| Author |
Message |
/mr
Joined: Aug 05, 2007
Posts: 223
Location: Elektron City, Sweden
Audio files: 1
|
 Posted: Sun May 02, 2010 4:39 pm Post subject:
|
  |
|
| emdot_ambient wrote: | | /mr wrote: | Because of the varying bandwidths the sharpness of the filter peaks will have to be a little different between bands, but hopefully it will be worthwhile. It might be the world's first(?) fullrange vocoder, but still with a good voice quality.  |
Very hoopy piece of thinking...and makes total sense to me. |
The obvious drawbacks would be that the series of component values of the different bands are not as easily related, and that shifting the band matrix does not shift all bands the same number of octaves. And that Jürgen has already something else in mind.  |
|
|
Back to top
|
|
 |
zthee
Joined: Feb 20, 2008
Posts: 414
Location: Stockholm
|
| Posted: Sun May 09, 2010 5:31 pm Post subject:
|
|
|
22 bands?  
I'm in for one!
Are you still planning on keeping the filter bank morph function?
_________________
http://www.thehumancomparator.net/ |
|
|
Back to top
|
|
|
jhaible
Joined: May 25, 2007
Posts: 2014
Location: Germany
Audio files: 24
|
| Posted: Mon May 10, 2010 4:04 am Post subject:
|
|
|
| zthee wrote: | 22 bands?
I'm in for one!
Are you still planning on keeping the filter bank morph function? |
Just theoretical thinking and simulations, at the moment.
22 Bands is the EMS 5000, of course.
I found *that* would be easier to (re)design and to adjust than the 14-band I had originally in mind. Not sure if better sounding, though.
JH.
_________________
"I tell you the truth, if anyone says to this mountain, 'Go, throw yourself into the sea,' and does not doubt in his heart but believes that what he says will happen, it will be done for him. Therefore I tell you, whatever you ask for in prayer, believe that you have received it, and it will be yours." (Mk 11,23f) |
|
|
Back to top
|
|
|
Dego
Joined: Apr 22, 2008
Posts: 139
Location: Sweden
|
| Posted: Tue May 11, 2010 12:43 am Post subject:
|
|
|
| zthee wrote: | 22 bands?
I'm in for one! |
+1
This is a fantastic project! |
|
|
Back to top
|
|
|
SolderSmoke
Joined: May 21, 2007
Posts: 15
Location: Sweden
|
| Posted: Tue May 11, 2010 3:21 am Post subject:
|
|
|
| I second that. Count me in too! |
|
|
Back to top
|
|
|
jhaible
Joined: May 25, 2007
Posts: 2014
Location: Germany
Audio files: 24
|
| Posted: Tue May 11, 2010 5:57 am Post subject:
|
|
|
Well of course I wouldn't make a direct EMS clone.
More thinking brought up this:
I could use an equally simple filter design as EMS did in the 5000, but optimize it for 20 channels instead of 22. That, of course, means that you can use a standard 20x20 Ghielmetti matrix, a rather inexpensive small type with shorting pins, for cross-patching:
http://www.ghielmetti.ch/forums.html?nav=19,48,107&Category=Matrix-Boards-50VAC/6A&prod_id=55
Doesn't mean you need a matrix. You can go for 2 * 20 jacks a for cross patching a la Moog, or build it without cross patching at all. But a matrix would be nice, wouldn't it?
You will have to buy heaps of 1% capacitors, but this helps to keep trimpot count low, and I'll choose capacitor values that are easily available.
JH.
_________________
"I tell you the truth, if anyone says to this mountain, 'Go, throw yourself into the sea,' and does not doubt in his heart but believes that what he says will happen, it will be done for him. Therefore I tell you, whatever you ask for in prayer, believe that you have received it, and it will be yours." (Mk 11,23f) |
|
|
Back to top
|
|
|
numbertalk
Joined: May 05, 2008
Posts: 992
Location: Austin, TX
Audio files: 5
|
| Posted: Tue May 11, 2010 6:25 am Post subject:
|
|
|
| jhaible wrote: |
You will have to buy heaps of 1% capacitors, but this helps to keep trimpot count low, and I'll choose capacitor values that are easily available. |
1% or even .1%? |
|
|
Back to top
|
|
|
jhaible
Joined: May 25, 2007
Posts: 2014
Location: Germany
Audio files: 24
|
| Posted: Tue May 11, 2010 7:03 am Post subject:
|
|
|
| numbertalk wrote: | | jhaible wrote: |
You will have to buy heaps of 1% capacitors, but this helps to keep trimpot count low, and I'll choose capacitor values that are easily available. |
1% or even .1%? |
1% - anything else would be insane.
