Emax

I did some more testing with EMXP...

The answer to my problem with distortion really has to do with headroom, not with the conversion algorithm or with Emax's 12bit sound!

If you convert a sample that is normalized to 0 dB with EMXP it will be distorted after the conversion. If you want to be sure there is no distortion you need to normalize the files to -4 dB before converting them to Emax format.

If anybody is interested in the results, I have some audio files and pictures demonstrating this with a 909 kick drum. 

Even so, there will still be a difference between a converted .wav file and a sample taken manually with Emax. This is only natural. Which one sounds better really depends on the sample and your taste.




--- In emax@yahoogroups.com, "esynthesist" <esynthesist@...> wrote:
>
> I did some tests myself today.
> 
> The lower quality of the Emax samples is not caused by the lower samplerate, but by the 12-to-8 bit compression algorithm.
> If you would convert the samplerate of a 44100Hz WAV to 41667Hz first with some audio tool, and then load this WAV into an Emax-1 bank with EMXP, the same "lower quality" would be audible.
> I also loaded some 44100 Hz WAVs into an EMAX-II bank with EMXP, making sure it converts them to 41667 Hz. Result: the sound is top quality on the Emax-II with both samplerates. So the samplerate converter in EMXP is definitely not causing the difference.
> 
> Of course you could ask me now the enhance the 12-to-8 bit compression algorithm in EMXP :-)
> 
> But when I designed this algorithm a few years ago, I already noticed these quality issues on both the Emax and the Emulator II, especially with drum sounds (I used TR808 samples).
> So I did some additional tests at that time: I checked the compression/decompression algorithms in SoundDesigner and they suffered from the same problem.
> I also saved some "good" 16 bit banks on the Emax-II as "compressed 8 bit" Emax-I banks and loaded those on the Emax-I: the same distortion problems appeared.
> So I concluded that my algorithm was as good (as bad ?) as the ones used by Emu and Digidesign, and I left them as they were.
> 
> The hardware compression on Emax-I during sampling is much better indeed. I guess there must be an explanation, but I'm not a hardware specialist.
> 
> ///E-Synthesist
> 
> --- In emax@yahoogroups.com, "jammie" <jammie.emma@> wrote:
> >
> > drum samples sound better at lower sample rates any way as they dont contain the whole frequency spectrum 
> > 
> > thats why hiphop producers prefer the 12bit sample sound and the emax is converted to 12bit so you get 4bits of data diference that means more stepping but it adds to the quality hence why people still buy the sp1200 which is an emax engine
> >   ----- Original Message ----- 
> >   From: bermudaben2006 
> >   To: emax@yahoogroups.com 
> >   Sent: Sunday, July 12, 2009 10:04 AM
> >   Subject: [emax] Re: Converting wav files to Emax samples using EMXP
> > 
> > 
> > 
> > 
> > 
> >   Well, if you could take samples with Emax at both 42 and 44 KHz you probably wouldn't hear any difference. But converting 44.1 Khz samples is another thing, because any samle rate conversion results in a degradation of quality. 
> > 
> >   I did some testing yesterday with drum samples that I either played back and sampled at 42 Khz or converted from 44 Khz to 42 Khz using EMXP, and the difference was quite radical! For example, the 909 bass drum sample I converted was quite distorted (the original is normalized at -0.3 dB). On the other hand, when I sampled it manually it sounded really good. Also, converting hi hat samples with EMXP added a lot of low frequencies that are not present in the original. Again, manually sampling them gave better results.
> > 
> >   Do you know if the .wav files I want to convert with EMXP should have some headroom for better results?
> > 
> >   --- In emax@yahoogroups.com, "esynthesist" <esynthesist@> wrote:
> >   >
> >   > In fact EMXP never allows samplerates higher than 41667 Hz for Emax-1.
> >   > I wanted to be sure that the banks would be compatible with all Emax-1 samplers.
> >   > 
> >   > The maximum samplerate at which Emax-1 can sample is 41667 Hz (Emax-II even only 39 kHz) 
> >   > As far as I know the option to change the samplerate to 44100 Hz on Emax (in the Digital Effects section) was only added for SE versions. I only have an SE and Plus version (which both support playback at 44100 Hz), so I couldn't test if non-SE versions are also able to playback samples at 44100 Hz.
> >   > 
> >   > If someone can confirm that ALL emax-1 samplers support playback at 44.1 kHz, I might change EMXP. If not all Emax samplers are compatible, I could still add the 44100 Hz option in EMXP "on user's risk" :-)
> >   > But... does anyone hear the difference between 41.7 and 44.1 kHz samples on the Emax ??
> >   > 
> >   > ///E-Synthesist
> >   > 
> >   > --- In emax@yahoogroups.com, "jammie" <jammie.emma@> wrote:
> >   > >
> >   > > it might be that to get all samples to fit in the emaxs memory that a bit of sample rate is lowered
> >   > > ----- Original Message ----- 
> >   > > From: bermudaben2006 
> >   > > To: emax@yahoogroups.com 
> >   > > Sent: Saturday, July 11, 2009 2:46 PM
> >   > > Subject: [emax] Converting wav files to Emax samples using EMXP
> >   > > 
> >   > > 
> >   > > 
> >   > > 
> >   > > 
> >   > > This is a question for esynthesist or anyone who uses his software:
> >   > > 
> >   > > Why does EMXP convert my 44.1 KHz wav files to 41.7 Khz Emax samples, given that Emax supports sample rates of up to 44.1 Khz?
> >   > > 
> >   > > Thanks,
> >   > > Ben
> >   > > 
> >   > > 
> >   > > 
> >   > > 
> >   > > 
> >   > > 
> >   > > ----------------------------------------------------------
> >   > > 
> >   > > 
> >   > > 
> >   > > No virus found in this incoming message.
> >   > > Checked by AVG - www.avg.com 
> >   > > Version: 8.5.387 / Virus Database: 270.13.10/2231 - Release Date: 07/11/09 05:57:00
> >   > > 
> >   > > 
> >   > > [Non-text portions of this message have been removed]
> >   > >
> >   >
> > 
> > 
> > 
> >   
> > 
> > 
> > ------------------------------------------------------------------------------
> > 
> > 
> > 
> >   No virus found in this incoming message.
> >   Checked by AVG - www.avg.com 
> >   Version: 8.5.387 / Virus Database: 270.13.11/2232 - Release Date: 07/11/09 17:56:00
> > 
> > 
> > [Non-text portions of this message have been removed]
> >
>