About saxophone voices quality

[BogdanH]

To be specific, it's about S.Art Saxophone (#008-032-083), which is listed first in Woodwind section.
All notes from C4-G4 contain very unpleasant/disturbing noise, which sounds like speaker membrane would be damaged. At first I though there's something "vibrating" inside keyboard (loose wire or similar), but after checking with headphones and external speakers, it's definitely "inside" voice.
Btw. the same thing can also be heard by using some of other (similar sounding) saxophone voices -which kinda confirms my theory, that the same samples are used for more than one voice only.

Yes, I'm (again) disappointed by quality of Yamaha voices.

Bogdan

[Jeff Hollande]

Hey Bogdan :

I am always removing these noices immediately when I am editing tracks in XGWorks.
Not only the Sax noises but also the guitar noises e.g. are removed by me.
It is simple and easy to get rid of them in XGW.

Wished Yamaha had upgraded XGW after the year 2002 but that will never happen.
XGW is no longer existing for Yamaha but there are still enough Yamaha customers today who are using XGW frequently.
XGWorks, the best midi editing prog ever made for ( older ? ) Yamaha arrangers.
Nobody knows nor understands why there has never been an XGW upgrade.

I wonder why Steinberg Germany ( 100% owned by Yamaha since 2002 ) never made ( up to now ) a real Yamaha arranger editing program ... but I am still hopeful and dreaming ... one never knows. Miracles happen sometimes. 


Best wishes, JH

[BogdanH]

hi Jeff,
I think we don't talk about the same thing... It's not the noise/distortion that is added intentionally by some effects (in that case it would be applied to all notes). The distortion I'm talking about is only present in C4-G4 key range... and so, it can't be removed (the distortion is inside voice).
The distortions sounds almost the same as damaged (ripped) speaker membrane. Distortion is not that loud, but is audible and makes sax voice very unpleasant to listen. When you hit one of C4-G4 keys, distortion doesn't start immediately -it starts when sax sound starts to vibrate and is easily noticeable.

Bogdan

[Jeff Hollande]

hi Jeff,
I think we don't talk about the same thing... It's not the noise/distortion that is added intentionally by some effects (in that case it would be applied to all notes). The distortion I'm talking about is only present in C4-G4 key range... and so, it can't be removed (the distortion is inside voice).
The distortions sounds almost the same as damaged (ripped) speaker membrane. Distortion is not that loud, but is audible and makes sax voice very unpleasant to listen. When you hit one of C4-G4 keys, distortion doesn't start immediately -it starts when sax sound starts to vibrate and is easily noticeable.

You are right ... that is another situation ...
Sorry ...
JH

[BogdanH]

No problem at all, Jeff... thank you for showing interest

Bogdan

[EileenL]

I have just tried this on my SX900. C4-G4 the breath noise is a little more pronounced but not unpleasant.

[overover]

Hi Bogdan,

Do you perhaps play with Initial Touch switched off, i.e. with Fixed Velocity? Since the sound changes with most voices due to the velocity ("playing strength"), I recommend always playing with Initial Touch switched on, especially with acoustic instrument voices such as the saxophone.


Best regards,
Chris

[pjd]

Only a few observations --

I agree that the breath noise is a little heavier (more distinct) in this note range. To my ears, it sounds like this key range is near the bottom range of the instrument samples. Get too far below C4 and it sounds unnatural.

I'm also hearing a little bit more sax key off noise in that note range.

This sax voice is probably based on the MegaVoice Tenor Sax (8/0/83) samples. There are breath noise, key on and key off samples. Could be a small issue with balance between the basic note samples and noises, i.e., a sound design/programming issue.

Hope this helps your exploration -- pj

P.S. I'm playing and listening to this voice on Genos. It is a legacy voice on Genos.

[BogdanH]

hi Chris,
Yes, I do play with Initial Touch ON. Thank you for asking! -because of your question I asked myself what happens if I turn Initial Touch off.. and here's my finding:

As we know, if we turn Initial Touch off, then touch sensitivity depends on Fixed Velocity (FV) setting, which goes from 1-127. The distortion I am talking about only occur at FV values 1-97. If you increase FV value to 98 (or higher), you can notice that sax vibrato speeds up and distortion practically disappears (btw. after each FV change, key needs to be pressed again, of course).

That can only mean, that for velocities above 97 different voice layer is used than at lower velocities. And why does it only happen on C4-G4 keys? Well, here I can only guess... because on Yamaha, we can't edit preset voices (or see voice settings) at layer level.

I believe that most of us use Initial Touch and that nobody really hit keys when playing (sensual) sax voices... and so distortion is always present when playing those notes. Of course, if you play some Rock'n'Roll, distortion won't be noticed, because it's lost in other instruments. But if you play some slow (solo) music, then it's hard not to hear it.

@pjd
-thank you for commenting. Well yeah.. we can discuss if that's "breath" noise or not.. if it is, then it's extremely poorly implemented.

Bogdan

[overover]

Thanks for your detailed answer, Bogdan!

