Absolutley agree on this. But it seems that also how you do the DC removal is important - I have a better version actually.. I think I need to upload these.
//
//
Then again, without dc removal it shifts the stage a lot to the front for B. I still think a DC removal is the right thing to do, though. But this shouldn't be audible if done right?! As you said, it's probably non trivial.
I'm surprised of this DC aspect too. I hadn't noticed it so clear before as with this particular recording. DC is interesting... like a singularity... 🙂 but I really agree - it should not be present.
//
//
Maybe one can only perceive a tone down to ~20 Hz, but variations in level on a longer timescale are still evaluated by the brain?
I know from doing precision EQing on PC (only a small band with minimal change <1dB) that it takes about a second for my brain to recognize the change. This is when I listen for where a sound comes from, not how it sounds. I also watched a video a while back, where different tonewoods for guitar were compared and they cut back and forth between two guitars within 2-3 seconds playing the same song. This is too short for me grasp the tone of each guitar. I hear the switch, but I need longer to fully capture the sound.
I know from doing precision EQing on PC (only a small band with minimal change <1dB) that it takes about a second for my brain to recognize the change. This is when I listen for where a sound comes from, not how it sounds. I also watched a video a while back, where different tonewoods for guitar were compared and they cut back and forth between two guitars within 2-3 seconds playing the same song. This is too short for me grasp the tone of each guitar. I hear the switch, but I need longer to fully capture the sound.
I wonder if it is really a notion of the DC itself but rather it has impact on the whole audible frequency range is some way... Maybe every "tone" is shifted in some way and some kind of distortion is created from this situation. When the DC is removed, everything is levelled and in "harmony".. ? I consulted my usually infinit knowledge resource but there was no answer... The ear-brain is fantastically adaptable and think the re-programming time is about the duration you mention.
//
//
I did just download the files in post #632, #636 & #638 and listened to them just via my computer and headphones, the difference between A and B are very clear, but the difference in bit resolution and sampling frequency isn't much discernible, if at all, on the other hand I suspected one of the ADC's recording may have one of the channels inverted and so I changed that in PulseEffects and voila... inverting one channel (didn't matter if L or R) on recording A and it sounded much more like recording B, ie. the spatiality disappeared, instead the vocalists voice was much more focused, and the sound overall "colder".
The same tonal difference could be observed when inverting one of the channels on recording B, the vocalists voice got more spatiality but less focused, as did the rest of the sound that got a bit warmer, ie. it did sound much more alike recording A.
I still find it a bit difficult to compare the recordings due to them being made at different part of the music performance.
However, my guess is that one of the ADC's (I am suspecting ADC A) may have one channel inverted.
The same tonal difference could be observed when inverting one of the channels on recording B, the vocalists voice got more spatiality but less focused, as did the rest of the sound that got a bit warmer, ie. it did sound much more alike recording A.
I still find it a bit difficult to compare the recordings due to them being made at different part of the music performance.
However, my guess is that one of the ADC's (I am suspecting ADC A) may have one channel inverted.
Last edited:
btw, one more thing I forgot to mention, I did also try summing together L & R, and recording A seem to dwindle away, kind of canceling out to some degree, more so than recording B which sound more like a typical mono playback, but I still think there's a little more going on here although inverting one channel seem to get recording A and B much closer to each other in resemblance.
OK so @Ultima Thule is correct about the phase reversal which is in the Left channel. I will upload a new A with the phase corrected and at the sae time add the the same DC blocking filter so also B will be updated.
Thanks again for finding this - it was really a puzzle 🙂
I'll be back but just now I need to complete the left channel of the Big Flower ;-D
//
Thanks again for finding this - it was really a puzzle 🙂
I'll be back but just now I need to complete the left channel of the Big Flower ;-D
//
That comment was regarding summing L & R without any phase inversion. Although I did a bit forth and back testing with both combinations, it's still some differences between A and B, but in particularly with case A in mono seemed to cause the vocalists voice to shrink more and becoming located farther away than B.
Maybe using the word "dwindle" is not the correct description as this could imply that the change is kind of decaying in one way or other slowly over time, which it is not, it's a constant difference.
I had some more listening, but as mentioned I think there's something more to it than just mono vs stereo and/or inverting phase on one channel or not, also using recordings of different part of the performance complicates the evaluation even further, I have a bit difficult to put a finger on what exactly it is, but inverting one channel on A made it to sound more like B, that's clear to me.
ps. nonetheless it's a beautiful recording, sounds fantastic with headphones on.
Maybe using the word "dwindle" is not the correct description as this could imply that the change is kind of decaying in one way or other slowly over time, which it is not, it's a constant difference.
