sample rate ? 44.1 ?88.2?

[David Nahmani]

Funny, I just tried a test playing various instruments at various sample rate. I just have Logic play a MIDI region as I change the sample rate (takes about 2 sec before the sound comes back).

 

At first I couldn't hear much difference, until I opened Ultrabeat, used the default factory settings, and played the first default sequence in the step sequencer (you have to hit play again every time you change the sample rate). Listen to the hi-hat: it definitely sounds much better at 44.1KHz than any other sample rate!

 

I'm definitely not drawing conclusion from this rapid fire, 2mn test, but it's interesting.

[Jope]

Here's another example: Choose the ES1 with its default sound and set the octave to 2'. Then play some notes in the range above C4. You will definitely hear some oscillations that are no harmonics of the sawtooth, they are even lower. Now change the sample rate. The interferences are more decent at higher sample rates (they will not disappear completely at any sample rate, though). I just tried it out. The effect is strongest when the filter is completely open. Edited April 26, 2007 by Jope

[nikkik]

Certain (effect) DSP processes are actually better performed at lower rates, others at higher.

 

I wont even get into serous discussion on this subject, as people have pre-formed beliefs that prohibit substitution of true fact and Law for the misinformation that has been digested. Once beyond fact, there simply resides subjective opinion. The unfortunate part is that the misinformation often influences this subjective conclusion. Had the person been properly educated, rather than fed information fuelled by false/improper analogies, their subsequent decisions/findings/opinions might differ. Equally *disturbing* are those who, when properly educated, disallow for the subjective side to be contradictory to what may seem an obvious conclusion based upon those facts and Laws, as the final conclusion is one of perception, which is singular and unique. After all, while Law will remain constant and unyielding, implementation itself alows for flaw; such flaw being "beauty or beast" in the eye of the beholder thus provides for the good old, "But, even though the Laws say this, when I use product 'Super-Duper,' I KNOW I am preferring this setting, even tho Law says I should not hear a difference from this lower setting."

 

Being educated allows one to understand (better) WHY their opinion might be what it is, and what the contributing factors *probably* are, rather than drawing conclusions based upon false logic (which can create a most vicious circle of nastiness). How about PCM vs DSD? Hehehehe... "How come it's only 1-bit? Oh- I heard it was 8-bit...don't we need at least 16 bits? And why such a high rate?" lol...

 

If you have free time (say, a day or two...or ten... ), check this thread for a decent debate and a slew of info, with the participants being tops in the active field- Massenburg, Aldrich, Lavry and a host of others show up. At 49 pages, it is one of the longest threads discussing these concepts/opinions I have ever seen online.

 

Thankfully, Mozart did not care what was used when in the recording studio...he just wrote incredible music...well, some of us consider it incredible

Edited April 24, 2007 by nikkik

[David Nahmani]

Here's another example: Choose the ES1 with its default sound and set the octave to 2'. Then play some notes in the range above C4. You will definitely hear some oscillations that are no harmonics of the sawtooth, they are even lower. Now change the sample rate. The interferences are more decent at higher sample rates (they will not disappear completey at any sample rate, though). I just tried it out. The effect is strongest when the filter is completely open.

Wow, thanks. Huge difference. The 44.1KHz almost sounds like it has a chorus on it, while the 192KHz sounds totally dry, pure digital sawtooth (I tried setting the mix slider all the way to the top osc, setting the octave on 8', 4' or 2' all produce very different results at different sample rates)!

[Jope]

Certain (effect) DSP processes are actually better performed at lower rates, others at higher.

 

I wont even get into serous discussion on this subject, as people have pre-formed beliefs that prohibit substitution of true fact and Law for the misinformation that has been digested. Once beyond fact, there simply resides subjective opinion.

 

Sure it does. But my ES1 example can easily be reproduced by any Logic user, and the effect is far beyond any mist of belief. Oh, I don't want to seem fanatic about it... But really, try!

[David Nahmani]

Certain (effect) DSP processes are actually better performed at lower rates, others at higher.

 

I guess both Jope and I just proved exactly that! In my example, Ultrabeat sounded much better at 44.1, while one could argue that in the ES1 test, 192KHz sounds better. And as Jope said, it's an OBVIOUS difference.

 

Believe me, when I started my original little test, I was really, really skeptic I would hear any difference at all. This is a revelation for me.

[Jope]

This is a revelation for me.

 

I am happy you still can be surprised...

