Technics WSA1 :: A minor observation / question about Drivers
Skip to toolbar


NEW MEMBERS - REGISTER HERE


Membership is FREE!
It is best to log in before posting. Anonymous posting is disabled.

A minor observation / question about Drivers

Hi all, Firstly, biased as I may be being a WSA1 owner, I am most grateful for the presence of this site/forum. I hope that those in charge are not disheartened by the level of activity from members, however. It is of course a very very niche area. …

Viewing 1 to 18 (18 Total)
A minor observation / question about Drivers

DBGalaxy5

DBGalaxy5
Australia

Total Posts: 12
Joined: August 10, 2015

Hi all,

Firstly, biased as I may be being a WSA1 owner, I am most grateful for the presence of this site/forum.

I hope that those in charge are not disheartened by the level of activity from members, however. It is of course a very very niche area. On top of which. I guess there are those of us that just arent the posting type (such as myself......... I dont tweet, and am a fairy infrequent facebook user). I am an almost constant computer user however, but more a reader and researcher than contributor.

Anyway, before I get off track too much. This is really just something that cropped up last night, and though it isnt very important I thought "why not make use of the forum?"

I was doing some back to basics programming and following a tutorial that used a modern virtual analog synth. Overall I was pretty pleased with the similarity of results I achieved. Being a bit 'pernickety' though (some say perfectionist.. but I dont like to use that term as it implies that one would often produce things to a high standard, which sadly I do not), I eventually started to wonder why I couldn't get the EXACT same sound. Of course, my more logical side realises that there are many many reasons why this likely wouldnt be possible.. but still, it didnt stop me wondering.

So I went back to the beginning of the tutorial and re-compared the initial basic square wave. Hmmm. The WSA1 square wave seems, for want of a better description, slightly 'softer'. As if someone has lightly sanded the corners of the square with a fine wet and dry. I guess this is in keeping with the nature of comments made about the WSA over the years. Also I realised that these are sampled squares (it is an assumption that the VA is generating waveforms). But really, if you are disabling all filters, effects, DSP, flattening the global EQ and using JUST the square, why shouldnt all synths sound the same at that particular point.

As I say, I realise that 'all synths have their own character', plus youve got your analog circuitry, your digital circuitry, sample compression, bit rates and depths etc etc etc.

But isnt a square a square?

Posted on October 14, 2015 at 4:05 AM
Display Messages: Threaded     Flat
17 Replies

admin

admin
United Kingdom

Total Posts: 1102
Joined: February 2, 2014

Re: A minor observation / question about Drivers

Hi DBGalaxy5,

It's great to see a post in here smiling I'm going to answer your post in reverse order...

I'm certainly not a perfectionist. Do I wish I were? Probably not smiling And I'm an absolute beginner with synthesizers. Ok, a toddler perhaps, who is learning a lot from the articles that Marc is working on with me (the Articles section of the website will have some updates in the next few days). So I can only approach your question from an 'intuitive' position (and I think that synthesizers are magical and mysterious but shouldn't be elitist).

I would say that yes, a square is a square but I can barely begin to imagine how many variations of square that are possible (and that is from looking at the usual illustration of a square wave).

My musical background, for what it is worth, is almost exclusively from playing Technics instruments (a few other makes along the way) and the topic of your post fascinates me because when I think back to my first endeavours on keyboards I remember how accurate I thought they sounded at the time. I remember playing an early Technics keyboard in the music shop and being astonished at how the piano sounded.

Looking back now, it sounded nothing like a piano! Did we used to have more imagination? If it was next to a piano, would my young mind (at 13) have discerned the huge difference?

Moving slightly forward, Johnny Cash made use of a Technics K450 on one of his recordings, I forget which one, and I read a comment somewhere that he found this keyboard to be 'state of the art'. This keyboard has no touch sensitivity, a tiny amount of sound memory and so little PCM that I expect the inventors would laugh today if they are still around. Of course, it was perhaps still a beautiful sound and I expect that many people listen to that recording today and enjoy it (wish I could remember the recording).

With virtual instruments today, we (might) believe that we are achieving sounds that are virtually imperceptibly different from the originals. I often wonder whether we will look back in days to come and feel the same about these sounds. Perhaps not. They really do sound so very accurate (should I really write that down to be re-read in a couple of decades?).

Back to a point I was trying to make. From the inception of the KN range of Technics keyboards, let's say KN800, which made a huge impact on the market, through to the last keyboards they made (KN7000, KN2600, PR804), I believe that the same samples were utilised. Recorded on high resolution equipment that was above the quality of instruments that we use even now. WSA1 came along between KN2000 and KN3000 in terms of technology. From what I understand, taking a Steinway piano for example (Technics famously used Steinways) it is the same source in an early KN keyboard as in a later one. Placing WSA1 somewhere in the middle of the lineup I can assure you that the KN800, WSA1 and KN7000 sound like the same piano sample (I own all three) but by no means do they sound the same.

KN800 sounds very "soft and sanded" to use your very appropriate analogy. WSA1 is cleaner. KN7000 is an extremely good piano. By today's virtual instrument standards, I can not say that they are as accurate as a virtual instrument or a brand new alternative brand of (high-end) keyboard. But they are all delightful and usable in their own way. And very immediate (which is a word I love to use in connection with hardware keyboards). My son loves his virtual instruments on his Mac but I just press the button and start playing in three seconds, I don't have the patience or desire to boot the computer and set everything up for ten minutes, generally, everything has its place of course but when I want to make music for myself or to an audience (as opposed to a recording environment), it is a fairly spontaneous need that requires immediacy.

