Resampling integrity is faulted.

Any problems with your Ultra sampler should be reported here, hopefully someone might know what's wrong.

Moderators: ezman, stu, Ole

Resampling integrity is faulted.

Postby drayon » Wed Jun 02, 2004 9:38 am

I noticed when internally resampling a sound in the EMU the resulting sample does NOT sound like the original. Im using headphones btw.

This has become somewhat of an annoyance. I have been working on a bass sound in a preset, where i set the Peak/Shelf Morph filter up to taste. When i play different notes there is obviously tambre (tonal quality) shift. There is a section around F3-B3 that sounds similar and consequently the sound i desire. Im liking the sound of A3 so i decided to jump into the New Sample page, lock in Resample mode 16, 18 or 20bit, and switched Monitor ON - I executed the sampling process. Upon inspection of the new sample (triggering the same note) , i noticed it sounded nothing like the same tone of the original note from the Preset when monitoring.

Now, the idea of the 'Monitor On' function is so u can hear exactly the sound that is too be sampled. Unfortunately, this does not seem to be the case as my resulting sample sounds unlike the sound that im Monitoring. This really is buggin the hell outta me, id expect an internal resample to be an exact copy of what is played.

Others experienced this? Find a solution?? I tried various ideas like manipulating the Output Boost and the Output Headroom parameters in the Master section, tried adjusting the Volume pot, tried the various resampling bit rates. The resampled sounds are either overly clipped or lack the fullness in the bass. I even tried sampling from the headphone output (monitor off) back to the machine via the regular analog sampling method. No dice there either...never enuff bass.

Hope some ppl ere have some comments to make.

laiterz :sad:
drayon
 
Posts: 211
Joined: Thu Jul 24, 2003 3:46 am
Location: a void where there should be ecstasy

Postby sampleandhold » Wed Jun 02, 2004 8:26 pm

I have had this happen a few times as well. i tried sampling a cymbal from a cd once and it sounded ass. i don't know why it does this or how to fix it. it almost seems like the sampler finds errors in the recording and exposes them. also remember, you are using a filter, and filters do affect certian notes more then others. for example if you set the lpf at 440hz, you will get quiter notes the more you play above A4 or whatever the A it is. so that one note you play that you like the sound of has the filter effecting it in whatever way you set the filter up like. so then your idea of resampling should work to maintain that sound. but of course the timbre will change as you interpolate the sound on the sampler.

maybe what you could do is make sure you turn every auto system off on the sampler, like normalize, if you have them on. maybe the damn thing is compressing the sound a little or something and screwing it up. if that doesn't work, record you note out of your sampler into a cd format, or on your computer and send it back by resampling through the inputs. I know my breaks that i sample actualy tend to gain low end when i sample them this way. everything for me actualy seems way too bass heavy. a few things to try when sampling is play with the sample rate. if you are sampling a sine wave bass for example at say 100 hz, you should, technically get away with retaining the original sound but sample at a lower rat like 22.1 or whatever. perhaps that is something you could try is sampling at a lower rate, which will make the bass stand out more in theory since you are not going to capture the higher harmonics as well. or you could go the oppisite direction and sample at 48 and see what happens then.

that is about all i have for ideas.

it sounds to me that the sampler is trying to maintain something that you can't hear that it is picking up. you know the peak shelf filter is a really strong filter and when you resample it, it may freak the sampler out because the low end is spiking so much that it is pushing the low end down to prevent clipping or something. i posted a little thing about the low pass filters about two or three months ago that talk about this. I had a drum loop i was sweeping with an lpf and i was getting major distortion on just those parts i was sweeping thus causing me loudness issues. i later found out the lpf was spiking a really fast boost of low energy when it was filtering even with out q. so i wonder if that may be the problem...

maybe some one else has an idea,...
"{jU$t-n3Rv0U$-N-+h3-@Ll3y-W@y}"
sampleandhold
 
Posts: 581
Joined: Tue Apr 15, 2003 4:38 am

Re: Resampling integrity is faulted.

Postby illinformed » Thu Jun 03, 2004 8:57 am

drayon wrote:Now, the idea of the 'Monitor On' function is so u can hear exactly the sound that is too be sampled. Unfortunately, this does not seem to be the case as my resulting sample sounds unlike the sound that im Monitoring. This really is buggin the hell outta me, id expect an internal resample to be an exact copy of what is played.


When you say 'Monitor On' I can't work out if you've read this workaround Here. It basically tells you to set the Monitor field to off in the 'create sample' screen. When I set the monitor to off, I can still here what I'm playing. This backs up the idea that when monitor is set to on, there are 2 lots of signal present which cause phase canellation problems. It's a bit confusing because you can't really here 2 lots of signal when you are recording, only the destructive effects after. I'd love to see emu's detailed routing diagrams for resampling.
illinformed
 
Posts: 259
Joined: Thu Sep 11, 2003 2:12 pm
Location: London

Postby elemental » Thu Jun 03, 2004 4:16 pm

yeah it does sound like phasing problems. this would cause filtering-like effects.

try it with monitor off.
elemental
 
Posts: 32
Joined: Wed Feb 04, 2004 6:02 pm
Location: London

Postby drayon » Thu Jun 03, 2004 4:57 pm

hrm, turned monitor off, no change.

The resampled sound doesn't have nearly as much bass an it really is alot less dynamic almost if like a compressor squeezing the life out of it.

Interesting note here, i checked out the DC Offset on the resampled samples, an man there is a huge offset. However, correcting this offset distorts the sound very heavily.

Something is definitely wrong here. Ill shut the machine down, load the preset an perform the resampling procedure again.
drayon
 
Posts: 211
Joined: Thu Jul 24, 2003 3:46 am
Location: a void where there should be ecstasy

Postby illinformed » Fri Jun 04, 2004 11:06 am

Damn it, that sounds like a real pain.

I remember once on my non ultra emu I used to resample internally alot and had similar problems. The ones which sounded rubbish all had a huge DC offset, and this was mainly on the ones which had signals going through the old 24bit dual effects card which you could get for the non ultra. It would also be bad when using an lfo in the voice, and I remember the LFO would behave differently when resampling internally, the timing would go slightly out of sync.

I can't remember your setup Drayon.

Ever since having that inconsistent problem with my old non ultra, I've always resampled back from a desk using balanced cables. It got rid of the LFO sync and offset problems. Before I got my desk I used to loop back a balanced cable from Sub3 to my sample inputs (remembering to turn monitor off :shock: ;) ), not ideal but it got around the problems.
illinformed
 
Posts: 259
Joined: Thu Sep 11, 2003 2:12 pm
Location: London


Return to EOS:Troubleshooting