[David Nahmani]

Oh with Logic, I will still be surprised and learn new tricks 25 years from now!

 

The more I learn the more I realize the amount of stuff I DON'T know.

[BobTheDog]

Certain (effect) DSP processes are actually better performed at lower rates, others at higher.

 

I wont even get into serous discussion on this subject, as people have pre-formed beliefs that prohibit substitution of true fact and Law for the misinformation that has been digested. Once beyond fact, there simply resides subjective opinion. The unfortunate part is that the misinformation often influences this subjective conclusion. Had the person been properly educated, rather than fed information fuelled by false/improper analogies, their subsequent decisions/findings/opinions might differ. Equally *disturbing* are those who, when properly educated, disallow for the subjective side to be contradictory to what may seem an obvious conclusion based upon those facts and Laws, as the final conclusion is one of perception, which is singular and unique. After all, while Law will remain constant and unyielding, implementation itself alows for flaw; such flaw being "beauty or beast" in the eye of the beholder thus provides for the good old, "But, even though the Laws say this, when I use product 'Super-Duper,' I KNOW I am preferring this setting, even tho Law says I should not hear a difference from this lower setting."

 

Being educated allows one to understand (better) WHY their opinion might be what it is, and what the contributing factors *probably* are, rather than drawing conclusions based upon false logic (which can create a most vicious circle of nastiness). How about PCM vs DSD? Hehehehe... "How come it's only 1-bit? Oh- I heard it was 8-bit...don't we need at least 16 bits? And why such a high rate?" lol...

 

If you have free time (say, a day or two...or ten... ), check this thread for a decent debate and a slew of info, with the participants being tops in the active field- Massenburg, Aldrich, Lavry and a host of others show up. At 49 pages, it is one of the longest threads discussing these concepts/opinions I have ever seen online.

 

Thankfully, Mozart did not care what was used when in the recording studio...he just wrote incredible music...well, some of us consider it incredible

 

I would have thought that any operation that affected the harmonics of a signal will produce aliasing affects, it is fairly obvious that increasing the sample rate lowers audible aliasing, is my education wrong?

 

Cheers

 

Andy

[David Nahmani]

Interesting to note that when I switch from 96KHz to 192KHz in Logic, my interface (Metric Halo ULN 2), which only supports 96KHz, stays at 96Khz. I still hear a major difference in the sound.

 

That leads me to think that the difference in sound has probably nothing or little to do with the sample rate of the D/A converters, but more with the sample rate conversion that Logic does on the fly.

[nikkik]

@David: Bingo!

 

@Bob: I said I wouldn't get into serious discussion on this...hehehe...I would rather not ever argue Laws n facts re: sample rate with PCM again, if possible; IMO, implementation simply allows that "YMMV" will apply.

 

...But, to answer your question: various methods. It depends upon how the coder writes, and then implements the tools available. FFT/iFFT to wavetable types, for instance. As a generalism tho, the plug-in coder *should* account for this type of behaviour, if it were possible to occur. I have a friend who codes for a certain other app I wont mention (hehehe...), and if you would like a detailed, technical explanation, I am sure he would be happy to pass the info on, as would I.

 

@virtual instruments re: sample rate of session...This will vary.

Sample based instruments, such as BFD for instance, might use samples originally done at 44.1k. Thus, using the plug-in at 96k would require upsampling at some point, which *would/could* be degrading to the quality. For instruments that do not use samples, the sample rate of the session/project should never make a difference in the output of the synth internally, unless the coder(s) intended it as such (for whatever reason...). Notice i said "should." As an example: if you produce a (pure) waveform at 3k with a synth, it will be the same regardless of session/project sample rate. What you hear at each sample rate will probably vary, due to the hardware involved. If the sound produced internally by the synth is different, then the coder(s) is(are) the one(s) responsible. Think of it this way: If you take an old man with a hearing range of 40Hz-10kHz, and then a young girl with a range of 24Hz to 19kHz and play the same square wave of 10kHz, the old man might struggle to hear it, and it might be "dull" sounding...while the girl will hear it as perfectly as she will ever be capable of hearing it...and years later, she might be faced with a similar sound as that old man experienced. Regardless, the synth itself is STILL producing an identical *sound* each time. Substitute "DAC" for "Ears" and you have one of the vital components in any experiement you might attempt.

[fader8]

I would have thought that any operation that affected the harmonics of a signal will produce aliasing affects, it is fairly obvious that increasing the sample rate lowers audible aliasing, is my education wrong?