Where better precision is required, there will be trimming.
But I was actually surprised that you can get affordable thru-hole polypropylene caps in the 2- and 3-digit Nanofarad range with 1%. Which will reduce the need for trimming a lot.
I'll design the circuit such that you get away with just a handfull of different values, so you can go and buy bags of them.
JH.
_________________
"I tell you the truth, if anyone says to this mountain, 'Go, throw yourself into the sea,' and does not doubt in his heart but believes that what he says will happen, it will be done for him. Therefore I tell you, whatever you ask for in prayer, believe that you have received it, and it will be yours." (Mk 11,23f) |
|
|
Back to top
|
|
|
jhaible
Joined: May 25, 2007
Posts: 2014
Location: Germany
Audio files: 24
|
| Posted: Tue May 11, 2010 7:14 am Post subject:
|
|
|
| zthee wrote: |
Are you still planning on keeping the filter bank morph function? |
Yes. That's the main thing I take from the Sennheiser concept. Where EMS just crossfaded to the dry signal for silence bridgeing (sp?), Sennheiser morphed to a filtered version with arbitrarily set channel levels. That's what set the VSM apart, IMO.
You will be able to implement this in my version (needs 20 extra potentiometers), but as with so many other things: you don't have to, and can simply omit it.
You can build a minimal version like a EMS2000 or Moog/Bode, but with guts that are closer to an EMS5000, or you can build a monster with patchmatrix a la big Synton / EMS5000, and Filterbank function a la Sennheiser.
JH.
_________________
"I tell you the truth, if anyone says to this mountain, 'Go, throw yourself into the sea,' and does not doubt in his heart but believes that what he says will happen, it will be done for him. Therefore I tell you, whatever you ask for in prayer, believe that you have received it, and it will be yours." (Mk 11,23f) |
|
|
Back to top
|
|
|
/mr
Joined: Aug 05, 2007
Posts: 223
Location: Elektron City, Sweden
Audio files: 1
|
| Posted: Tue May 11, 2010 12:20 pm Post subject:
|
|
|
| jhaible wrote: | I could use an equally simple filter design as EMS did in the 5000, but optimize it for 20 channels instead of 22. That, of course, means that you can use a standard 20x20 Ghielmetti matrix, a rather inexpensive small type with shorting pins, for cross-patching:
http://www.ghielmetti.ch/forums.html?nav=19,48,107&Category=Matrix-Boards-50VAC/6A&prod_id=55
Doesn't mean you need a matrix. You can go for 2 * 20 jacks a for cross patching a la Moog, or build it without cross patching at all. But a matrix would be nice, wouldn't it? |
Oh yes. That is definitely an important design choice!
| jhaible wrote: | | You will have to buy heaps of 1% capacitors, but this helps to keep trimpot count low, and I'll choose capacitor values that are easily available. |
Sounds good. This worked out well in your frequency shifter.
Btw, when the boards are finished it would be nice with some advice for choosing good capacitor types... it's important that they stay within 1% for years and years to come. This is a device that should last for ages! |
|
|
Back to top
|
|
|
emdot_ambient
Joined: Nov 22, 2009
Posts: 667
Location: Frederick, MD
|
| Posted: Tue May 11, 2010 1:15 pm Post subject:
|
|
|
| jhaible wrote: | | ...or you can build a monster with patchmatrix a la big Synton / EMS5000, and Filterbank function a la Sennheiser... |
 More = More!
 |
|
|
Back to top
|
|
|
zthee
Joined: Feb 20, 2008
Posts: 414
Location: Stockholm
|
| Posted: Wed May 12, 2010 3:14 am Post subject:
|
|
|
| jhaible wrote: | | zthee wrote: |
Are you still planning on keeping the filter bank morph function? |
Yes. |
Great! I've never used a Sennheiser, so I have to admit I really don't understand the functionality... But I find the idea very interesting!
