SuperAnalog808, weird preview problem

I just bought the gold baby superanalog808 kit, and i converted all the samples to 16 bit with audacity (ive done this before with other kits, no problem), and then loaded them into my 2500. When i preview the sounds (before i actually load them into memory) they sound like they've been time stretched to be rrreeeeaaaalllyyy slow, in a destructive way, and sound terrible. But when i actually load them into memory and assign to a pad and play it, they sound how they should be sounding. Its only with this kit.

I might try and re convert all the samples and delete the ones now from the MPC and put the reconverted ones on later, but for now, any help

just guessing - but could they be stereo files that are encoded as interlaced?

if the preview function is somehow not cool with this format - it might sound the file is half speed.

tapedeck wrote:just guessing - but could they be stereo files that are encoded as interlaced?

if the preview function is somehow not cool with this format - it might sound the file is half speed.

if any of them are as you said then i dont think it matters because the bass drums are mono and it happens with them too

is it possible to encode mono files like you said for stereo and have these problems too? if yes how do i check if they are encoded as interlaced (not sure what it is)?

ThatsMisterMooreToYou wrote:is it possible to encode mono files like you said for stereo and have these problems too? if yes how do i check if they are encoded as interlaced (not sure what it is)?

they could be called 'mono' but really be stereo tracks with the same info on both channels - i dont know, was just a blind guess.

you could probably get some info by opening the file in audacity, using a program called 'sox', or you could post up a file somewhere for someone else to check.

I use Extreme Sample Convertor myself. I've had similar conversion errors with GoldBaby sounds too.

Judging from some of the bit rates listed on the original files I think tapedeck is onto something here.

44100 samples x 1 channel x 24 bits = 1058400 bits per second listed as 1058.

A lot of the 1058 ones are the ones that don't convert right.

Well looks like some of the bass drums are actually stereo, but the problem is the same.
So im not sure what can i do then? Im not really familiar with bits and things like that all i know is for the MPC i always can convert to 16 bit, 44.1 or whatever and im good to go, and now i cant (with this goldbaby pack)

I used pro tools to convert all them and the sounds dont sound timestrteched or slowed down in preview anymore but there is a chunk of the mid-high frequences sound like they're missing in preview, but once its loaded to a pad i can hear the full sound.

Def better than before but still wondering if theres more i can do?

From the site: "This sample pack has been recorded at a higher sampler rate"
But that all gets taken care of when i convert down to 16 bit and 44.1 right?

ThatsMisterMooreToYou wrote:But that all gets taken care of when i convert down to 16 bit and 44.1 right?

until you post some files for someone to look at, i have no idea

a before and after file (one straight from goldbaby, one converted) would help.
also - are other sounds you bounce from protools ok in your mpc?
also - you started out saying you converted with audacity, now you are converting with protools...so not sure what changed since the switch.
anyway lots of confusion on my end that could be addressed with some demo files.

audacity should be able to handle this no problem - however i know it is a pain to convert so many files one by one. when i converted a goldbaby pack, i used foobar2000 and had no problems...that's on windows - not sure what os you are on.
if you are on mac i could maybe help you with a sox script - it gets nerdy but its awesome.

i have been sampling all my super808 goldbaby stuff into the analog ins on my mpc and importing them that way, that works for me everytime, but it takes forever to get all these sounds in there.

when i started recording stuff in, i have found that you only need 1-3 samples in each one though because the round robin feature doesnt work in the mpc (unless your a guru and follow that guide, but thats a whole different topic) so i have been sampling bass 1, bass 2, bass 3 and labeling them K_808 k1 a1, k_808 k1 a2, k_808 k1 a3

broken down like this becuase K is for kick,
808 is what it is,
k1 stands for kit 1 and
A1 stands for the first set of kicks in that kit,

the second set would be labeled as this
K_808 k1 b1, k_808 k1 b2, k_808 k1 b3.

then the next kit is labeled as k2, and then i do the whole process over again.

i know you didnt ask for it, but i hope this helps you to organize your kits when you finally get them in there, i know i would be lost with all these duplicate samples if i didnt sit there and go through trial and error with my sample names.

i find that i only need 3 sounds (that are in the same set) because i can set one for a low velocity, one for a mid and one for a high, and it will cycle through them like that, instad of round robin.
with the mpc, 9 of the same sound (or very similar) is overkill. so this cuts down on the time it takes to manually record all this stuff as well.
i would just do 1 sample per round robin, but i am kinda ocd and figure i should get at least the few i can use with the mpc if anyone will really notice when im making a track or not.

damien907 wrote:with the mpc, 9 of the same sound (or very similar) is overkill

just to point out for the op, that is defeating the purpose of round robin.
i think only jj can do this anyway in an mpc, but the point is, if you want to benefit from the 'super analog' bit made possible by the round robin, you need to have many different similar samples to cycle through. this is not the same as using different velocities, which imo is a waste of memory to use different samples for.

i have batch converted entire goldbaby directories to work successfully in my 2kxl using foobar for the conversion, and 1-4a rename to convert all the filenames to 8 characters (that was a challenge).

so it is possible, and relatively easy to do, something in the op workflow is messing this up.