 

No, I think that's a generally accepted notion. I think it accounts for most of the difference we actually hear. But I don't think you have to get much higher than 44.1 for it to be noticable. Back when I worked with 16 bit MDM's and analog mixers, I always noticed a quality difference between 44.1 and 48. I still hear it now albeit not as much.

 

Going from 48 to 96k, with 24 bit, I don't think I hear much difference if any at all. Except! Plug-ins. To be sure, some plugs sound way better at 96k. In particular the UAD stuff. I'm not alone in that perception either. Then again, there could be some plugs out there that sound worse at 96.

 

Then again, my old departed Akai S950 played some 22k 12 bit snare samples that were just KILLER. I'm sure Mozart had something to do with that.

[nickt66]

Yeah, there's some interesting things going on with the soft synths when you change sample rate. Check out Sculpture's 'Charming Flute' preset at 44.1 and then at 192. It kinda loses its charm

[BobTheDog]

@David: Bingo!

 

@Bob: I said I wouldn't get into serious discussion on this...hehehe...I would rather not ever argue Laws n facts re: sample rate with PCM again, if possible; IMO, implementation simply allows that "YMMV" will apply.

 

...But, to answer your question: various methods. It depends upon how the coder writes, and then implements the tools available. FFT/iFFT to wavetable types, for instance. As a generalism tho, the plug-in coder *should* account for this type of behaviour, if it were possible to occur. I have a friend who codes for a certain other app I wont mention (hehehe...), and if you would like a detailed, technical explanation, I am sure he would be happy to pass the info on, as would I.

 

@virtual instruments re: sample rate of session...This will vary.

Sample based instruments, such as BFD for instance, might use samples originally done at 44.1k. Thus, using the plug-in at 96k would require upsampling at some point, which *would/could* be degrading to the quality. For instruments that do not use samples, the sample rate of the session/project should never make a difference in the output of the synth internally, unless the coder(s) intended it as such (for whatever reason...). Notice i said "should." As an example: if you produce a (pure) waveform at 3k with a synth, it will be the same regardless of session/project sample rate. What you hear at each sample rate will probably vary, due to the hardware involved. If the sound produced internally by the synth is different, then the coder(s) is(are) the one(s) responsible. Think of it this way: If you take an old man with a hearing range of 40Hz-10kHz, and then a young girl with a range of 24Hz to 19kHz and play the same square wave of 10kHz, the old man might struggle to hear it, and it might be "dull" sounding...while the girl will hear it as perfectly as she will ever be capable of hearing it...and years later, she might be faced with a similar sound as that old man experienced. Regardless, the synth itself is STILL producing an identical *sound* each time. Substitute "DAC" for "Ears" and you have one of the vital components in any experiement you might attempt.

 

Sorry i was no talking about D/A or ears, I was talking about the internal sample rate that the plugin was using and how it will affect the final sound of the algorithms.

 

Lets take a simple example of a FM modulated triangle wave through a comb filter, the sample rate used by the algorithm for FM modulation directly affects the next stage the comb filter.

 

Now plugins may over-sample/filter/anti-alias (whatever you want to call it) to effectively raise the plugins internal sample rate but you cannot get away from the fact that if you increase the sample rate you decrease affects of aliasing at higher frequency.

 

When we convert this back to audio at the end of the chain I agree that 48K is good enough for all humans hearing and 44.1K good enough for nearly everyone.

 

Would your friend think this is wrong?

 

Cheers

 

Andy

[Jope]

Interesting to note that when I switch from 96KHz to 192KHz in Logic, my interface (Metric Halo ULN 2), which only supports 96KHz, stays at 96Khz. I still hear a major difference in the sound.

 

That leads me to think that the difference in sound has probably nothing or little to do with the sample rate of the D/A converters, but more with the sample rate conversion that Logic does on the fly.

 

You can turn it as you like - all problems and interferences occur where frequencies do not fit into the grid of a given sample rate. In such a case high frequencies (also edges in waveforms which means strong overtones) can cause audible lower frequencies to occur, and this usually is annoying to our ears.

When you record natural sounds, the filters in your interface will round the edges and thus cut overtones that might cause interferences before the waveforms are sliced by the sampling process.

One more summary:

- First case: Recording of natural sound with sample rates as low as 44.1kHz will just cut frequencies/harmonics that are beyond the audible spectrum, and the effect of this has much to do with belief. This is my point of view, leave me there.