I'm so looking forward to this project! Thanks!
_________________
http://www.thehumancomparator.net/ |
|
|
Back to top
|
|
|
mono-poly
Joined: Jul 07, 2004
Posts: 937
Location: Rotterdam, Netherlands
Audio files: 2
|
| Posted: Wed May 12, 2010 4:31 am Post subject:
|
|
|
Sennheiser is super overated.
I compared it with the EMS 3000 and the last one imo was much nicer and musical sounding and to work with. |
|
|
Back to top
|
|
|
jhaible
Joined: May 25, 2007
Posts: 2014
Location: Germany
Audio files: 24
|
| Posted: Wed May 12, 2010 6:30 am Post subject:
|
|
|
| mono-poly wrote: | Sennheiser is super overated.
I compared it with the EMS 3000 and the last one imo was much nicer and musical sounding and to work with. |
I don't know - I had the chance to play with a Sennheiser twice. First time was an overwhealming experience, I was really blown away by te sounds it produced.
2nd time I hardly got anything usefull out of it, had a hard time to get the input levels right, got very bad signal to noise ratio. My conclusion was the second unit must have been broken. (They both were out of my price range, so I lost track of them.)
From an engineering point of view, it's a strange beast. Very complex filter topology - 3 stages of staggered-tuned Sallen-Key-BPF, built with discrete transistors and select-on-test resistor values. Probably the worst thing you could do to keep manufacturing cost reasonable, but arguably the best way to achieve really precise filter courves and consequently, clarity of simulated speech. There's an example on Greenslade's "Pentateuch" album that is (IMO) musically awfull, but awesome in terms of clarity of speech simulation.
The envelope detectors are similarly over-engineered. I compared a lot of full wave rectifiers some years ago, and found that Sennheiser's variation had the best dynamic range. You buy this with a lot of components, and EMS found a way to be almost as good with a ridiculously low component count. And then, Sennheiser paired this super-envelope-detection with a cheap switched modulator for VCAs.  And the had no slew control or freeze feature.
But that filterbank / vocoder morphing thing of the Sennheiser is a great feature. Never had the chance to try it myself, but there's a demo floating around on the web where natural noises start to speak, without a glitch or change of tone colour in between - really, really nice. And not much is needed for this, electronically. Mainly the extra potentiometers.
What was it that you didn't like on the VSM, or loved better on the EMS3000?
Jh.
_________________
"I tell you the truth, if anyone says to this mountain, 'Go, throw yourself into the sea,' and does not doubt in his heart but believes that what he says will happen, it will be done for him. Therefore I tell you, whatever you ask for in prayer, believe that you have received it, and it will be yours." (Mk 11,23f) |
|
|
Back to top
|
|
|
StephenGiles
Joined: Apr 17, 2006
Posts: 507
Location: England
|
| Posted: Wed May 12, 2010 10:45 am Post subject:
|
|
|
Are the Sennheiser Vocoder Schematics in circulation, I have a passion for complicated envelope detectors (as well as my wife of course  ) |
|
|
Back to top
|
|
|
mono-poly
Joined: Jul 07, 2004
Posts: 937
Location: Rotterdam, Netherlands
Audio files: 2
|
| Posted: Wed May 12, 2010 3:06 pm Post subject:
|
|
|
Juergen, somehow the EMS sounded easyer right to me maybe the Sennheiser is better on paper.
But in the end you have to judge it by how you get good results and your ears like it.
The price difference between the two units is totaly not worth it imo. |
|
|
Back to top
|
|
|
zthee
Joined: Feb 20, 2008
Posts: 414
Location: Stockholm
|
|
|
Back to top
|
|
|
dcreatorx
Joined: Jun 02, 2009
Posts: 17
Location: Barcelona
|
| Posted: Fri May 14, 2010 4:03 am Post subject:
wow |
|
|
| Wow that design kicks ass !!!! |
|
|
Back to top
|
|
|
jhaible
Joined: May 25, 2007
Posts: 2014
Location: Germany
Audio files: 24
|
| Posted: Fri May 14, 2010 4:10 am Post subject:
|
|
|
Looks great! Yes, sliders make sense for 20 bands + 20 filterbank settings ...
