Author |
Message |
xav
Joined: Mar 21, 2005 Posts: 164 Location: paris
Audio files: 8
G2 patch files: 7
|
Posted: Mon Feb 06, 2012 8:22 pm Post subject:
How can I build a square wave like this? |
|
|
Look at those square waveforms, they look all different. The one I would like to copy is produced by the Studio Electronics ATC-1, an analog device. I tried different ways, with shapers modulated by the square, but it's far from that sweet and powerful analog sound. Varice's classic VCO style multi waveform Oscillator Bank is the best tool I found, to build such a waveform from a visual point of view (especially the variation number 2) but the sound isn't as beautiful as the analog one.
If you have ny idea....
Thank you
Xavier
Description: |
|
Filesize: |
7.66 KB |
Viewed: |
17085 Time(s) |
|
Description: |
|
Filesize: |
8.45 KB |
Viewed: |
17085 Time(s) |
|
Description: |
|
Filesize: |
8.3 KB |
Viewed: |
17085 Time(s) |
|
Description: |
|
Filesize: |
8.21 KB |
Viewed: |
17086 Time(s) |
|
|
|
Back to top
|
|
|
varice
Joined: Dec 29, 2004 Posts: 961 Location: Northeastern shore of Toledo Bend
Audio files: 29
G2 patch files: 54
|
Posted: Mon Feb 06, 2012 10:27 pm Post subject:
|
|
|
The ATC and Predator images that you posted look like square waves being fed through a high pass filter with a relatively high cutoff frequency.
To reproduce that ATC waveshape in the G2, you need to start with a square wave and then add a HPF at some point before the output module. The resulting waveshape will depend on the frequency (pitch) of the square wave, the cutoff frequency of the HPF, and the cutoff slope of the HPF (6dB, 12dB, etc.).
By the way, the CEM Osc Bank building block patch already has a 6dB/octave HPF in the FX Area named “DC Block”. You can adjust the cutoff frequency and slope with the square (pulse) wave selected until it resembles the ATC waveshape.
Description: |
|
Filesize: |
35.78 KB |
Viewed: |
469 Time(s) |
This image has been reduced to fit the page. Click on it to enlarge. |
|
Description: |
|
Filesize: |
33.99 KB |
Viewed: |
475 Time(s) |
This image has been reduced to fit the page. Click on it to enlarge. |
|
_________________ varice |
|
Back to top
|
|
|
xav
Joined: Mar 21, 2005 Posts: 164 Location: paris
Audio files: 8
G2 patch files: 7
|
Posted: Tue Feb 07, 2012 1:39 am Post subject:
|
|
|
Thank you very much Varice,`What a good knowledge of waveforms deformations !!!
I think it is a very good way to design the wave. I will compare this with the ATC as soon as possible. The sound seemed wide.... but this kind of description isn't helpful...
All the best
xavier |
|
Back to top
|
|
|
xav
Joined: Mar 21, 2005 Posts: 164 Location: paris
Audio files: 8
G2 patch files: 7
|
Posted: Tue Feb 07, 2012 3:34 pm Post subject:
|
|
|
Very nice advice Varice!!! and your patch is definitely useful... I love it.
Here are the new waveforms and the sounds. The original G2 square is more agressive (and more square). The ATC and your patch are quite similar... there is still something sweet I cannot really imitate in the ATC... but it is a detail.
Thank you very much
All the best
xavier
Description: |
|
Filesize: |
8.27 KB |
Viewed: |
16980 Time(s) |
|
Description: |
|
Filesize: |
8.24 KB |
Viewed: |
16980 Time(s) |
|
Description: |
|
Filesize: |
8.05 KB |
Viewed: |
16980 Time(s) |
|
Description: |
|
Download |
Filename: |
ATC.mp3 |
Filesize: |
69.73 KB |
Downloaded: |
919 Time(s) |
Description: |
|
Download |
Filename: |
CEM + HPF.mp3 |
Filesize: |
69.12 KB |
Downloaded: |
979 Time(s) |
Last edited by xav on Wed Feb 08, 2012 2:28 am; edited 1 time in total |
|
Back to top
|
|
|
xav
Joined: Mar 21, 2005 Posts: 164 Location: paris
Audio files: 8
G2 patch files: 7
|
Posted: Tue Feb 07, 2012 3:35 pm Post subject:
|
|
|
... and the last mp3
I posted it twice, because it seems overloaded when I normalizes it... but I normalized the same way the other samples... Maybe that pure squarewave is more powerful since it stays longer around the peak levels (I hope that phrase makes sense in English... In french, we use food words when unable to describe scientific sensations: "ça a trop la patate" - "it has too much potato" ).