- Second case: Generating sounds by algorithms that are in some way stuck to certain time grids, and sample rate conversion, i.e. transferring a waveform from one grid to another grid introduces more or less disturbing interferences.

Keep im mind the ES1 sound in the example is very immoderate since a plain sawtooth has not much musical value and the interferences will usually be buried by other notes.

[blueintheface]

nikkik, thanks for your input - and your wit

 

I have a friend who codes for a certain other app I wont mention (hehehe...), and if you would like a detailed, technical explanation, I am sure he would be happy to pass the info on, as would I.

 

I think that would be tremendously interesting - if it's not too much trouble - it might also be interesting to see the extent to which such information is 'interpretted'. .

 

. . . . you cannot get away from the fact that if you increase the sample rate you decrease affects of aliasing at higher frequency.

 

BobTheDog - why do you think aliasing is inevitable, and what do you suppose might be the cause?

[nikkik]

OK..I will try to get a few to compose something to explain it all in understandable terms.

 

Bottom line: yes, different sample rates CAN have an effect on how a plug-in sounds. It is not consistent though, as every coder or team of coders can have different approaches, use different algorithms, etc. If a coder were to approach things by upsampling, applying processes, and then bandpassing the result according to session rate, you might end up with a synth that sounds the same internally at any rate, and thus DAC would be the only variant.

 

This is actually an excellent point IMO, in that it can be beneficial to try your virtual instruments at different internal *song sample rates*. Sample based, and certain wavetable based, would be exempt, and would be best run at their native sample rate, obviously.

 

Personally, I stick to 44,1k/24 bit for most of my recording/composing, but am quite tempted to switch to 48k, which I have tried to use more, but since I have sample based instruments running 44.1k samples, I have not been as comfy when doing so all the time.

 

I look at it this way: if I find a synth sounds different at a higher rate, why would I assume it to be *better*, or even a difference I would appreciate enough to sacrifice in other areas just to have this one particular sound? I would first try to reproduce it thru tweaking at my preferred rate...or, I would simply dismiss that "different sound" as yet one more "piece of clay" I choose not to even try to sculpt with, just as I avoid trying every single virtual, and hardware, instrument I hear about. Or, render at that rate and import the rendered audio.

 

It is an intersting phenomenom, it can be explained (maybe not easily), and is completely, 100% subjective in qualitive appreciation. If it came down to an argument of, "Well, at 96k session rate, I can use synth YYY to produce an even more accurate and proper sawtooth!", I would have to walk away, and shrug my shoulders. True sawtooth is difficult, digital or analogue, and why would I use a pure, unfiltered sawtooth? In the end, much as Mr Reznor might see it- I just want to find a new way to mangle, warp, cajole or coerce it into something that is close to a sound I hear...or never have before. Sample rate does not dictate this for me, especially when I am always working with at least 44.1k.

 

As I said- it is an interesting observation, one that has been discussed at length elsewhere many times, and one that can be fascinating, or boring. I find the whole universe, and what might be beyond...all of this...very fascinating. Oh- and Logic. LOL!!! Very cool to see this observed and discussed!

[tiny333]

all very intresting ....mmmm

 

the pro sound thread is the bomb kinda says it all dont it...

 

intresting the way pluggins are affected....'

 

not that i use them...

 

just intrested in the audio

 

:}

 

gonna stick at 88.2 myself just because

 

thanks gentlemen

[Jope]

BobTheDog - why do you think aliasing is inevitable, and what do you suppose might be the cause?

 

Although I'm not Bob and at the risk of boring you, here is just another perspective on aliasing.

If a digital or digitalized tone does not fit into the sample grid (as the period of a 1kHz oscillation is not a multiple of the time of one sample when using a 44.1kHz sample rate), every cycle of the digital waveform has to look somewhat different from each other, i.e. the "steps of the staircase" have to be distributed differently on every cycle. Your output Interface will round the steps, and the challange is to put the steps right prior to analog conversion to make the cycles look equal after conversion and filtering.

And this is a huge challange as the algorithm of the digital oscillator or whatever has to anticipate what the output filter will do. No matter how accurate the algorithm works - it can always be just an approach.

[nikkik]

@jope:

 

I am not attacking, so please do not take this as such...please? I am curious: are you aware of oversampling in converters?