I'm really digressing here so to be more brief, I suppose that the resolution of the WSA1's square is not so high as the KN7000 (well, I know it's not) so that will have an influence on 'blunting' the result as will internal amplification and the speaker system. The original source too of course.

I wonder whether you could take the WSA1 square and work on it to make it more similar to your VA's square BEFORE you do all the other editing and whether that would then come closer to what the VA is producing. However, I also wonder whether you could start on the VA and arrive at what the WSA1 is producing. Which is more a matter of which end result you prefer isn't it?

Ok. No answers from me in all of that but some food for thought and hopefully some other more proficient 'synthesizers' than me will join in (Question: Do we call synthesizer users Synthesizers?).

Regarding the beginning of your post. I've established several Technics keyboard websites all of which have a unified membership. Because I hope that all keyboard players have similar and interconnected interests but also because I'm only one person and the administration of several websites (which have about 3000 members, thousands of pages and hundreds of thousands of downloadable files) is slightly too much for me on top of my dayjob, so I have to find efficiencies grin

I'm never disheartened by the relatively small numbers of posts because I can see the statistics. Thanks to the wonders of Google and various webserver services I know that at the moment there are around 30,000 unique visitors to the websites monthly and that people interested in WSA1 are a significant component.

Like you, I read websites more than I participate so I completely understand that we have many, many more people interested here than it seems on the surface of things. Thousands of visitors and members go by for each forum post, gleaning knowledge coming back (the stats show me that they come back) and downloading hundreds of the resources files that are available here every day (ok, downloads are in the tens for WSA1).

And this WSA1 site is my baby for a few reasons. I started it before I owned a WSA1. I read that it takes about 7 years to become an expert in something, whatever that means, and at 48 I have a plan to become a very good player by the time I'm 55 grin I'm not a bad player now but I want to be able to read music to a very accurate level and be more of a 'pianist'. Along with that I want to become very good at using my WSA1. These should be achievable objectives. So, that is one of the things that i get from these sites for myself.

As I say, I don't get very involved in other websites on a regular basis as a participant, so I understand what you mean, everyone is busy these days. And it is rare that I make a long post even on my own websites. I spend a lot of time writing them already and I put the resource here for other people to use rather than for me to write everything. That makes me even more appreciative when someone takes the time to post as you have.

Of course there are other spinoffs. Every post helps us to be visible in Google and attracts other WSA1 visitors and Members. And all of that feeds back into making the website current and interesting as well as ensuring that it continues to exist (I can assure you that the sites are here to stay for a very long time).

Which is also why I have the 'Verified WSA1 Member' status. You will notice that you have a new badge against your profile. And that gives you access to my high resolution manuals (and more future things to come). Regular Members can access lots of resources here and verified members can access even more. My high-res Manuals are an encouragement for those quiet users such as you and me to join in sometimes smiling The manuals constitute about 30 hours of my own work up to now, so you've given something to everyone and I've given something to you. A very good deal for all of us grin

http://wsa1.technicskeyboards.com/wsa1-downloads/wsa1-manuals/...

I hope that wasn't too rambling!

Best wishes,
Mike

Posted on October 14, 2015 at 6:46 PM

DBGalaxy5

DBGalaxy5
Australia

Total Posts: 12
Joined: August 10, 2015

Re: A minor observation / question about Drivers

Mike,

Your response is much appreciated.

I should say that the programming excersize that I undertook had no real purpose other than to initiate a bit of a return to programming, which is something I havent done in years. However I felt that having previously joined the forum, it would be a good place to discuss the thoughts that arose.

I have to confess that shortly after posting the question I had a little look around the net and established various bits of information that also contributed to the answer I was looking for. Most notable was the manner in which square waves are 'constructed' in some synths from a sine wave and fundamentals. In essence meaning they could never be true squares in the first place. Another site showed several sawtooths (they didnt have examples of squares) generated by various synths, and it was quite amazing how different they were.

More annoyingly I began to recall whilst looking at that information, that I had historically been aware of much of it (the memory isnt what it used to be...... quite the opposite to most keyboards, which nowadays have much improved memory!)

This does not diminish the value of your reply and inight's though. In fact you raise some thought provoking points.

The comparison of the 'quality' of virtual instruments (I probably should say simulated synthesised instruments, as I dont mean software per se) is one that has always had me tied up in knots. As stated before, I can be quite pernickety, and despite being a poor musician (and that's hyping myself up!!!) I often find myself striving for better and better sonic reproduction whether that be through better interfaces, cables, mic's , effects etc. But this half of my musical split personality is in stark contrast to the other, as I am certain if you provided some excellent musicians with the likes of a 1970's casio keyboard, a set of dustbin drums, and a playstation microphone, they would still produce better "music" than I would ever dream of with the best gear in the world..... So why does the first half of me STILL pay attention to how good individual sounds are ?

To my mind, the issue with reproducing known instruments is twofold.

Firstly, if a sound (piece of music) is played and it gives a strong impression of what it "should" sound like... but doesn't, then it sounds bad. That's probably unclear... sorry. To take extreme examples. If you were at a great cathedral for a gentry wedding, and the wedding march fired up on a 1980's Bontempi organ, it would sound just plain wrong. In a slightly less extreme example, if you are playing a cover song in a manner that suggests you are trying to replicate the original, then again any 'timbres' that do not sound like the original will stand out.