Sorry for the first bad format uploads.
Description: |
|
Download |
Filename: |
G2 osc without normalization.mp3 |
Filesize: |
66.37 KB |
Downloaded: |
942 Time(s) |
Description: |
|
Download |
Filename: |
G2 osc.mp3 |
Filesize: |
66.37 KB |
Downloaded: |
922 Time(s) |
Last edited by xav on Wed Feb 08, 2012 2:40 am; edited 1 time in total |
|
Back to top
|
|
|
varice
Joined: Dec 29, 2004 Posts: 961 Location: Northeastern shore of Toledo Bend
Audio files: 29
G2 patch files: 54
|
Posted: Tue Feb 07, 2012 5:07 pm Post subject:
|
|
|
Hey Xavier, glad that you found that patch useful!
I’ll check your sound clips later. I’m going to my sister’s place for dinner tonight!
All the best “Evariste LeMire” (as my name may be if my French ancestor Pierre LeMire had not immigrated to the New World almost 300 years ago…). _________________ varice |
|
Back to top
|
|
|
varice
Joined: Dec 29, 2004 Posts: 961 Location: Northeastern shore of Toledo Bend
Audio files: 29
G2 patch files: 54
|
Posted: Wed Feb 08, 2012 3:54 pm Post subject:
|
|
|
The ATC and CEM + HPF mp3s still have clipping distortion on the low frequency notes probably caused by the normalization process. The higher notes are not clipped.
You have the two wave shapes very similar. It seems to me that the ATC recording has a little bit of high frequency roll off compared to the CEM + HPF. Maybe some LPF treatment in the G2 patch would help. Do you have a frequency spectrum analyzer? Comparing the square wave spectrums of your ATC-1 synth and G2 CEM patch may reveal other differences. I would use my analyzer, but the mp3 clips are too short and have mixed pitches (and I wouldn’t recommend trying to do an accurate analysis of an mp3 recording anyway). I use the free software program “Visual Analyser” for Windows from here:
http://www.sillanumsoft.org/ _________________ varice |
|
Back to top
|
|
|
xav
Joined: Mar 21, 2005 Posts: 164 Location: paris
Audio files: 8
G2 patch files: 7
|
Posted: Thu Feb 09, 2012 2:48 am Post subject:
|
|
|
Oh, you are completely right! They clip. I normalized them with Ableton Live... Danger!
And I noticed this : the diference is much more noticeable with monitors (I use Dynaudio BM15) than with headphone (especially my Koss portapro). It seems to be a little snob, and I don't like that... but it's what I feel. With a LPF, it's better... What I try to know is : do I really need this ATC? is it really possible to get the same smooth and wide sound?
So these are the non normalized looped versions... unfortunately, I think it is not possible to post non mp3 sample here.
Thank you for the link. I soon have a try to that program... There are more analyze softwares for the PC... I should switch...
Thank you for the interest and the advices
Description: |
|
Download |
Filename: |
8 ATC.mp3 |
Filesize: |
48.59 KB |
Downloaded: |
1016 Time(s) |
Description: |
|
Download |
Filename: |
CEM Hpf.mp3 |
Filesize: |
49.52 KB |
Downloaded: |
936 Time(s) |
|
|
Back to top
|
|
|
blue hell
Site Admin
Joined: Apr 03, 2004 Posts: 24079 Location: The Netherlands, Enschede
Audio files: 278
G2 patch files: 320
|
Posted: Thu Feb 09, 2012 7:16 am Post subject:
|
|
|
xav wrote: | I think it is not possible to post non mp3 sample here. |
It should allow for wav, ogg, flac, mp4 ... couple of others ... what did you want to post? _________________ Jan
also .. could someone please turn down the thermostat a bit.
|
|
Back to top
|
|
|
ian-s
Joined: Apr 01, 2004 Posts: 2669 Location: Auckland, New Zealand
Audio files: 42
G2 patch files: 626
|
Posted: Thu Feb 09, 2012 9:45 am Post subject:
|
|
|
xav wrote: | do I really need this ATC? is it really possible to get the same smooth and wide sound? |
When I hear wide sound I thing of those "wide" switches on old stereos. They worked by mixing HPF filtered and inverted copies of the opposite channels.
