This is a drinking song performed a British baritone Leo Stormont in 1903. The original wax recording has been restored by Historyradio.org, and is published...
This song is actually quite catchy. And it would have been quite popular even today with this baritone.
That you don’t like the song, I cannot comment on. That is an individual taste. But the original recording was a wax recording, remember that. To get from a wax recording to this is not bad. But I do not know who you are, and you may have much superior skills and resources. I just wanted to mention this wax recording point.
That you don’t like the song, I cannot comment on.
TBH, I could barely even hear it. I mean, I get that going from wax to digital is a big leap in itself, just that as I recall it, some similar AI-restoration projects have produced much richer end results, with more clean highs & lows, and less ambient noise. But yeah, that’s just me. Possibly I’m remembering poorly.
In any case, I enjoy history and learning about history, so I do appreciate your taking the trouble to share this. Thank you.
Well, the problem is that on these old cylinders there is quite a lot of static noise, and then in addition sometimes a sort or repeating thump every time the cylinder rotates. When you then remove the static and thump, you take some of the fullness of the original voice. I do not doubt that those with better equipment than myself would get a better outcome. But it is not easy. They have a recording of both Queen Victoria and Lord Tennyson. And it is very hard to extract something you can be sure is genuine. I hope these Ai services would set up some template for this. It should actually be quite easy, as the noise profiles on all of these are almost identical, and that subtraction process would be much better. It could be very interesting to hear the oldest voices on wax cylinder. I did this because enjoyed the song, probably because I also used to enjoy drinking :) Some people like other things, that’s ok…:)
Just for fun, I tried a test in one such service below. I chose the ‘keep vocals and music, stripping out noise/etc’ option. It seems to have greatly improved the vocals, while sadly losing the music. See what you think:
NOTE: that’s without fiddling with any settings, which likely would have produced a better result.
So, far as I can tell, most online services of this type seem to be paid, if that works for you. And yeah-- in previous examples of excellent preservation results, I did get the sense that most were ‘DIY,’ with all the custom hardware & software that entailed.
Or at least, they hired human restorers or mixing engineers to do the work. So yeah… it seems that in order to get a quality result, at least some money needs to change hands.
You can use a couple other logged-in services in place of an acct, like Google. Probably worth signing up either way so that you can test the service out for your own needs.
I’m just sad that it lost the music. Maybe another pass could be made, keeping the music and stripping the vocals, then you could use Audacity (free) to combine the two tracks. Several possibilities there…
That you don’t like the song, I cannot comment on. That is an individual taste. But the original recording was a wax recording, remember that. To get from a wax recording to this is not bad. But I do not know who you are, and you may have much superior skills and resources. I just wanted to mention this wax recording point.
TBH, I could barely even hear it. I mean, I get that going from wax to digital is a big leap in itself, just that as I recall it, some similar AI-restoration projects have produced much richer end results, with more clean highs & lows, and less ambient noise. But yeah, that’s just me. Possibly I’m remembering poorly.
In any case, I enjoy history and learning about history, so I do appreciate your taking the trouble to share this. Thank you.
Source: just an amateur idiot. :-)
Well, the problem is that on these old cylinders there is quite a lot of static noise, and then in addition sometimes a sort or repeating thump every time the cylinder rotates. When you then remove the static and thump, you take some of the fullness of the original voice. I do not doubt that those with better equipment than myself would get a better outcome. But it is not easy. They have a recording of both Queen Victoria and Lord Tennyson. And it is very hard to extract something you can be sure is genuine. I hope these Ai services would set up some template for this. It should actually be quite easy, as the noise profiles on all of these are almost identical, and that subtraction process would be much better. It could be very interesting to hear the oldest voices on wax cylinder. I did this because enjoyed the song, probably because I also used to enjoy drinking :) Some people like other things, that’s ok…:)
Just for fun, I tried a test in one such service below. I chose the ‘keep vocals and music, stripping out noise/etc’ option. It seems to have greatly improved the vocals, while sadly losing the music. See what you think:
https://auphonic.com/engine/status/RngqonNazCTnz4ZwdU5aGj
EDIT: (see alt DL link in comment below)
NOTE: that’s without fiddling with any settings, which likely would have produced a better result.
So, far as I can tell, most online services of this type seem to be paid, if that works for you. And yeah-- in previous examples of excellent preservation results, I did get the sense that most were ‘DIY,’ with all the custom hardware & software that entailed.
Or at least, they hired human restorers or mixing engineers to do the work. So yeah… it seems that in order to get a quality result, at least some money needs to change hands.
It appears we can’t access that without an account
You can use a couple other logged-in services in place of an acct, like Google. Probably worth signing up either way so that you can test the service out for your own needs.
Anyway, I’ve uploaded it below. See if that works for you:
https://drive.google.com/file/d/172XAHq6cqgcb7b5CekBidLi7XvO4XPJm/view?usp=sharing
Ohh, thank you. It sounds a lot clearer than the original!
I’m just sad that it lost the music. Maybe another pass could be made, keeping the music and stripping the vocals, then you could use Audacity (free) to combine the two tracks. Several possibilities there…