The second issue is that distinct 'impersonations' become identifiable, and in turn become associated with certain era's, and ultimately 'dated' (Of course, like any fashion, some of these become 'trendy' again). This is perhaps more true with synthesized instruments than any other (DX7's, Simmons drums etc).

Yet, despite these issues, I believe original compositions have more leeway. I am not suggesting that if I composed a song using a period drum machine, DX7, sax solo, and a two girl backing that it wouldn't sound entirely retro. But I do feel that the use of elements of these - or any era associated sounds - should be quite reasonable considerations as and when they suit the piece. I know that some people would not agree with this.. but I believe a good song is a good song, and should not be dictated by faddishness.(though, if you are talking commercial considerations, then perhaps this doesnt hold up so well.) I mean, if you took some of your favourite 80's songs and re recorded them today as if they never previously existed, would they really be better if you changed all the sounds and instrumentation to it's modern equivalent. Maybe in some cases.. but certainly not in all cases.

OK.. so how can I bring my waffle back to relevance.

Well, you are going to look for a downside, the WSA is clearly a digital synth and to that end may not be suitable or desirable to many because of that and may sound 'dated' to some for the same reason. But I think that this is selling it short. It certainly doesnt suffer from having an overheard sound. Also, if my memory serves (as back in my early days of ownership, I did dabble in a bit of programming) it can get somewhat grungy as well as downright weird (though I am often disappointed that it is only this 'weird' aspect that many seem to appreciate). Sadly I overwrote my own sounds without saving them after checking out if my old factory floppies still worked.

Although my own experience is limited, to this day I feel the WSA is very versatile though. As the original point skirted around, I do feel you can probably achieve a fair approximation of a very wide range of sounds (in fact perhaps I should be open to a challenge on that front). It may not have all the knobs and sliders of a VA, but the controllers are pretty flexible (and there are enough tweakable parameters to justify an external set of midi knobs, if thats your bag).

I doubt I would ever sell my WSA (I have had it since new, and have all the original documentation, and sound disks.. even my receipt). What would you replace it with? There may be many things better at doing specific things, but where am I going to get the flexibility I get with this for anything like the price. (OK... I have always coveted an EX5...... and whilst not completely out there price wise - in Australia at least - they still arent cheap cheap.)

One final digression if I may regarding synth sounds. And that is regarding presets. Personally, whilst I may have physically 'used' presets more often than anything by virtue of ease or time, I have never consider a synth as a preset device. In the days when I 'tried' to compose more, I always ultimately created my own sounds. And nowadays - where I use keyboard much less - I still tend to tweak things at least a little.

And I always honestly believed everyone else was the same. I mean yes, I have known plenty of keyboard players that have used presets, but I havent known them enough to assume this was ALL they did. Fair enough if you bought a workstation or similar. But why by a device intended primarily as a synthesiser rather than a tone module, and only use presets. Recently(ish) I met a keyboard player at some jams that I go to who has quite a collection of synths, yet he ONLY uses presets. It's sadly amusing, as sometimes we 'jam' over parts of old songs, and these tend to be songs that come up every now and then. Each time he is beavering through the presets saying he cant find one that is 'quite' right. You should see his face when I propose that he modifies a tone - let alone create one from scratch !!!! I hope there are not many WSA owners that only use presets... that would be a shame.

For a less than occasional poster I feel I have used up my yearly allowance (or maybe my verbosity is part of the reason why I do not post often).

So I will thank you again for your response, and add my thanks for the access to the manuals, which is most kind even though, as already stated, I do have my originals..

Whilst I will try not to be a complete stranger, it may realistically be a while before I bore you further (which I am sure anyone who has read this far will be grateful for).

All the best, and good luck with your goals.

Posted on October 15, 2015 at 4:10 AM

DBGalaxy5

DBGalaxy5
Australia

Total Posts: 12
Joined: August 10, 2015

Re: A minor observation / question about Drivers

OK.. I rather implied a promise that I wouldn't be returning in a contributive manner for a while (not because I actively do not wish to, but because that was the most likely scenario). However, as I have been doing some further investigation regarding the original 'on topic' issue of waveforms I thought it may interest you to see my findings to date. Well.... I found it interesting at least.

(I should note that I am having trouble putting images in the post, and although I can "attach a file" I had quite a few, so they wouldn't be in context, so for now I am just going to add the most pertinent couple. If you could advise how I could otherwise insert images, I would be grateful to know)

So. I obtained a VST plugin (spellcheck to the rescue).... ' Oscilloscope' ... and looked at the basic waveforms from the WSA. I actually quickly scanned ALL of them, but looked more closely at those in Group K. (Square, sine, etc)

I looked at the sine first as I thought that would give an indication as to whether the plug in was in the ballpark of being accurate, and it looked like a perfect sine.

So I then went to the one that initially bugged me. The square.

Long story short, it looked NOTHING like a square. I mean perhaps I can see how the wave can be derived from a square, but then again one could argue - although it would be more extreme - that one could see how a sine could be derived from a square.

I tried hard to find other images of waves generated from synths online and other than many based upon home-made synths (where I guess it is quite appropriate to be analysing the waves generated) I could find little, though eventually I did find one showing three basic waves (Square, Saw, Triangle) from three mainstream synths.