The human ear is very sensitive to minute changes at high frequencies, we depend on it for accurate sound location.
I notice that the G2 and ATC samples have slightly different duty cycles. The G2 is 50:50 (easy for dsp) while the ATC is (as best I can measure) 49.8:50.2
If you are using a SHPB oscillator, try adding a very small constant offset (maybe via a mixer input) to the Shape input.
Are you testing the two waves through the same LPF? |
|
Back to top
|
|
|
xav
Joined: Mar 21, 2005 Posts: 164 Location: paris
Audio files: 8
G2 patch files: 7
|
Posted: Thu Feb 09, 2012 11:25 am Post subject:
|
|
|
Blue Hell wrote: | xav wrote: | I think it is not possible to post non mp3 sample here. |
It should allow for wav, ogg, flac, mp4 ... couple of others ... what did you want to post? |
I only tried aif... I didn't know wav is possible to post wav. I'll make a try as soon as I can... I just spilled tea on my computer, so I try to dry it before. |
|
Back to top
|
|
|
blue hell
Site Admin
Joined: Apr 03, 2004 Posts: 24079 Location: The Netherlands, Enschede
Audio files: 278
G2 patch files: 320
|
Posted: Thu Feb 09, 2012 12:22 pm Post subject:
|
|
|
xav wrote: | I only tried aif... |
Added aif too now on the list of allowed extensions.
Hope your computer didn't end up too bad ... _________________ Jan
also .. could someone please turn down the thermostat a bit.
|
|
Back to top
|
|
|
xav
Joined: Mar 21, 2005 Posts: 164 Location: paris
Audio files: 8
G2 patch files: 7
|
Posted: Thu Feb 09, 2012 12:42 pm Post subject:
|
|
|
Thank you Jan....
The actual format of my hard drive is .water |
|
Back to top
|
|
|
blue hell
Site Admin
Joined: Apr 03, 2004 Posts: 24079 Location: The Netherlands, Enschede
Audio files: 278
G2 patch files: 320
|
Posted: Thu Feb 09, 2012 12:58 pm Post subject:
|
|
|
Liguid drives? _________________ Jan
also .. could someone please turn down the thermostat a bit.
|
|
Back to top
|
|
|
xav
Joined: Mar 21, 2005 Posts: 164 Location: paris
Audio files: 8
G2 patch files: 7
|
Posted: Thu Feb 09, 2012 4:24 pm Post subject:
|
|
|
Here are the wav files.
it's interesting to visualize a 5 seconds waveform. It reveals slow modulations not noticeable in one cycle. I already found out that those modulations make a real live in organ simple waves, hard to imitate with the G2 (for me).
Description: |
|
Download |
Filename: |
0002 ATC.wav |
Filesize: |
1.21 MB |
Downloaded: |
877 Time(s) |
Description: |
|
Download |
Filename: |
0002 CEM.wav |
Filesize: |
1.21 MB |
Downloaded: |
895 Time(s) |
|
|
Back to top
|
|
|
varice
Joined: Dec 29, 2004 Posts: 961 Location: Northeastern shore of Toledo Bend
Audio files: 29
G2 patch files: 54
|
Posted: Fri Feb 10, 2012 3:58 pm Post subject:
|
|
|
xav wrote: | Here are the wav files... |
These wave files are much better
xav wrote: | ...it's interesting to visualize a 5 seconds waveform. It reveals slow modulations not noticeable in one cycle... |
In the ATC wave? Are you sure that you have all modulation sources at zero?