 

The actual sample rate that is initially used when receiving analogue input is not (or rarely) the rate at which it will (eventually) be sent to the digital out (and then to the host, processor, DSP chip, etc, whatever). The...ermmm....anomalies? that one would/could associate with implementation of *raw* Nyquist/Shannon upon non-pure sinusoidal/non-complex sinusoidal when implementing this in analogue/digital conversion can be *mostly* overcome by doing this. A common reference/curiosity is presented when considering something such as the sawtooth, or triangle, waveform type. Since one would need an (almost) infinite rate times two to properly "sample" this, one could then falsely conclude that the Shannon works (and subsequent works of others) would not be fully applicable to these non-sinusoidal types. But, in fact, when one incorporates the actual engineering of how we as humans interpret sound (ears->brain), it becomes a bit clearer as to why this has no bearing (or, as I and many others may prefer- little bearing, but valuable bearing nonetheless).

 

Hehehe...I said I wasnt going to get deep into this, and now look...pfft on me...

 

IMO, one of the biggest inhibitors in relating sample rate information in a manner that is *easily* understandable is that there has been so much misinformation, and too many bad analogies. Equating sample rate to stair steps, or to pixels in a picture, or frames per second in a film...these are all bad IMO. Why not liken sample rate to light spectrum, should we venture into a visual comparison/analogy? Do we need to capture those ultra-high frequency spectrae to more *properly* capture a moment in picture, static or time lapse? Are those highest radio waves necessary to capture to properly reproduce the "real world" we attempt to capture? Maybe so...but, equally, in the auditory world, do we need to capture those wavelengths that exceed our range of perception for that sense (hearing)?

 

Dwelling in this marsh of contention is simply a waste for me, as I am not interested in developing equipment for the purpose of recording/capturing moments. But, understanding just enough so that when I listen to brand X at 192k, and then brand Y at 48k, and am baffled as to why brand Y sounds better to me...clearer, deeper...it speaks ot me...well, I really enjoy having just enough info to be able to deduce that one distinct possibility is that the brand Y rig probably has better components, more R&D, more care placed in the production. Probably.

 

I feel above 96k in PCM based audio is a waste. I would rather spend the extra cash for better converters than on trying to have enough DSP to do the higher rate. I dont see the value, based on the facts I have. In practice, the facts have held up. For me. Funny thing is, I still work at 44,1k the majority of the time!

 

If I truly wanted to climb Mt Everest, I might be able to. I could then proclaim to the world of my achievement. But, would it be worth it? Nah, not for me. Having taken the time to learn, and leaving doors open to keep learning and expanding, and being satisfied with my decisions, I can now spend my time playing music, and not worrying if a new 384k sample rate 128 bit uber-verter will make my music sound better. More likely? My frustration and obsession would make it sound worse.

 

As Forrest said, "And that's all I got to say about that." (it has all been said before anyway...make music!!!)

[Jope]

@jope:

 

I am not attacking, so please do not take this as such...please? I am curious: are you aware of oversampling in converters?

 

I don't take it as an attack, and I am aware of oversampling in converters: It's just a quite successful attempt to relocate parts of the necessary filtering from analog to digital, and it always works with multiples of the destination sample rate, thus avoiding the introduction of artifacts. The trick is to set the real sample rate of the converter much higher than the destination sample rate and thus far beyond the cutoff frequency of the analog filters.

It doesn't matter for my point of differences in the occurence of disturbing frequencies when using software synthesizers with different sample rates.

Even shorter: If a software synthesizer produces side noises, using a higher sample rate is the way to go.

[blueintheface]

BobTheDog - why do you think aliasing is inevitable, and what do you suppose might be the cause?

 

Although I'm not Bob and at the risk of boring you, here is just another perspective on aliasing.

If a digital or digitalized tone does not fit into the sample grid (as the period of a 1kHz oscillation is not a multiple of the time of one sample when using a 44.1kHz sample rate), every cycle of the digital waveform has to look somewhat different from each other, i.e. the "steps of the staircase" have to be distributed differently on every cycle. Your output Interface will round the steps, and the challange is to put the steps right prior to analog conversion to make the cycles look equal after conversion and filtering.

And this is a huge challange as the algorithm of the digital oscillator or whatever has to anticipate what the output filter will do. No matter how accurate the algorithm works - it can always be just an approach.

 

The idea that the more the sample 'steps' resembles the analogue waveform is not only unhelpful, it's misleading.

 

A higher sample rates offers more 'steps' per sample and a visual representation of it appears to resemble the analogue waveform more closely, but the visual 'accuracy' is irrelevant.