If you've been playing with an Initial Touch Curve other than "Soft 2" (e.g. "Normal"), I recommend setting it to "Soft 2". This produces a relatively high volume (velocity) even with light playing strength, and it might also solve the problem you describe with some saxophone voices.

Another option would be to go into Voice Edit and set the parsmeters "Touch Sensitivity Depth" and/or "Touch Sensitivity Offset" so that the voice is always triggered by higher velocities. Then either save the voice in Voice Edit as a User Voice, or save the changed settings directly into a Registration.


Best regards,
Chris

[BogdanH]

hi Chris,
Thank you for suggestions. But the thing is, I don't wish to play sax at high velocities (either physically or by increasing sensitivity).. I wish sax to sound "gentle" -and for that, one just need to use lower velocities (to avoid harsh attack).

Bogdan

[overover]

Hi Bogdan,

Of course, these are all just suggestions (how I would go about it).

Of course, when played "quietly", i.e. with a low velocity, the sound of many voices is different. Maybe this saxophone voice has more "breath" noise at the beginning. On the other hand, you may be able to achieve a similar result with higher velocities, for example if you soften an attack that is too harsh by slightly increasing the Attack Time and/or turn down the Filter Cutoff or the Brilliance parameter to dampen the highs. If necessary, you can also lower the volume a bit in Voice Edit so that you don't have to turn the volume in the Mixer down too far.


Best regards
Chris

[Amwilburn]

Bogdan: yes those are (loud) breath samples; the thing is when you blow at a lower strength, you'll get relatively more breath noise (at high velocities, that breath noise practically disappears). But anyone playing a Tenor Sax at lower volumes, you do get a lot of breath sound. What we're hearing? Excess spit (technically it's condensation, but that's common in sax players. Yes, it sounds like a speaker membrane cracking. It's condensation aka splitty sax). Why at lower velocities? At high velocities (ie a hard blow) the spit noise gets expelled immediately along with the very growly loud note, so it's completely covered up.

Sax players are supposed to suck in the condensation before blowing a long gentle note for this exact reason. What Yamaha probably should've done is used a shorter sample *before* the condensation gets spat out, and loop from that. But that's a real sound, and authentic, if also somewhat gross :p

However, to get rid of it, you could just use the voice edit -> Sound -> Cutoff, and change cutoff to -40 or so; that will get rid of most of the breath noise (you'll still hear it come in faintly). It will change the sound of the sax., but  once you change the cutoff filter and start playing, you won't really notice it.

Mark

[BogdanH]

hello Mark,
Thank you for explaining where from "breath" comes when playing real sax instrument. But that doesn't really address the issue that I mentioned.
I like how that sax voice sounds and it fits perfectly my particular song. The problem is C4-G4 range only. If that's how breath should sound (which I disagree), then how come that breath disappears outside that range? That is, why is this breath present only in this range?
I can understand, that for some this breath doesn't sound disturbing, that some don't care and some maybe don't even hear it. Still, I stand to my opinion: it's a glitch that shouldn't be there.
Is there a workaround? Not really. We can apply some settings, but that only changes overall sound of voice. It's like trying to remove only noise from old audio tapes, without affecting the sound -it's impossible.

Bogdan

[Bill]

Hi Bogdan

I'm not at my KB at the moment, however just a quick thought.

Does the noise shift when you use the Octave Shift, or does it always stay on keys C4-G4.  I'm thinking - could it be the "breath" effect that is being exaggerated by dirty / damaged key pads.

Bill

[icatalin42]

Hi all,

I noticed something is wrong with the s.art. sax voice on sx600 too, and i've avoided using it since. In the beginning,i thought something was wrong with my brand new keyboard sx600 but then again i realized  something was terribly wrong with the s.art sax voice. By the way, i have the feeling more s.art voices have those kind of sample problems on sx600, thats my impression, i gently avoided s.art. voices.

[Jeff Hollande]

Might it be an s.art problem or is it a keyboard problem or both ?

JH

[icatalin42]

I think it is a voice problem, something related to the samples used. I noticed this the moment i turned it on. Seems not all sx600 have this problem, take a look on the video below:
https://youtu.be/jADeNVT8QT4
May it be related also with styles?this could be something wrong in thekeyboard style+voice electronic management??? voice alone doesnt seem to have those issues or i may be wrong.

[BogdanH]

@Bill,
It's tone pitch specific (it doesn't occur at specific physical keys only) and so it's not a hardware issue.

@icatalin42
I really can't talk about SX600, because I don't have it to try. And if music is busy enough or style contain other noises (i.e. snares), then the distortion I'm talking about isn't really that noticeable.

@Jeff
In my opinion, it's definitely (s.art) voice problem. It's obvious that vibrato is artificially added and I think that this vibrato effect (in combination with some other setting) is causing that distortion -because it only happens in C4-G4 region at 1-97 velocities. Means, it only occurs in specific voice layer(s). And as I mentioned, on Yamaha keyboards we can't edit layer specific settings.