It looks like the G2 patch waveshape needs more tweaking. The scope reveals that the curve of the wave is a little more aggressive compared to the ATC. If you have the G2 CEM HPF slope set to 12dB, try 6dB instead. You may need to adjust the HPF cutoff frequency also. The spectrum also reveals that the G2 CEM has a weaker fundamental frequency (about -3dB) versus harmonics when compared to the ATC spectrum.
ian-s wrote: | … I notice that the G2 and ATC samples have slightly different duty cycles. The G2 is 50:50 (easy for dsp) while the ATC is (as best I can measure) 49.8:50.2
If you are using a SHPB oscillator, try adding a very small constant offset (maybe via a mixer input) to the Shape input. |
Ian mentions a very important point. The duty cycle of the two synth waves need to be as close as possible. The scope shows that the duty cycles are close, but the spectrum of the ATC reveals that there is a stronger second (even) harmonic at about 80Hz compared to the G2 CEM. This could be the result of the ATC wave not being perfectly square with a 50/50 duty cycle. A perfect square wave will have predominately odd harmonics with very little even harmonics. Even the slightest change in duty cycle will cause a noticeable difference in the timbre of the wave. The CEM block patch has a “P Width” shape control that you can use to adjust the pulse width (duty cycle).
I don’t understand what you mean when you say that the ATC-1 square wave sounds “wide”. That seems to me to refer to a stereo effect. This synth has a mono output, correct?
xav wrote: | … And I noticed this : the diference is much more noticeable with monitors (I use Dynaudio BM15) than with headphone (especially my Koss portapro)… |
For critical sound comparison, I would recommend using good headphones. Using stereo monitors in a reflective room with any echo/reverb will complicate things. Even the smallest change in the position of your head relative to the monitors will affect the sound that you hear. But, if you do use monitors, be sure to use only one for comparing mono sources (e.g. mix both the ATC-1 and G2 to the left monitor only).
Description: |
|
Filesize: |
31.42 KB |
Viewed: |
297 Time(s) |
This image has been reduced to fit the page. Click on it to enlarge. |
|
Description: |
|
Filesize: |
31.5 KB |
Viewed: |
289 Time(s) |
This image has been reduced to fit the page. Click on it to enlarge. |
|
Description: |
|
Filesize: |
38.62 KB |
Viewed: |
237 Time(s) |
This image has been reduced to fit the page. Click on it to enlarge. |
|
Description: |
|
Filesize: |
39.46 KB |
Viewed: |
259 Time(s) |
This image has been reduced to fit the page. Click on it to enlarge. |
|
_________________ varice |
|
Back to top
|
|
|
xav
Joined: Mar 21, 2005 Posts: 164 Location: paris
Audio files: 8
G2 patch files: 7
|
Posted: Sat Feb 11, 2012 8:08 am Post subject:
|
|
|
varice wrote: | xav wrote: | Here are the wav files... |
These wave files are much better
|
So different? I'm used to listen to mp3... I'll be more exigent... And I really don't know why Ableton cuts the waveform in the normalize process.
varice wrote: |
xav wrote: | ...it's interesting to visualize a 5 seconds waveform. It reveals slow modulations not noticeable in one cycle... |
In the ATC wave? Are you sure that you have all modulation sources at zero?
|
Yes: no modulation... wouldn't be the analog lack of precision?
varice wrote: |
It looks like the G2 patch waveshape needs more tweaking. The scope reveals that the curve of the wave is a little more aggressive compared to the ATC. If you have the G2 CEM HPF slope set to 12dB, try 6dB instead. You may need to adjust the HPF cutoff frequency also. The spectrum also reveals that the G2 CEM has a weaker fundamental frequency (about -3dB) versus harmonics when compared to the ATC spectrum.
|
I understand and will try more HPF settings... it's actually a 6db. But how can I change the fundamental frequency? The spectrum is a great tool i should use.