(But that's up to each who builds it, of course.)
Ahh, and so you want channel LEDs, also!
You're really coaxing me into realizing this project, it seems ...
But the *name* is set already. I think I have mentioned it already, and if not, you caertainly won't be surprised. It's "Living Vocoder", of course.
JH.
_________________
"I tell you the truth, if anyone says to this mountain, 'Go, throw yourself into the sea,' and does not doubt in his heart but believes that what he says will happen, it will be done for him. Therefore I tell you, whatever you ask for in prayer, believe that you have received it, and it will be yours." (Mk 11,23f) |
|
|
Back to top
|
|
|
/mr
Joined: Aug 05, 2007
Posts: 223
Location: Elektron City, Sweden
Audio files: 1
|
| Posted: Fri May 14, 2010 4:37 am Post subject:
|
|
|
Wow, juste!
I love your graphic design, as always. Sliders is definitely the way to go, ergonomically and visually much better than knobs. I'd place the input/output sections at the top, though, since I expect them to be "less playable" than the other controls. I love the fluffy wrist rest cushion.
Hmmm, "Carrier analysis section"?
It's the Signal that is analyzed, not the Carrier.
PS: The name is not bad at all. My Living Vocoder will be called Moroder.  |
|
|
Back to top
|
|
|
mono-poly
Joined: Jul 07, 2004
Posts: 937
Location: Rotterdam, Netherlands
Audio files: 2
|
| Posted: Fri May 14, 2010 5:00 am Post subject:
|
|
|
| Jeez that looks awesome Zthee! |
|
|
Back to top
|
|
|
zthee
Joined: Feb 20, 2008
Posts: 414
Location: Stockholm
|
| Posted: Fri May 14, 2010 5:23 am Post subject:
|
|
|
Thanks guys!
| /mr wrote: | Hmmm, "Carrier analysis section"?
It's the Signal that is analyzed, not the Carrier. |
I wasn't sure if the signal or the carrier was analysed. Though if I post it and I'm wrong I'm bound to find out
_________________
http://www.thehumancomparator.net/ |
|
|
Back to top
|
|
|
/mr
Joined: Aug 05, 2007
Posts: 223
Location: Elektron City, Sweden
Audio files: 1
|
| Posted: Fri May 14, 2010 6:22 am Post subject:
|
|
|
| zthee wrote: | | I wasn't sure if the signal or the carrier was analysed. Though if I post it and I'm wrong I'm bound to find out |
Hehe!
The spectrum of the Signal (from your voice) gets analyzed, and this spectrum is applied to the Carrier, which carries the spectrum at the output. |
|
|
Back to top
|
|
|
zthee
Joined: Feb 20, 2008
Posts: 414
Location: Stockholm
|
| Posted: Fri May 14, 2010 10:19 am Post subject:
|
|
|
| jhaible wrote: | | Ahh, and so you want channel LEDs, also! |
Yes! Channel LEDs would be great! And when looking at your teaser layout, it seems like you're adding LEDs for all the switches too? That would be a great option!
| /mr wrote: | Hehe!
The spectrum of the Signal (from your voice) gets analyzed, and this spectrum is applied to the Carrier, which carries the spectrum at the output. |
Yeah, I realized that if I'd googled some I'd been safe from the start, hehe
_________________
http://www.thehumancomparator.net/ |
|
|
Back to top
|
|
|
jhaible
Joined: May 25, 2007
Posts: 2014
Location: Germany
Audio files: 24
|
| Posted: Fri May 14, 2010 11:13 am Post subject:
|
|
|
http://www.jhaible.de/vocoder/Living_Vocoder_20_simulation.jpg
JH.
_________________
"I tell you the truth, if anyone says to this mountain, 'Go, throw yourself into the sea,' and does not doubt in his heart but believes that what he says will happen, it will be done for him. Therefore I tell you, whatever you ask for in prayer, believe that you have received it, and it will be yours." (Mk 11,23f) |
|
|
Back to top
|
|
|
|
Forum index » DIY Hardware and Software » Jürgen Haible designs
FAQ
Calendar
Search
Memberlist
Usergroups
Register
Profile
Log in to check your private messages
Log in
Chat Room
