TIP: Cool half-time effect

[Machinate]

The following is an awesome thing I discovered while doing a superslow sample for a videogame project for a friend. I don't know if this can be done in any other way, so bear with me.

The effect is made by using a very high sample rate, so that even when slowed down, the material will still retain a lot of high-frequencies, as per Nyquist blahblah, etc.
Render the bit you want to slow down, and do it in the highest possible setting (I found that 192k worked well,hehe)
Open it up in your wave editor. I use soundforge, and it has a function called resample. Set the sample rate to the original sample rate of the track, in my case 44.1. and then check the "set sample rate only. Do not resample"-box. This way the file will be interpreted as 44.1, and thus slowed way down.

BUT since it's been recorded at 192k, it still has a huge amount of high-end sound in there from my synths and digital drum machines and whatnot.

This way your "drop everything an octave and play it at half speed" breaks won't sound like that's what you're doing

[noisetonepause]

But then I won't sound exactly like Massive Attack either... so what's the point?

-Paws

[Diskclaimer]

[john gordon]

But then I won't sound exactly like Massive Attack either... so what's the point?

-Paws

speaking of massive attack i heard that they may have something new coming very soon.

[montrealbreaks]

Render the bit you want to slow down, and do it in the highest possible setting (I found that 192k worked well,hehe)
Open it up in your wave editor. I use soundforge, and it has a function called resample. Set the sample rate to the original sample rate of the track, in my case 44.1. and then check the "set sample rate only. Do not resample"-box. This way the file will be interpreted as 44.1, and thus slowed way down.

BUT since it's been recorded at 192k, it still has a huge amount of high-end sound in there from my synths and digital drum machines and whatnot.

I tried this just now. Great trick! If you use 44.1 normally, try 88.2 or 176.4. This way, everything stays in tune in the same scale.

I used it on my voice, saying "This is the Bassline", then dropping it from 176.4 to 44.1. It got really nice and thick on the bottom, but the syllibance caused problems, cause now those "S"es are sitting around 3000 hertz and VERY noisy.

I think I will run the original audio file through De-esser to make it a little cleaner.

[Machinate]

Cool! Good idea about the de-esser as well.

[Pitch Black]

i think the BEST half-time effect is Janet Jackson's areola.

[John Daminato]

i think the BEST half-time effect is Janet Jackson's areola.

LOL

[Meef Chaloin]

[mooncaine]

i think the BEST half-time effect is Janet Jackson's areola.

It was the best boob her money could buy.

[Stace]

Great tip, works a treat! Thanks very much guys.

[darthfinger]

Is there a way to do this with Peak 5.2?

There is a "convert sample rate" option, but it actually does re-sample. So the effect is lost.

[Khazul]

BUT since it's been recorded at 192k, it still has a huge amount of high-end sound in there from my synths and digital drum machines and whatnot.

If the source was a digital drum machine or synth, than lets be generous and say its running its audio out at 44.1Khz, which by the time it hits your pristine 192Khz sampler has gone through oversampling, a phase linear filter, some downsampling with noise shaping and other general frigging to ensure basically the end equiv of a very steep filters with a cutoff starting at around 20Khz and getting close to complete muting by 22Khz, then I guess finally a DAC, and analog audio...

Sample that at 192K, and perform the trick you just mentioned and erm... well - that filtering is now being applied at around 4.5Khz...

Anything above that from a 44.1 digital sound source is going to just show you how crap its DAC processing is... (ie how much of the above theoretical oversampling and filtering they didnt do, and/or got wrong..)

You will probaly get better result from analog gear, except that probaly had a left over non-phase linear LP filter in the path so is equally unlikely to make much about 20K in it...