[/quote]
varice wrote: |
ian-s wrote: | … I notice that the G2 and ATC samples have slightly different duty cycles. The G2 is 50:50 (easy for dsp) while the ATC is (as best I can measure) 49.8:50.2
If you are using a SHPB oscillator, try adding a very small constant offset (maybe via a mixer input) to the Shape input. |
Ian mentions a very important point. The duty cycle of the two synth waves need to be as close as possible. The scope shows that the duty cycles are close, but the spectrum of the ATC reveals that there is a stronger second (even) harmonic at about 80Hz compared to the G2 CEM. This could be the result of the ATC wave not being perfectly square with a 50/50 duty cycle. A perfect square wave will have predominately odd harmonics with very little even harmonics. Even the slightest change in duty cycle will cause a noticeable difference in the timbre of the wave. The CEM block patch has a “P Width” shape control that you can use to adjust the pulse width (duty cycle).
|
Alright, I understand now the duty cycle... it is the Pulse width, isn't it? I tried to reach the "perfect square". It is nothing but perfect in comparison with the original G2... or the real perfect Ableton Analog square.
I tuned the Pulse Width by ear. The graduation is from 0 to 127, and the value that gives me the sensation to be between 2 identical states is 58. All the other values seem to give more even harmonics (or more harmonics). It seems even more difficult to imitates the wave of the ATC in other pulse width values.
[/quote]
varice wrote: |
I don’t understand what you mean when you say that the ATC-1 square wave sounds “wide”. That seems to me to refer to a stereo effect. This synth has a mono output, correct?
|
You're right... it isn't the right word, and I might not find it since it's difficult to explain the sensation. Maybe I should say ATC wave has more life and deepness. But with all your tips and advices, the difference is less perceptible.
varice wrote: |
xav wrote: | … And I noticed this : the diference is much more noticeable with monitors (I use Dynaudio BM15) than with headphone (especially my Koss portapro)… |
For critical sound comparison, I would recommend using good headphones. Using stereo monitors in a reflective room with any echo/reverb will complicate things. Even the smallest change in the position of your head relative to the monitors will affect the sound that you hear. But, if you do use monitors, be sure to use only one for comparing mono sources (e.g. mix both the ATC-1 and G2 to the left monitor only). |
Usually, I listen to differences with headphone, like Sony MDR7506 for example, but for those squarewaves, without moving the head, I think I have a better point of comparison, maybe thanks to the high bass energy. |
|
Back to top
|
|
|
varice
Joined: Dec 29, 2004 Posts: 961 Location: Northeastern shore of Toledo Bend
Audio files: 29
G2 patch files: 54
|
Posted: Sat Feb 11, 2012 3:20 pm Post subject:
|
|
|
xav wrote: | varice wrote: | xav wrote: | Here are the wav files... |
These wave files are much better
|
So different? I'm used to listen to mp3... |
In my opinion, an mp3 recording is OK for casual listening, but for critical listening (or sound analysis) I prefer 16bit/44.1kHz or better digital recordings.
varice wrote: |
xav wrote: | ...it's interesting to visualize a 5 seconds waveform. It reveals slow modulations not noticeable in one cycle... |
In the ATC wave? Are you sure that you have all modulation sources at zero?
Yes: no modulation... wouldn't be the analog lack of precision? |
Yes, that could be an imperfection that adds to the “charm” of analog synths.
xav wrote: | varice wrote: |
It looks like the G2 patch waveshape needs more tweaking. The scope reveals that the curve of the wave is a little more aggressive compared to the ATC. If you have the G2 CEM HPF slope set to 12dB, try 6dB instead. You may need to adjust the HPF cutoff frequency also. The spectrum also reveals that the G2 CEM has a weaker fundamental frequency (about -3dB) versus harmonics when compared to the ATC spectrum.
|
I understand and will try more HPF settings... it's actually a 6db. But how can I change the fundamental frequency?... |
You may have the CEM block HPF cutoff frequency set too high, that will reduce the level of the fundamental. Try a lower freq setting. Using the G2 Demo software, I got the CEM block waveshape closer with a 6dB HPF set to 29.1Hz. But you need to set this value lower using your G2 synth because there is in effect a 6dB HPF (DC blocking cap) already built into the output of the G2 (as there obliviously is in the ATC-1 also). In fact, the whole reason that we are going through this G2 HPF exercise may be because the ATC-1 apparently has a smaller value DC blocking cap than the G2, which is causing the ATC-1 square wave to exhibit the higher HPF cutoff shape that it has.