Whilst none of these synths showed perfect squares either, even the most 'deformed' was much much closer to a square than that I am getting from the WSA.

This led me to question the accuracy of the plugin, despite its rendition of the sine. The WSA triangle though, looked very close to what I would expect (although still with an odd artefact that looked like the ramp up part was somehow modified to create a slightly imperfect peak. This said the triangles from the examples I found online were not all perfect either, though these imperfections seemed more symmetrical.

Finally, the saw was also way off the mark. In isolation, despite being rather unusual I was nonetheless initially 'happy' with it, and in some ways it was similar to the internet examples. But even the most similar of those suggest that the WSA is too peaky making it much more like a pulse than a saw. plus.. oddly, it ramps in the opposite direction from what appears to be the norm. This latter point acknowledges that some synths have both 'ramp up' and 'ramp down' saws (and others), but it still seems to be that the bread and butter saw would be a 'ramp up'... the opposite of the WSA. I also considered that it may be an image reversal in the oscilloscope, however it is easy to see the waveforms start from the right and move to the left, so this isn't the case as far as I am concerned.

So then what, I thought.

Well I then recalled that my Yahama sampler had some basic waveforms built in. I hadn't fired it up in years, and the encoders are shot to hell, so it was a bit of a painful process BUT, I was able to run the same waveform types from the Yamaha through the oscillator plug-in.

Given that I was actually leaning away from trusting the plug-in at this stage, I was rather blown away. The waves from the Yamaha were all but perfect. Maybe less of a surprise for the sine, as this seems to be universally consistent across synths. But the triangle could be used as a geometric drawing aid. The saw has a tiny peak (like a little resonance peak). And the square has similar, plus there is a very slight angle from the horizontal and vertical planes.... but it is nonetheless VERY square. Basically, it seemed to produce waves either equal to or better than the third example obtained from the internet... which in turn was the best of the three in that example. (Best being measured by geometric accuracy rather than sonic qualities).

(As an aside, as this level of detail really starts to go beyond my usually near-infinite curiosity, the little inconsistencies/peaks appear similar to artefacts that can be introduced by AD DA conversion and the like. I'm not 100% on that, but whilst looking for waveforms I recall noticing some discussion around that when seeing similar peaks elsewhere. I 'could' test this by simply importing the wave to the PC as a file, rather than playing the audio, to compare, but as I say, its a bit beyond what I am looking into here).

You may think then, that I must conclude that the basic waves in the WSA are rather 'off'.

Possibly, yes.

I am more inclined to think, though, that there is one of two explanations (which are more or less the same thing).

1. Despite my best efforts, I still have some aspect of the WSA processing enabled, that is interfering with outputting the audio generated by the driver and the driver alone (I can discount the audio path outside of the WSA as the sampler used the exact same path), or

2. There are 'effectors' (be they filters, eq's, effects etc) that cannot be truly bypassed in the WSA.

Certainly I don't wish to spend too much more time on this. As much as it fascinates me, it distracts from actually playing music (I wasted the best part of a decade in the past being distracted by the technical side of music making, hence why I have a room full of rack gear, pedals, recording devices, and a large collection of sound on sound, future music, and computer music magazines (albeit a very VERY small room). But I am going to see if I can track this down. If I don't post further, you can assume I didn't find anything. I am not optimistic to be honest..... but I would LOVE to be wrong.

Given that I am only attaching three pictures, they should be quite self evident. One is the internet example of three synths. Another is the square from the WSA, and the final is the rather better square from the sampler. So even if they are not reproduced in that order, they should be clear as to which is which.

As a final note, whilst a lot of the waveforms (drivers) in the WSA are, understandably, gobbledegook from a visual standpoint, some are really quite beautiful (especially some of the bell type ones).

OK.. Off for a final bit of analysis.

Posted on October 16, 2015 at 1:37 AM

DBGalaxy5

DBGalaxy5
Australia

Total Posts: 12
Joined: August 10, 2015

Re: A minor observation / question about Drivers

I am about ready to wrap this up (and sadly, contrary to my previous mail I am not returning to proffer tales of my success).

I am as convinced as someone that doesn't actually know for a fact can be, that there is 'some' processing occurring within the WSA to throw off the original waves. I still cannot believe that the sample waveform itself would be so far away from a square in the first place. Yet, I have disabled, deactivated and zero'd every parameter that I can find within the sound, the part, the system and global, and have not got any closer to a square than the image I provided in my last post.

It has crossed my mind that for one or more parameters a zero position is possibly skewed one way or another. Of course, if this applies to more than one parameter it would be nearly impossible to track down. I guess a trained audio engineer might be able to see from the waveform what type of effect or filter may take a square such that it ends up with what I am getting, and as such get closer to pinpointing a solution (if I were terming this as a 'problem', which really I am not).

Ironically, I feel I am able to get closer to a square just by putting the triangle through a different resonator (Not really 'that' close, but like a triangle chopped off halfway down). Though the digi lead seems, to my eyes to be even closer still, to the extent that I would probably call it a square if I saw just one waveform cycle, however there is a little 'movement' (sonically a good thing I guess) such that it only really repeats identically every two cycles. Its not a perfect square either, but in some ways it doesn't differ from a perfect square any more than those internet examples I provided, its just the manner in which it 'differs' is 'different', as it seems to have a little extra harmonic content. (three or four of the resonators - noting that my other tests used 'through' - also provide different, but equally 'not too far off' results with the digi lead as well.... but not so the square).

