Tempo/pitch stretching finished mix

[Keni]

Ok... This seems to be far more difficult than it used to be.

 

i have a file that was digitized from an old tape recording. I don't know who or how, but both the tempo and pitch is too fast/high.

 

i used to simply stretch the clip and both tempo/pitch adjusted relatively, but now only tempo changes.

i've been trying to do this now using melodyne, but as a very new melodyne user, the learning curve is daunting.

 

is there some simple method in CbB that allows the pitch to change with the stretch?

 

thanks...

 

[John Sandlin]

You're looking for élastique Pro time/pitch which is now included in Cakewalk by BandLab.

[Keni]

8 minutes ago, John Sandlin said:

You're looking for élastique Pro time/pitch which is now included in Cakewalk by BandLab.

Thanks John, but how do I make it shift audio as well as tempo? Currently, stretching with it slows the tempo of the track, but not alter the pitch.

[John Sandlin]

Well, I'm not having any luck on how to use élastique. I cheat and usually use Audacity for pitch work. Cakewalk by BandLab documentation needs to be updated, I think, to include How To Use for the new toys.  Melodyne should do it too - if you have a non-Trial/Demo version.

[Keni]

10 minutes ago, John Sandlin said:

Well, I'm not having any luck on how to use élastique. I cheat and usually use Audacity for pitch work. Cakewalk by BandLab documentation needs to be updated, I think, to include How To Use for the new toys.  Melodyne should do it too - if you have a non-Trial/Demo version.

Thanks John.

 

yeah, funny how things that used to be "normal" are now so hard to do.

i do have melodyne which i just upgraded to editor but the learning curve there is daunting. I've been on this for a few days now.

 

i can stretch the audio in CbB and then use transpose, but trying to get the values to coincide is currently beyond my math.

 

...and using this method I must do each song individually. There are two long sets. If only stretch would do both i could easily repeat the process n one song until the pitch matches the known pitch of the song, then apply that percentage to the entire 2 set recording...

 

How good is audacity results? I could d/l a copy and use that?

 

Edited December 29, 2018 by Keni

[John Sandlin]

Audacity is in the "OK" range for results. I don't think you have a 64 bit engine in it (I could be wrong). I usually really only use Audacity for special effects, and pitch shifting for that reason - making sounds "odd" on purpose. It does work, though.

 

[Keni]

49 minutes ago, John Sandlin said:

Audacity is in the "OK" range for results. I don't think you have a 64 bit engine in it (I could be wrong). I usually really only use Audacity for special effects, and pitch shifting for that reason - making sounds "odd" on purpose. It does work, though.

 

Thanks...

 

i guess i'll continue to work without audacity then...

 

it's supposed to be easy to do in melodyne's universal mode but so far i'm not having any luck figuring it out.

 

i've been using v-vocal all this time as melodyne essentials doesn't have the tools i needed and it eas very simple. I upgraded to assistant the other day and finally editor today but other than my normal pitch editing, the myriad of melodyne tools, options, and methodology i find it a bit difficult...

 

[noynekker]

Hey Keni . . . as Mr. Sandlin says above . . . inside CakeLab's latest "big" update . . . "élastique Pro time/pitch" should work to do what you're looking for.

I must admit I've had some troubles trying to get a larger complicated project to adjust tempo upwards / downwards using élastique, but I'm thinking to adjust one track as you're looking to do should work pretty easy.

When they released that update, there was a great thread by Craig Anderton that I found very useful over in the old forum regarding this:

http://forum.cakewalk.com/Changing-Tempo-of-Entire-Song-with-201809-m3787767.aspx

 . . . hopefully you'll find it useful for what you're trying to do ? . . . (oh well, just noticed you're there in post #3)

[John Sandlin]

3 minutes ago, noynekker said:

Hey Keni . . . as Mr. Sandlin says above . . . inside CakeLab's latest "big" update . . . "élastique Pro time/pitch" should work to do what you're looking for.

Sadly, elastique doesn't do pitch yet as far as I can tell.

 

[noynekker]

5 minutes ago, John Sandlin said:

Sadly, elastique doesn't do pitch yet as far as I can tell.

