Chances are it's just some rogue signal left over from the format conversion - you should be recording in 44.1 for a CD master - it's just a lot simpler that way.

I'll look more into that hiss and why it might be there... I'm too hungry to think right now, but I will figure it out.

Chris

I made a mix for a friend of mine, and used some effects to get a mix over to a CD. I used 2 bridged mono limiter compressors and 1 *ph2n(cs bass resurector .to get the low from the LPs on to ADAT's 16 bits/48.200Khz of bandwith. I must say it sounded great, but when i converted the ADAT tape to CD i heard a very high hiss (only somewhat hearable on headphones and my 2 monitor speakers, but still anoying). Does ADAT have a bad 48 to 44 convertion or do i need to modulate to 44.1 on the master??

And I'm sure DJ Vastari can give you the technical reason as well. I believe it had something to do with ghost-signals appearing when converting to another sample rate. A bulletproof way of avoiding this sort of a hassle is to make sure you record on the same samplespeed as your eventual target will have, or a multiple of that samplespeed.

Yesterday I did some thinking on this issue and came up with an explanation on this thing.

When you sample/record digital you lose information. That's obvious, because you cannot effectively record >22kHz waves when you sample at 44kHz. So far so good.

Now when you have made a recording on 48kHz and try to simply resample it to 44.1kHz you get some sort of artifact sound. This is because, and now it becomes tricky, you are effectively using one extreme cut-off filter on top of another!. First you are cutting down to 24kHz and right after that you cut down to 22kHz. This causes some sort of cut-off pattern to appear (don't know the official name).

Think of it as luxaflex or window blinds or whatever. When you use 2 layers of luxaflex behind one another with one having a few more 'blades' than the other: (�=open �=closed)

������������������������������
������������������������������

When you add those two up you can see a strange pattern appear:

������������������������������

Note: if you did a resample from 88 to 44, you would not see (hear) this effect. Just draw it out yourself.

Hopes this helps,
Leo

Ah so it's a by-product of resampling to a bitrate that isn't a factor or multiple of the original bit-rate?

I know a few guys who record DATs but then down-sample them to 44.1... and they haven't had any problems. I would guess it depends on what program/hardware is doing the downsampling. Also, if this was such a big problem, nobody would be recording DATs; I'm sure a lot of producers go from DAT to CD because it's convienent. That or to a computer.

My guess is the hiss is left from the compressors. When you recorded, there might have been a slight high-frequency hiss which was amplified by the compressors...are you sure it's not on the DAT masters? Sometimes you don't catch things until the recording, etc is in the final stage. You might also try a different CD-player as well... some are noisier than others.

--
Dave

[This message has been edited by SpinThis! (edited 01-31-2001).]

I have heard the complaint before from other sources, that the sound changes more than you would expect when converting from one samplerate to another. I also know that if I'm right, there is a rather simple way to circumvent the problem, and that is to filter out the unwanted subsignals. I'll go and see if I can get any official info on the matter.