Given the certainty that I am not immortal, I really need to leave it there now. Bugs me though. Not because I am any less happy with the WSA because of these findings, but I would really love to know 'what is going on'. Shame there are none of the 'creators' around that can be asked such things.

One final side note though. I would very tentatively suggest that the modelling element of the WSA is based upon elaborate filters rather than independent algorithms. I realise that both require complex calculations (and this likely sounds like hot air coming from someone who a few days ago could recall any reason why a square wouldn't look essentially perfect in every synth), but just as a theory, what I mean is that the various resonators often seem to give results that look familiar to the results I obtained from EQ-ing and filtering. So rather than being individually constructed mathematical models for each resonator, I would guess the designers played around with existing filters and EQ's until they achieved reasonable simulations of the real instruments. The various fitting, muting etc parameters would then be assigned to the various available parameters to again, reflect the findings from the real life sounds. Just a guess.

So there we are.

As much as I am going to release my dogged grip on this issue now, I nonetheless welcome any input, insight or question anyone may have on the topic at anytime. I will say upfront though, that my responses may not be particularly prompt.

All the best.

Posted on October 16, 2015 at 3:38 AM

admin

admin
United Kingdom

Total Posts: 1102
Joined: February 2, 2014

Re: A minor observation / question about Drivers

I'm fascinated by what you've written here and the images you included.

I want to do your post justice so I need to do some research when I have time and I'll get back to you with another post.

In the meantime, I wanted to answer a couple of technical points:

Regarding putting images inline in posts - They have to be on the Internet before they can be included inside the actual post. This means that you have to make the post and attach the images, THEN right-click on the image links to copy the address of the image and THEN edit the post to insert the image where you want it.

I realise that this is long-winded. It's a limitation of the forums at the moment and I hope to address it when I get (a lot of) time spare.

The second point I wanted to make is that the Technical Guide that is available in high resolution was not distributed with the keyboard, so you probably don't have it. Some of the information is the same as in the Basic Manual but there is a lot of additional, very interesting (to me) information. So you might want to download that one.

Best wishes,
Mike

Posted on October 17, 2015 at 5:54 AM

DBGalaxy5

DBGalaxy5
Australia

Total Posts: 12
Joined: August 10, 2015

Re: A minor observation / question about Drivers

Mike,

I am grateful for your enthusiasm, and for the fact that you managed to get through my initial drivel to get to the 'on topic' posts.

To be honest, most of the 'musos' that I associate with are by way of a regular jam that I attend. I play Bass guitar as my main instrument (though I am pretty useless at that as well, so that should give an indication of my fairly non existent keyboard skills). Many, if not all, at the jam are fairly technophobic as far as music making is concerned. We have one keyboard player that attends, but even he isn't really into the nuts and bolts of things (he is the one I referred to with regard to using only presets).

Actually at the last jam I mentioned this particular matter to a couple of the attendees. I thought I spiced it up and gave only an outline, but you could see the eyes glazing over!!!

I did form a suspicion that images needed to be attached in that manner after I posted, but I had no 'immediate' way of going back to try it out. (More specifically I am sure I have a few different web presences that I can utilise, I just need to spend a few minutes to remember what and where) I am at work now, and the images are back home, however I may try and revise the previous post to include all the images I mention in the near future.

Finally, thanks for clarifying the nature of the manuals you have available. I am most curious and am off to down load the technical manual right away. Many many thanks.

No rush for any follow up... I am not going anywhere.

All the best.

Posted on October 18, 2015 at 2:22 AM

DBGalaxy5

DBGalaxy5
Australia

Total Posts: 12
Joined: August 10, 2015

Re: A minor observation / question about Drivers

Mike,

That Tech manual will definitely keep me entertained and confused for many hours to come (although most of it is certain to remain over my head).

I did notice on a quick glance through that there does appear to be filters on the outputs that seem separate to the 'effector' filters (though I reiterate it was a quick glance through),

However, a quick broad brush search such as the following:-

http://www.google.com.au/search?q=square+(%22low+cut%22+OR+%22lo+cu...

... and you can see that there are a few images on that page that closely resemble the WSA ouptut.

Still at work, which is now frustrating, as obviously I wish to both quadrouple check my WSA settings - I still 'could' have made an error*** - and see if knowing the nature of the wave deformation means it is somehow possible to compensate for it.

I also want to process a square on the computer to replicate the results for myself. I could do it here but I hate loading up non work related apps on the shop PC.

One other thing I learned (another thing I really 'should' have realised for myself, but didn't), is that the sine wave would not be deformed by filtering as it has no harmonics (ignoring matched frequencies or filter resonance)...... edit... hmmmm maybe this isn't strictly true, but I think as a generalisation for the visual representation of the sine in my example I'll let this ride because if the effect is phase shift then the 'tools' I am using would not make this immediately obvious)


***.... I really really want to get home as I am now wondering if I didnt reset the global cut after one of the many 'initialise's" that I performed... so definitely hold of any investigation at your end until I establish how much of a numpty I may have been!!!

Posted on October 18, 2015 at 4:47 AM

admin

admin
United Kingdom

Total Posts: 1102
Joined: February 2, 2014

Re: A minor observation / question about Drivers

Hmmm.... 5am here and I have my WSA1 and KN7000 set up, comparing squares smiling)))))))))

They sound remarkably similar to each other. However, like you I keep wondering whether I have turned everything else off... and of course they are probably sampled from the same original source. The KN7000 sounds very marginally brighter but it is also sampled at a higher rate.

