘DSP’ is only concerned with the unwanted sound reaching the opposite ear which is what happens when you put your headphones as the opposite ear cannot hear the sound meant for the respective ear/channel.  
 

Can you follow this far?   

 

 

Thank you - very much appreciate the explanation and information.

 

I'm not sure I understand 100%, so forgive me if I am misunderstanding. But when you write "this is done by using DAW with just delays and phase," that is the alteration of the original signal, yes? And it includes altering the phase and timing of some of the signal, yes?

[tmtomh]

4 minutes ago, STC said:

The original signal is not touched. All the cancellation is done by the copy of the the original signal. if you take the measurement of the first sound emerges each speakers of the processed and unprocessed file, they are identical. What is altered is the copy of the original signal delayed and inverted and send to the other channel. 
 

A better way to understand and to do is, Have one pair of speaker that is producing the original recording and another pair for the processed file. It is important that the original signal is produced untouched to avoid any coloration otherwise this method is no different from any other DSP attempt. 

 

Again, many thanks for your patience and detail.

 

I do not mean to sound disagreeable or cranky, but "the copy of the original signal delayed and inverted and sent to the other channel" is exactly what I hypothesized several comments ago when I said it sounds like a phase-inverted portion of each channel is added to the other channel. You said that was not the case - but it turns out it is indeed the case.