 

It's true, no pitch, but if Cakewalk can at least get the tempo correct, then Keni's version of Melodyne Editor should be able to use "select all" and adjust up (or down) one cent at a time to find the desired pitch.

[Keni]

13 minutes ago, noynekker said:

Hey Keni . . . as Mr. Sandlin says above . . . inside CakeLab's latest "big" update . . . "élastique Pro time/pitch" should work to do what you're looking for.

I must admit I've had some troubles trying to get a larger complicated project to adjust tempo upwards / downwards using élastique, but I'm thinking to adjust one track as you're looking to do should work pretty easy.

When they released that update, there was a great thread by Craig Anderton that I found very useful over in the old forum regarding this:

http://forum.cakewalk.com/Changing-Tempo-of-Entire-Song-with-201809-m3787767.aspx

 . . . hopefully you'll find it useful for what you're trying to do ? . . . (oh well, just noticed you're there in post #3)

Thanks noynekker...

 

you might notice my post in that thread immediately beliw Craig's first 3...

 

Sadly, trying to stretch the pitch separately from tempo has been less than optimum and this recording is two long sets of songs so the issues are difficult.

 

melodyne free standing does a decent job of pitch shift for it but time stretch not so good as it doesn't average tempo so there are many serious fluctuations that occur...

 

right now i'm missing the good old days when stretching affected both pitch/tempo in relative ratio...

[Keni]

4 minutes ago, noynekker said:

It's true, no pitch, but if Cakewalk can at least get the tempo correct, then Keni's version of Melodyne Editor should be able to use "select all" and adjust up (or down) one cent at a time to find the desired pitch.

Good idea, except i can't tell the correct time value while it sounds chipmunk.

 

i just tried a few more attempts and i'm burnt for the night. Too many problems with a 2 hour recording that has little clarity to begin with. Tomorrow i'll try again with fresh ears...

 

thanks for the help.

 

[marcL]

Hi Keni, I have exactly the same problem (some waves saved from a tape recorder). But in my case the tape recorder went to slow (sound to low). So I asked the same question in my own words some days ago. Here is the thread:

It is really a hard thing to accomplish this stuff in CbB and it had been so easy with my old tape recorder (Varispeed). My first attempt was using Audacity and it really worked. But on one hand I am not sure about the quality (even if I exported 32-bit float format) and on the other hand it was complicate to find out how much speed change is necessary. But because my tracks were recorded with a click and I also know their song keys, I could calculate the new file length and then use Audacity's "Change Speed ..." function.

[Keni]

16 minutes ago, marled said:

Hi Keni, I have exactly the same problem (some waves saved from a tape recorder). But in my case the tape recorder went to slow (sound to low). So I asked the same question in my own words some days ago. Here is the thread:

It is really a hard thing to accomplish this stuff in CbB and it had been so easy with my old tape recorder (Varispeed). My first attempt was using Audacity and it really worked. But on one hand I am not sure about the quality (even if I exported 32-bit float format) and on the other hand it was complicate to find out how much speed change is necessary. But because my tracks were recorded with a click and I also know their song keys, I could calculate the new file length and then use Audacity's "Change Speed ..." function.

Thanks Marled...

 

I'll give this a try today.

 

i wish i had the original tape as i still own a deck with varispeed, but i was given a file.

 

i previously did this for another cluent/friend as i had the tape and did the transfer. The issue there was simple as the original tape was 3.75ips but my deck only does 7.5/15ips. Varispeed wouldn't go that far so i digitized it at 7.5, stretched to 200%, then transposed the octave.

 

but in this case, the digitized file is by some unkown amount less than that. Maybe tape stretch or recoded on dying batteries or such. Using a piano i determind the pitch center to be close to Bb on the first tune and found it to sound more natural near f#/g. Guessing the feel/sound of the guitar lead be to believe g is probably right. But when i use melodyne it shows the pitch near d? Weird!

 

the recording is also extremely muddy. Recorded from a less than optimum location...

 

difficult...

 

thanks. I forget about the stretch plugin...

 

[Keni]

Odd...

 

i found stretch.ax and .hlp in my sonar platinum folder.

i could not get regsvr32 to register it.

 

i booted an old machine with vista32 and as far back as sonar 8.5

opening plugin manager shows stretch as active though it is not in any fx menus so i added to a menu and saved it, but it still doesn't show in the menu...