 

There, I've just said the same thing twice, both times ineloquently. But it's early, and my caffeine levels are low.

 

As we approach the Nyquist frequency - for any sample rate - a frequency will be represented by as few as two steps and bear little visual resemblance to the analogue display (that's the part people seem to have trouble with) but this lack of visual correlation of the digital representation has no bearing on the (continuous, analogue) waveform that it reconstituted post D/A.

 

If all else fails, any discussion of sample rates always provides an illumination into the 'successes' of religion. . . .

Edited April 25, 2007 by blueintheface

[Jope]

I the idea that more a display of the sample 'steps' resembles the analogue waveform is not only unhelpful, it's misleading.

 

A higher sample rates offers more 'steps' per sample and a visual representation of it appears to resemble the analogue waveform more closely, but the visual 'accuracy' is irrelevant.

 

It's not. If the waveform of a constant tone varies from cycle to cycle (and I am not talking about the steps of the digital part but the filtered output of the DAC, visualized on an oscilloscope), you will hear the variations as side noise in some way.

Definitely.

I just talked about steps because they are what's in the digital process and what has to be "justified" to avoid the waveform variations after conversion and filtering.

[tiny333]

marsh of contention??!!

 

i live near cambridge so i know all about marshes!

 

[Jope]

marsh of contention??!!

 

Oops - you are right. Let's not fight about such things. I am perfectly sure most of us know about the same things and we are just argueing around some terms.

BTW, I'm afraid I have to get the work done what I'm paid for...

[tiny333]

The ideal sampling rate and for that matter sample size are points of debate among industry professionals. And, there are some inherent problems with digital audio theory:

 

1. mathematical theory and engineering practice often conflict

 

2. converter clocks are not stable

 

3. converter voltages are not linear

 

4. filters introduce phase distortion

 

Many people hear information in the 20 KHz range and beyond as "air" and as desirable. For some people, even up to age 41, the frequency range extends to over 23 KHz as it did for Rudolf Koenig. It seems strange that a new CD would have less frequency bandwidth than a phonograph record made in the 1960's or a new digital recorder have less frequency bandwidth than a 20 year-old analog recorder. Many analog systems produce frequencies beyond 25 KHz and the study of psychoacoustics has revealed that sound has effect above 22 KHz in both the physiological and subjective areas.

 

We are moving to a 96 KHz sampling rate which will greatly extend the realizable upper frequency limit. Even at 16 bit sample length this would be a significant achievement. With larger word lengths, digital audio will finally through off the limitations of technology that have characterized the first 20 years

 

 

 

 

 

sounds good !!!

 

hey anyone else remember the old 3m 16 bit 32 tracks??

now they where crusty!!

 

or how about the 1st emulators !!

 

wow we where so impressed!!

 

and a bloody fairlight was £250 000 !!!!

 

you kids dont know your born

[fader8]

We are moving to a 96 KHz sampling rate which will greatly extend the realizable upper frequency limit. Even at 16 bit sample length this would be a significant achievement. With larger word lengths, digital audio will finally through off the limitations of technology that have characterized the first 20 years

 

Now if we could have a significant breakthrough in loudspeaker technology, we'd be all set. We're long overdue.

 

hey anyone else remember the old 3m 16 bit 32 tracks??

now they where crusty!!

 

Sure, 45ips and when it clock errored it went off like a gunshot in the monitors. We did a Paul Winter album with one of these. It came with a technician. They weren't for sale, lease only items. If I recall there was only one studio in NY that actually had them and you'd rent from them. They were 48k SR. They sounded excellent.

 

Even more crusty . . . the Sony Umatic video decks with outboard PCM converters. Not a bad 2-track setup. Cheaper than a Scully, but didn't really sound as good. 1979 maybe?

[Neil Parfitt]

So.. is anyone here writing music anymore or did everyone turn into a scientist?

 

 

-Neil

[Jope]

So.. is anyone here writing music anymore or did everyone turn into a scientist?

 

As far as I am concerned, I'm everything I like to be

 

Since Wendy Carlos gave us Switched-On Bach in the early seventies, I am tinkering about both the musical and the electronical side of music... Psychologists call this imprinting in the early childhood or so...

[fader8]

So.. is anyone here writing music anymore or did everyone turn into a scientist?

 

Actually, I think "Mad Scientist" is the more appropriate terminology. However, it's more polite and respectful to use the formal "Mr." before addressing anyone in this way.