It is interesting but I'm not getting anywhere, probably because I don't fancy installing lots of different apps and setting up my recording kit grin

I should probably go and actually play some music for a while instead but you have my mind fascinated by the subject.

Posted on October 18, 2015 at 5:17 AM

DBGalaxy5

DBGalaxy5
Australia

Total Posts: 12
Joined: August 10, 2015

Re: A minor observation / question about Drivers

5am !!!

I assume you are up early and not pre bed.

Play some music Mike.

I'd hate to find out I missed something obvious and have you waste your time on this (and really it IS a waste of time, as even if I / we establish a satisfactory conclusion, neither of us will be able to play better / run faster / live longer / make additional strides toward global peace as a result )

I would admit a curiosity to "seeing" the square from your KN7000, but 'd still like to check some stuff out first.

I head home soon, so should be able to recheck this with my finest toothcomb in a couple of hours.

Posted on October 18, 2015 at 6:14 AM

MarcBrassee

MarcBrassee


Total Posts: 8
Joined: September 17, 2015

Re: A minor observation / question about Drivers

Hi Guys,


Boy, I thought I used too many words. I'd like to throw in a few bits but let me first remark that I just scanned your previous stuff. So please forgive me if I should repeat anything you already said above.

Within that mindset these are some general remarks which might still be of use to you:

1. A square indeed is not a square in the sense that many synths actually use waveforms that are only bad approximations of a pure square. The same however also applies to all the other basic waveforms. Take a snippet of a pure synth wave, put it in a graphic audio editor like Audacity, zoom in and be surprised. Especially those hallowed analog synths produce very crooked shapes.

2. Looking at it from the other side: A(n almost) pure square wave will probably sound a lot less exiting then what you might expect from it. The “problem” with many 80ties and 90ties sampled waves was actually that they where too perfect for their won good and thus sounded rather dull

3. More important even is to be aware of what you expect from a certain waveform and why. Your auditory picture of the perfect square wave might in reality be more akin to a sort of hybrid , for instance a “sawtoothy” square wave or one with an attack peak.
In basic analog synth courses the square is actually often described as the mellower wave which one typically uses for a woodwind type sounds like a clarinet, bassoon or English horn. A sawtooth wave is advised for brighter sounds like violins and strings.

4. Do also not confuse a pure square wave with a distorted sound. Clipping might lead to squared off sounds and can thus add overtones but then the remaining, non clipped content of the waveform will still be present. The chopping then again leads to a hybrid which we perceive as a more aggressive sound.

5. More important even: To me it does not make much sense to complain about the exact shape of a single waveform in a 4 layer synth. Why not use the waveshape you want to dominate the sound you are after in the first layer but then add a bit of low level spice in the other layers at a more subconcious level? Why for instance not use that square wave and beef it up with and added less loud, sawtooth or even add some noise? Other options are to use the modeling and interaction parameters I described in my last article. Furthermore you can add “real” distortion in the effects section.

6. My last and main point. Talking about the exact nature of waveforms is great but first off all one has to use ones ear and be creative with the many possibilities one has on board. I have programmed some “analog” sounds sounds into the WSA that imitate some of the classics very convincingly. Knowing how to do that is one thing but still I use my ears more then anything else.

Yours truly,


Marc Brassé

Posted on October 18, 2015 at 9:25 PM

DBGalaxy5

DBGalaxy5
Australia

Total Posts: 12
Joined: August 10, 2015

Re: A minor observation / question about Drivers

Marc,

Nice to get a little extra input on the subject.

The lengths of my posts may well be responsible it being more practical to 'scan' the thread, but in turn, whilst interesting and valuable comments, I feel some elements were missed or misinterpreted. However, if I may firstly address the 'undertaking' I had been following up on.

So last night I checked and checked but am really convinced I have not missed any settings that I need to get as pure a driver sound as I can from the WSA. Because I was quite tired by the time I finished I left me comments until this morning (whereupon I had 1 final final check) and I stand by the waveform as per the image previously attached for the WSA Square. As before, however, I would be more than happy to be proven wrong or to have something pointed out to me that I may have missed.

I also went back to the sampler with the information established yesterday, and ran the sampler's square through some filters. It took a minute or so to arrive at the attached image. This shows three traces in white green and red (I can recall which is which, but when you see you will understand it isn't that important. One is the WSA1 square. The second is the sampler square with Hi pass filter with a mild setting and no resonance, and the last is the combined trace. As you can see all but the same. In fact I get closer, but you can no longer see the three traces.

My own conclusion then is that there is a small amount of low cut / Hi Pass filtering permanently applied (which is particularly odd given that he technical manual shows "low Pass" filters on the analog outs). Or the sample has been stored with filtering already applied.

I would still love to see the square from your KN7000 Mike. One day. Also, never before have I been bother about not having the digital out board, but now I am so curious what the results would have been through SPDIF (though I am guessing, the same).

So Marc, sorry, back to your points.

1. You are quite right. This and many other points I am re-learning (and I feel rather stupid for having forgotten). However I had acknowledged this point, as well as providing images showing this in one of my earlier posts.

2. Again correct and well stated, but again I have also acknowledged this previously. The main thrust of the thread has been (for me at least) to establish the facts of a point of curiosity, rather than to suggest the perfect waveform was desirable.