 

any idea what's going on?

 

thanks

[marcL]

6 hours ago, Keni said:

but in this case, the digitized file is by some unkown amount less than that. Maybe tape stretch or recoded on dying batteries or such. Using a piano i determind the pitch center to be close to Bb on the first tune and found it to sound more natural near f#/g. Guessing the feel/sound of the guitar lead be to believe g is probably right. But when i use melodyne it shows the pitch near d? Weird!

2 days ago I made an even more strange discovery with Melodyne! In my case the files come from an 8-track 1/2 inch recorder, so I have different tracks for the instruments. As I know the song keys I checked the bass track with Melodyne (IMO this is easiest to check pitches). When I compared the Melodyne analysis in the detection mode with the one of the edit mode I noticed about 48 cents difference. It seems that the detection mode shows about the original values played on the instruments (although the wave file is too low) and the edit mode shows the values caused by the slower playing! Hell, like a miracle! Thereafter I checked other recordings that should be correct (recorded via USB or VSTi) in Melodyne and noticed also there pitch differences (only small in this case)! IMHO this should not be the case, so I wrote an email to the Melodyne support and I am curious what will be there answer.

[Keni]

12 minutes ago, marled said:

2 days ago I made an even more strange discovery with Melodyne! In my case the files come from an 8-track 1/2 inch recorder, so I have different tracks for the instruments. As I know the song keys I checked the bass track with Melodyne (IMO this is easiest to check pitches). When I compared the Melodyne analysis in the detection mode with the one of the edit mode I noticed about 48 cents difference. It seems that the detection mode shows about the original values played on the instruments (although the wave file is too low) and the edit mode shows the values caused by the slower playing! Hell, like a miracle! Thereafter I checked other recordings that should be correct (recorded via USB or VSTi) in Melodyne and noticed also there pitch differences (only small in this case)! IMHO this should not be the case, so I wrote an email to the Melodyne support and I am curious what will be there answer.

That is strange. I'm guessing the different algos have disceet reasoning behind them. Melodyne seems to discreetly consider many aspects separately and differently. All documented of course.

 

for me most of the issues are of little importance as i really use pitch correction for small location edits. I don't auto correct anything. So my primary focus is on those tools, but it is still daunting. V-Vocal is far easier and requires almost zero learning curve.

 

[marcL]

Received a message from Celemony support (wow, very swift). It is all very simple:

In the detection mode Melodyne displays the pitch values per default not in relation to 440 Hz, but in relation to the detected tuning (e.g. 439.2 Hz)! Very odd! But when you change the tuning (in detection mode) to 440 Hz it matches the values of the edit mode.

But even if it is that simple, it may help us to find out how much our recordings are detuned, at least if the detuning is not more than 50 cents or in the other case it helps to find out the fine-tuning value. Well, that's another use case for Melodyne.

[Keni]

8 hours ago, marled said:

Received a message from Celemony support (wow, very swift). It is all very simple:

In the detection mode Melodyne displays the pitch values per default not in relation to 440 Hz, but in relation to the detected tuning (e.g. 439.2 Hz)! Very odd! But when you change the tuning (in detection mode) to 440 Hz it matches the values of the edit mode.

But even if it is that simple, it may help us to find out how much our recordings are detuned, at least if the detuning is not more than 50 cents or in the other case it helps to find out the fine-tuning value. Well, that's another use case for Melodyne.

Hmmm... How could it know the tuning standard of the source if the source is askew?

[marcL]

On 1/2/2019 at 2:24 AM, Keni said:

Hmmm... How could it know the tuning standard of the source if the source is askew? 

It analyzes all the note pitch values and uses an average tuning. That's all that is possible, I guess.

By the way I have found now a good solution for my pitch/stretching problems: I have also Samplitude and there is the menu item "Effects > Time / Pitch > Resampling /Time Stretching..." and if I use the algorithm "Resample", then it does exactly what I was searching for (maybe you too?). You can even choose if the tempo or the pitch is leading!

[Keni]

Thanks marled...

 

I managed to get Sonar X1 which I still have on an old 32bit os machine to work using Time Pitch Stretch.

 

The only hard part was the math to get the stretch equal to the pitch change...