I'm in a serious pickle, please help . . .

[chessrock]

I just got done tracking 12 songs this past weekend.

As I sit down to have a listen, I notice something peculiar.

All of the tracks . . . and I'm talking every one of them, are slow. To be precise, they're all about 1.67 steps too slow. What the f&*k? ? ? Anyway, no time to worry about what went wrong. I just need to fix the tracks and worry about what went wrong later.

Anyway, I tried the Sound Forge time-stretch / pitch-shifter, and it seems to do the trick just fine.

. . . Except that I'm still hearing some artifacts. It's not a totally smooth timestretch, and I've noticed this with Sound Forge in the past.

I remember once using a dedicated program that did nothing but time-stretching and as I recall it did a VERY good job, with almost no artifacts that I was aware of.

Only problem is I can't remember the name of the damn program ! Ughhh. It was made by a relatively well-known company. Anyone have any idea, or who can recommend another program that does close-to-flawless timestretching?

Thanks!

 

[drstawl]

The SF Time Stretch is virtually worthless because of those artifacts. You've got to get to the bottom of the sampling rate change, or redo the tracks.

Maybe another app does a better job, but that's an ugly thing to do to a file. Fix the problem at the source.

 

[Blue Bear Sound]

I just got done tracking 12 songs this past weekend.

As I sit down to have a listen, I notice something peculiar.

All of the tracks . . . and I'm talking every one of them, are slow. To be precise, they're all about 1.67 steps too slow. What the f&*k? ? ?

Sample rate mismatch between what was recorded and playback? ie - if you record at 48Khz and playback at 44.1, you'd hear a slowdown, but not necessarily a huge pitch change...

 

[sweetnubs]

h o l a.....c..h..e....s...s..n..u...b...s... . . . . . t..h..i..n..g..s.....a..r..e.....s..u..r..e.......s..l..o..w.. a..r..o..u..n..d.......h...e.....r....e........t...o...d...a...y...


you have a sampling rate mismatch (assuming you are all digital) between your software/hardware devices on playback or you tracked at different sampling rate than what you are playing back. Always make detailed session notes and include bit depth and sampling rate. check all your "clocks", sequencer sampling rates and the "clock settings" on your software control for your hardware if you have such a suchity such. peace out, trouser trout.

 

[Skysaw]

Re: Re: I'm in a serious pickle, please help . . .

Sample rate mismatch between what was recorded and playback? ie - if you record at 48Khz and playback at 44.1, you'd hear a slowdown, but not necessarily a huge pitch change...

I've gone through that, and there is a big difference in pitch. Makes my vocals sound a bit like Herman Munster. Not kidding.

 

[chessrock]

Whheeeewww ! ! !

(sigh of relief)

So what you're saying, if I'm not mistaken, is that I probably had the damn thing set to 48 instead of 44.

Well, Goddamn.

How the bloody cluster f%^k did that go and happen on me?

Now how the heck do I go back and fix this crap? I'm assuming I just reset the sample rate to 48 . . . but without actually re-sampling it, am I correct?

Thanks again, guys. I'm an asshole.

 

[Skysaw]

You've got Sound Forge so, you're in luck! The sample-rate conversion works quite nicely for this, and there should be no artifacts. I de-Munsterized two of my own tracks this way, so I should know.

 

[chessrock]

I wouldn't think there would be (any artifacts) . . . since you're not even resampling, right? I mean, it was recorded at 48 to begin with, so basically I'm just correcting it. Don't even think it would have to perform any calculations. I'm not very good with this stuff, though, (obviously) so I could be wrong.

Anyway, thanks for the help, guys. It's all fine now. Problem is completely rectified, and I'm extremely happy.

 

[Skysaw]

I wouldn't think there would be (any artifacts) . . . since you're not even resampling, right? I mean, it was recorded at 48 to begin with, so basically I'm just correcting it. Don't even think it would have to perform any calculations. I'm not very good with this stuff, though, (obviously) so I could be wrong.

I would think just the opposite. I imagine there is lots of math involved, but I also imagine the algorithm has been painstakingly worked out for this type of transfer, since it comes up often. By contrast, a pitch-shifter/time-stretcher has to be ready to do anything.

Glad it worked, by the way. Losing stuff sucks.

 

[chessrock]

I still want to know the name of that pitch-shifter I used . . . for future reference.

Anyone know which one I'm thinking of? ? ?

It's a good one, and I believe it's made by the same guys who make Nuendo or Logic . . . one of those guys.

 

[TexRoadkill]

You shouldn't have to convert anything. Just double check all your settings on your soundcard control panel and audio software to make sure they are all set to the same sample rate.

Did you have any digital devices hooked up then or now through SPDIF or AES? If you set your recorder to sync of their clocks you may have inadvertantly changed the sample rate

 

[Kryogh]

Last year, I got a similar problem. I found the problem right after the recording process.

I'm using a 01V, and at this time my soundcard was a Echo Mona.

The project was recorded in 48k, via lightpipe (optical ADAT from the 01v to the Mona). The point is, I forgot to set the 01V in 48k.
So, when I upgrade my soundcard to a Nuendo Multiset, everything was slowed down, because the Multiset "speaks" with my 01V and, when the project is in 48k, the 01v automaticaly go to 48k.

So I was f**ked.

I tried many solutions, but changing sample rate, resampling the waves... all this made me sick cuz I wasent satisfied with the results. So I called the band and told em we will redo all the recording, no charge obviously, minus 10% of the final amount of the project.

First they were obviously pissed, but at the end, they were happy cuz.. they correct some errors and their performance were better the 2nd time.

Anyway....I still hope you wont have to do this :|

Good luck !

 

[chessrock]

You shouldn't have to convert anything. Just double check all your settings on your soundcard control panel and audio software to make sure they are all set to the same sample rate.

Did you have any digital devices hooked up then or now through SPDIF or AES? If you set your recorder to sync of their clocks you may have inadvertantly changed the sample rate

Yea, I was recording 12 people, and it was a zoo.

Someone knocked over my A/D converter, and although it wasn't damaged, apparently the sample rate switch was inadvertently knocked to the 48 position in the fall. I checked it and sure enough . . .

Anyway, in Soundforge, it gives you the option of either resampling or just "switching the sample rate," which I'm assuming has the effect of telling my sound card what rate to play it back at . . . so switching it to 48 isn't actually changing anything in the audio, per se, and involves no resampling or any other calculations.