3. Good point well made.

4. Absolutely. This was actually well represented in the oscillator plug in I am using, and is evident when overdriving sounds.

5. I really must be clear here. I am 'not' complaining. Obviously I would not have been pursuing this little project had I initially established a perfect square driver, but it has been a pursuit of curiosity, not complaint, nor because of problems achieving musical objectives.

6. And again, a good point. But I think I need to re-clarify. The first post I made pertained to trying to replicate a sound precisely, which in turn led to my findings regarding the square wave. That first posts discussions around matching the tones of the tutorial may be misleading you to believe that I have since been seeking to maybe 'correct' this if possible for sonic reasons. I apologise if it appears this way, but it simply isn't the case. All I have been trying to find out since is "is there any "reason" why the square waveform driver in the WSA1 does not appear close to square. That's it. I may feel a little disappointed that I didn't find a reason, but that's really just a 'pride' issue of not justifying the time spent.. but it makes no difference to my perception of or happiness with the WSA.

Thanks again for your input.

Attachments: WSASQ_A3KSQ_HPF.jpg
Posted on October 19, 2015 at 2:40 AM

MarcBrassee

MarcBrassee


Total Posts: 8
Joined: September 17, 2015

Re: A minor observation / question about Drivers

Dear DBG,


Be assured that I mean all my remarks in a similar spirit as with which you posted them. The only real point is the last one. I also like to know what I am talking about but in the very end, when programming new sounds or making music with them I will always choose to follow my ears and let the analytical half of my brain take a more monitoring role.

From a purely technical point fo view though: The WSA is not without it's "bugs", as is no synthesizer (luckily!). Only last week I found a setting within the amplitude envelopes that was not consequent. A pair of settings that should lead to a perfectly sustained sound (I was a the moment trying to generate a test tone) still led to a slow decay with a held key. If I remember correctly the maximum decay setting did not lead to 100 % sustain, although is must be said that the maximum value is actually 99, so one might still interpret that choice as consequent. I was able to circumvent the problem it in the end but conceptually it still is a dubious choice. The point is however so mute during everyday use that only such a specific goal as generating a 0 odB A=440 test tone made it that apparent.

Many more such remarks could be made. For instance: As much as I do like the WSA, the sequencer is a buggy sack of bones if ever there was one.

About that square wave. I once read in a test that the whole sample set has actually been "prepared" to work optimally with the modeling section. Your remark that some inbuilt filtering might be present could therefore be spot on. Maybe some filtering was actually done to the sample set to avoid certain transients from aliasing effects and such. Want some proof? Give the low setting in the WSA's EQ a big boost and you'll sometimes hear some very interesting artifacts. On mine this is permanently set on a +6dB boost at 400 Hz. It really makes the WSA do some great stuff in the lows.

In the same vain that might explain why it was often complained that the WSA sound a bit thin. In typical perfectionist Japanase fashion the rude behaviour I describe above was probably frowned upon so they rather overdid the cure a bit then leave it in. I do however love this extra bit of character.

Year after its design Robert Moog analyzed the success of the typical Minimoog sound in an article in the US Keyboard magazine. His conclusion was that all the tiny mistakes in the design that he could not avoid or was unaware of at the time (harmonic distortion and stuff) actually gave the instrument it's now hallowed character.

If you are interested in comparing waveforms. I have multi-sampled all the waveform in my Yamaha CS80 into WAV's, just in case the real thing ever packs in. I can send you a few examples so that you can look on in horror how crooked the waves of another of these most revered synthesizers in analog history actually are.


Marc

Posted on October 19, 2015 at 12:12 PM

MarcBrassee

MarcBrassee


Total Posts: 8
Joined: September 17, 2015

Re: A minor observation / question about Drivers

P.S.

Since my last answer I also had a look at your JPG. Non of these waveforms are acually pure square waveforms. Ergo: Even in sample sets distorted semi-squares already seem to be the standard.

Interesting indeed!

Marc

Posted on October 19, 2015 at 8:34 PM

admin

admin
United Kingdom

Total Posts: 1102
Joined: February 2, 2014

Re: A minor observation / question about Drivers

I thought it interesting that the "Square Lead" preset on WSA1 uses JazzClarinet1 as the second Tone. On the KN7000 preset both Tones are Square Lead.

The WSA1 does suffer from some noise in the background. A 'whooshing' that stops about a second after the key is released. This also applies to Technics keyboards earlier than WSA1 but is completely eliminated in the later keyboards (which also have hugely impressive sound editing capabilities very similar to WSA1 Sound editing but less realtime controllers).

Posted on October 19, 2015 at 8:42 PM

MarcBrassee

MarcBrassee


Total Posts: 8
Joined: September 17, 2015

Re: A minor observation / question about Drivers

And no modeling options which to my taste is the WSA's ultimate strength.

I am however not aware of the wooshing you describe on any of mine. I do however rarely use the on board effects. Maybe they are to blame? Maybe mutign them helps?


Marc

Posted on October 19, 2015 at 10:38 PM

DBGalaxy5

DBGalaxy5
Australia

Total Posts: 12
Joined: August 10, 2015

Re: A minor observation / question about Drivers

Hi all,

Sorry for the late reply (though I did intimate I am not a regular poster usually, so even this delayed post is more prompt than my norm.)