I had some more listening, but as mentioned I think there's something more to it than just mono vs stereo and/or inverting phase on one channel or not, also using recordings of different part of the performance complicates the evaluation even further, I have a bit difficult to put a finger on what exactly it is, but inverting one channel on A made it to sound more like B, that's clear to me.
ps. nonetheless it's a beautiful recording, sounds fantastic with headphones on.
Thanks! Yes, an Y-cable and parallel recording would have been better. I have yet to hear a fixed version myself. The strange thing is that A had the impression of having a lot more LF - this seem contradictory when it played out of phase.
//
//
Okey - so now I have two new version of the excerpts that should be correct. I tried to level align them this time. Both have the same DC removal process.
Happy listening 🙂
//
Happy listening 🙂
//
Thank you for taking the time TNT.
I had a listen to the newest files. My impressions from initial listening:
I find A very forward. Everything is in front of me. The room, the singer, everything. The singer and piano are very loud, I do not hear a lot ambient from the building. The recording is warm and dense.
I find B very centered. Everything is around me. The room is like I would sit in a church. I would position the singer rather to my back than to my front. I hear a lot of ambient from the church. The recording is very open sounding (you can hear the air) and cold compared to A. I do not have the lack-of-detail issue which I had with the original recording B.
Recording B does fully externalize for me. I close my eyes and try to remember, imagine how it sounds when I really sit in a church, singer, piano 5-10 meters in front of me. Yes, B is precisely how this sounds. The singer & piano position corrects for me if I imagine actually sitting in this scene. The singer is small and distant and the church is large, airy, cold and echoy and all around me.
A does not do this. A is more as if you positioned the microphone right at the singer and piano, not at a distance in the room. Because of the forwardness, I find B easier to listen to than A. The density (?) of A is still very nice to listen to. It's very rich.
Now that we established how A and B sounds, would you like to share which recording device was A and B? I have an U87 clone in the works which needs a proper preamp with ADC 🙂
I also had a short listen inverting the left channel of the original recording A. Funny how this changes the perception of where things are. Inverting the right instead of left channel didn't have the same effect for me.
I totally agree. Any chance to get a longer recording once you resolved on the right post processing?
I had a listen to the newest files. My impressions from initial listening:
I find A very forward. Everything is in front of me. The room, the singer, everything. The singer and piano are very loud, I do not hear a lot ambient from the building. The recording is warm and dense.
I find B very centered. Everything is around me. The room is like I would sit in a church. I would position the singer rather to my back than to my front. I hear a lot of ambient from the church. The recording is very open sounding (you can hear the air) and cold compared to A. I do not have the lack-of-detail issue which I had with the original recording B.
Recording B does fully externalize for me. I close my eyes and try to remember, imagine how it sounds when I really sit in a church, singer, piano 5-10 meters in front of me. Yes, B is precisely how this sounds. The singer & piano position corrects for me if I imagine actually sitting in this scene. The singer is small and distant and the church is large, airy, cold and echoy and all around me.
A does not do this. A is more as if you positioned the microphone right at the singer and piano, not at a distance in the room. Because of the forwardness, I find B easier to listen to than A. The density (?) of A is still very nice to listen to. It's very rich.
Now that we established how A and B sounds, would you like to share which recording device was A and B? I have an U87 clone in the works which needs a proper preamp with ADC 🙂
I also had a short listen inverting the left channel of the original recording A. Funny how this changes the perception of where things are. Inverting the right instead of left channel didn't have the same effect for me.
I totally agree. Any chance to get a longer recording once you resolved on the right post processing?
NYX, thanks for the feedback. A was the one you preferred last time with the faulty setting of, actually A. With this corrected, you have switch your preference to B. A and B are the same ADCs as last time.
Sometimes I also prefer something weird... as if it special, but it doesn't last long usually.
A is the AKM based and B is the ESS based ADC. This was already implicitly revealed actually.
B is unchanged from before but for one aspect - an other DC removal process was used (same as for A now). This change gave a good increase in SQ as I hear it. Old B and this B is comparable wrt to DC removal process.
I have promised the artist to not to share the recording but it was me offering it rather than them requiring it. I can make inquiries about if I could share the whole concert (25 min...).
Thanks again for listening and giving feedback.
//
Sometimes I also prefer something weird... as if it special, but it doesn't last long usually.
A is the AKM based and B is the ESS based ADC. This was already implicitly revealed actually.
B is unchanged from before but for one aspect - an other DC removal process was used (same as for A now). This change gave a good increase in SQ as I hear it. Old B and this B is comparable wrt to DC removal process.
I have promised the artist to not to share the recording but it was me offering it rather than them requiring it. I can make inquiries about if I could share the whole concert (25 min...).
Thanks again for listening and giving feedback.
//
Listening with headphones my impressions are quite different. A has more ambient and soundstage is more as in the picture. But balance is a bit to the left and noise is higher. B is clean but very centered and singer appears to be slightly to the left of piano which does not match the picture.