Something I didn’t mention in my previous reply: the ATC spectrum confirms the slight roll off of high frequencies that I mentioned hearing earlier when compared to the CEM spectrum. The stronger fundamental along with the HF roll off might be why you describe the ATC-1 square as sounding smoother than the G2. I would recommend trying the G2 EQ modules to tweak the sound a little more closely.
xav wrote: | ...The spectrum is a great tool i should use... |
Our ears are good at perceiving the differences when comparing sounds, but it is not always apparent *why* the sounds are different. A scope and a spectrum analyzer are good tools to help reveal the reasons.
xav wrote: | varice wrote: |
ian-s wrote: | … I notice that the G2 and ATC samples have slightly different duty cycles. The G2 is 50:50 (easy for dsp) while the ATC is (as best I can measure) 49.8:50.2
If you are using a SHPB oscillator, try adding a very small constant offset (maybe via a mixer input) to the Shape input. |
Ian mentions a very important point. The duty cycle of the two synth waves need to be as close as possible. The scope shows that the duty cycles are close, but the spectrum of the ATC reveals that there is a stronger second (even) harmonic at about 80Hz compared to the G2 CEM. This could be the result of the ATC wave not being perfectly square with a 50/50 duty cycle. A perfect square wave will have predominately odd harmonics with very little even harmonics. Even the slightest change in duty cycle will cause a noticeable difference in the timbre of the wave. The CEM block patch has a “P Width” shape control that you can use to adjust the pulse width (duty cycle).
|
Alright, I understand now the duty cycle... it is the Pulse width, isn't it? I tried to reach the "perfect square". It is nothing but perfect in comparison with the original G2... or the real perfect Ableton Analog square.
I tuned the Pulse Width by ear. The graduation is from 0 to 127, and the value that gives me the sensation to be between 2 identical states is 58. All the other values seem to give more even harmonics (or more harmonics). It seems even more difficult to imitates the wave of the ATC in other pulse width values. |
The P Width control might be too course. You can get finer control around the 50/50 duty cycle by reducing the shape modulation input control of the Pulse osc module. _________________ varice |
|
Back to top
|
|
|
xav
Joined: Mar 21, 2005 Posts: 164 Location: paris
Audio files: 8
G2 patch files: 7
|
Posted: Sun Feb 12, 2012 12:31 pm Post subject:
|
|
|
varice wrote: | It seems to me that the ATC recording has a little bit of high frequency roll off compared to the CEM + HPF. Maybe some LPF treatment in the G2 patch would help. |
Right Varice. I tried with a 6db LPF at 8k, and the waves were more similar, especially by ears. And I tried with Rob Tilt Filter, and it's fine too, maybe the best result for my ears... The oscilloscope shows a little deformation at the beginning of the wave (that tends to sinus), but it's more visual than audible.
And I also adjusted a little bit the HPF to conserve the fundamental, as tyou said.
I'm really happy with the result. Thank you for the advices... and for that fantastic patch. |
|
Back to top
|
|
|
xav
Joined: Mar 21, 2005 Posts: 164 Location: paris
Audio files: 8
G2 patch files: 7
|
Posted: Sun Feb 12, 2012 2:50 pm Post subject:
|
|
|
varice wrote: | In fact, the whole reason that we are going through this G2 HPF exercise may be because the ATC-1 apparently has a smaller value DC blocking cap than the G2, which is causing the ATC-1 square wave to exhibit the higher HPF cutoff shape that it has.
|
So the DC blocking cap of every preamp is a HPF? It explains those differences between the pure square of a software directly analyzed, and the output of a hardware synth. |
|
Back to top
|
|
|
varice
Joined: Dec 29, 2004 Posts: 961 Location: Northeastern shore of Toledo Bend
Audio files: 29
G2 patch files: 54
|
Posted: Sun Feb 12, 2012 6:10 pm Post subject:
|
|
|
Yes, that is true. The resulting HPF cutoff frequency will depend on the value of the cap and the impedance/resistive loads of the output and input circuits. Many circuit designers will be careful to select values that will keep the cutoff frequency low enough so that the bass response will not be affected too much. The G2 apparently has a relatively large cap. I get plenty of bass and subsonic energy out of it. _________________ varice |
|
Back to top
|
|
|
|