Marc. With regard to your post immediately after my last. Fortunately I haven't really come across the WSA's bugs. Not really being a keyboard player per se, I guess I am less likely to notice. It's the same with the sequencer. I do recall learning to operate it when I first got the WSA, but have not had occasion to utilise it since (unless you count the very occasional midi file playback).

I certainly would be interested in looking at those CZ waveforms though, if / when you have time to send them through.

You are quite correct that none of the waveforms I posted are pure squares. I guess I was just a bit excited when looking at the sampler 'square' compared the the WSA. In that context it looked pretty perfect, but of course, in isolation it isnt. In any case, having looked into the way analog gear creates a square, one wouldnt expect a perfect square from that type of gear as, in theory, it would require an infinite number of sine waves to create.

Whilst I am 'over' the original frustration, I may ask a favour of someone. If I saved my unaffected 'square' user patch and posted it here, perhaps someone could have a really quick look to see if I did miss anything (i.e. whether there is still some 'effector' in place that I have missed.) Obviously your global effects would still be in place if memory serves, so I guess you'd only be able to check tone specific parameters.

One other final final final thing I might look at is the sampler square imported into a waveform editor rather than analysed via the audio output. If it isnt perfect in that case, then I could create a perfect square and see if it plays back perfectly (I am guessing that it wouldnt).

Mike. I am once again at work, but I do not recall seeing a 'square lead' driver....... Oh. Hang on. Sorry, I guess you are referring to a combi patch made from 'square lead' individual patches. I must have a look at the drivers for those when I get home. I am sure you know, but my own reference to 'digi lead' was regarding the drivers rather than tones.

And I am with Marc regarding the 'wooshing'. I don't recall this myself (another thing to check when I get home). I originally used to monitor on some PA speakers (which I still use occasionally). I then used some Roger's DB101's which I really love even though it's not a perfect application for them. However I very recently purchased some JBL LSR305's and have only used these with the WSA in my recent programming experiments. It will be interesting to look out for this noise artefact playing some standard patches through the new monitors. They seem to have great clarity so I 'should' pick up on anything particularly notable. Having said all that, perhaps headphones would be better still??)

Marc. I noticed the other day via a link you posted on another thread, that you are the owner of the 'starship' set ups. I remember coming across those a number of years ago, and a few times since. Great set up, and nice to be able to chat to the proud owner.

Thanks guys.

Posted on October 22, 2015 at 6:50 AM

MarcBrassee

MarcBrassee


Total Posts: 8
Joined: September 17, 2015

Re: A minor observation / question about Drivers

“Marc. With regard to your post immediately after my last. Fortunately I haven't really come across the WSA's bugs. Not really being a keyboard player per se, I guess I am less likely to notice. It's the same with the sequencer. I do recall learning to operate it when I first got the WSA, but have not had occasion to utilize it since (unless you count the very occasional midi file playback).”

What's in a name. But that sequencer really is buggy. I produced the full length of Zamisdat part 2 on it and it drove me mad. The typical thing is that the interface seems well thought out at first glance but as soon as you start to do some real multitracking on it ti drives you berserk. So in practice it is little more then a sketchpad.


“certainly would be interested in looking at those CZ waveforms though, if / when you have time to send them through.”

CZ eh, I only know a Czechoslovakian Motorcycle brand named CZ and I think Casio once had CZ synthesizers. Hardihar. Couldn't let hat one slip. But seriously. Can I just attach WAV's to these messages, Mike?


Whilst I am 'over' the original frustration, I may ask a favor of someone. If I saved my unaffected 'square' user patch and posted it here, perhaps someone could have a really quick look to see if I did miss anything (i.e. whether there is still some 'effector' in place that I have missed.) Obviously your global effects would still be in place if memory serves, so I guess you'd only be able to check tone specific parameters.

Again: Looking at your JPEG I am hardly surprised. A real square wave has to look like a row of brick. Using approximations like these in stead has the advantage of added overtones. Basically it seems that what we all call a square wave is in reality normally just a halfway shape between sawtooth and square.


“One other final final final thing I might look at is the sampler square imported into a waveform editor rather than analysed via the audio output. If it isnt perfect in that case, then I could create a perfect square and see if it plays back perfectly (I am guessing that it wouldnt).”

As you have already proven the results would only be different on a microscopic level, Think about it: Even an old sampling system like the WSA's (I actually expect an 8 bit system because of the relative graininess of many sounds) will still have a sampling frequency of about 30 kHz to get anything near 15 kHz in its output signal (ask Mr. Nyquist), If you play an A = 440 Hz your resolution will be 30.000 divided by 440 = 68, 18 samples per single waveform on the X axis. And that is for an A above middle C. With every octave down the “resolution” per single waveform doubles. You will have to zoom in extremely to see any differences that are appreciable.

So again: Crooked, so called square waves are just an integral part of synthesis lore. Live with it! smiling


“Marc. I noticed the other day via a link you posted on another thread, that you are the owner of the 'starship' set ups. I remember coming across those a number of years ago, and a few times since. Great set up, and nice to be able to chat to the proud owner.”

I've done more impressive things. Basically it is just a custom case filled with off the shelve equipment. Only the MIDIMIXFIX is rather unique because I personally inspired it's creator to write the poly aftertouch to multichannel software. What I cannot believe is that people are not buying the darn thing in droves since then. It makes every multitimbral synth listen to poly aftertouch via MIDI!

Posted on October 23, 2015 at 9:39 PM