Some maybe think I'm making a big fuss about nothing.. so be it. I'm only reporting my finding and opinion. Is a information some may find useful.. and some not 

Bogdan

[Jeff Hollande]

Hey Bogdan :

Personally I find it very useful and interesting members are sharing
findings, opinions and comments even if they do not always agree.
The only way to learn something new, IMO.

An arranger is a complicated piece of technology that has been called " a music instrument ".
I wonder if it is a real music instrument even if it has piano keys ? ... Sorry, I cannot answer that question ... maybe others have an opinion and/or comments.

Mostly  " real " music instruments are made for a life time.
An arranger is mostly made for a period of time and becomes outdated very soon, just like a computer.
Updates and upgrades are very usual for an arranger.

It is also called a One Man Band ( OMB ).
Perhaps DOMB ( Digital One Man Band ) might be a better name.

Take care, JH

[Amwilburn]

hello Mark,
Thank you for explaining where from "breath" comes when playing real sax instrument. But that doesn't really address the issue that I mentioned.
I like how that sax voice sounds and it fits perfectly my particular song. The problem is C4-G4 range only. If that's how breath should sound (which I disagree), then how come that breath disappears outside that range? That is, why is this breath present only in this range?
I can understand, that for some this breath doesn't sound disturbing, that some don't care and some maybe don't even hear it. Still, I stand to my opinion: it's a glitch that shouldn't be there.
Is there a workaround? Not really. We can apply some settings, but that only changes overall sound of voice. It's like trying to remove only noise from old audio tapes, without affecting the sound -it's impossible.

Bogdan

It's not just breath, it's basically spitting condensation after a long note. It only shows up like that after a gentler blown note after a while; as I mentioned above it won't happen with a faster velocity note because the "spit take" gets expelled immediately. It's not a mistake on the recording per se; that *really happens* playing the sax. It's just the player who performed the sample should have sucked in all the condensation prior to recording gentle blows in that note group (hence why it only shows up in those notes). And as mentioned above, if you change the cutoff to -40 or so, it will minimize that nasty spit take sound that comes in after a while.

What Yamaha should have done is chopped the recording off before the spit take happened, and looped it based on the first part, but they didn't, and now that you've pointed it out I can't *not* hear the spit. But since they didn't, all we can do is filter it out as best we can; again as I'd already mentioned above.

This guy doesn't demonstrate a *really* spitty sax, but his description says it all: 30 seconds in: too much saliva and condensation, and your saxophone "sounds like a crackly old vinyl record".

https://www.youtube.com/watch?v=unydN0iyPuc&t=54s

No it's not a keyboard problem, it was recorded that way. It's the same S.Art sax sample on sx600 and sx900 as well, so no difference there. And it only shows up at lighter key velocities. No it shouldn't have been recorded that way. It's a recording problem, not a hardware problem.

This lady describes the issue perfectly
https://www.youtube.com/watch?v=Rze8SJOwIxU

Again, that's an issue with the sax player, not the hardware of the PSR. And most likely they recorded the lower notes first, allowing the condensation to really build up when they got to these high notes (the highest note on a standard Bb sax is F#, which would be the highest note sampled).

Mark

[BogdanH]

Thank you for participating Mark, I really appreciate.
I used word breath to keep post short, but we can also call it condensation issue, distortion, playing technique, noise... Point being, it only happens for C4-G4 notes and that fact alone proves it's not something musician did at recording the samples.
Cutoff filter can hide the distortion to some degree, of course. But it also greatly influences the overall voice sound: sax loses all brilliance and at setting below -20 becomes totally dull. Which is not surprising, because at the end, cutoff filter does what it's supposed to do.

Bogdan

[travlin-easy]

Well, maybe because I've experienced significant hearing loss over the past 8 decades, but even wearing my Bose Hearing Aids, I cannot hear what you are talking about using the S-950's sax voices. No spitting sound, no distortion, no exaggerated breath sounds. In fact, it sounds more like a real sax than the sax one of my old band members used to play.

Bogdan, how about posting that sax sound here so we can all listen to it.

All the best,

Gary

[Amwilburn]

Thank you for participating Mark, I really appreciate.
I used word breath to keep post short, but we can also call it condensation issue, distortion, playing technique, noise... Point being, it only happens for C4-G4 notes and that fact alone proves it's not something musician did at recording the samples.
Cutoff filter can hide the distortion to some degree, of course. But it also greatly influences the overall voice sound: sax loses all brilliance and at setting below -20 becomes totally dull. Which is not surprising, because at the end, cutoff filter does what it's supposed to do.

Bogdan

It's the fact that it only shows up at the higher notes and at lower velocities that proves it was something the musician did! Because they sample note ranges sequentially. (usually lowest to highest). Which is exactly why the condensation would only show up in the highest notes on the sax (notes above that aren't based on the new S.Art sax). I work in a music store, I hear violins, cellos, saxophones, guitars daily. WHY no condensation crackling above that? It uses a completely different sample above that (Alto sax perhaps, but I don't know which one).

Unfortunately, since Yamaha took *way* too long a sample for those notes, filtering (or switching to a diff sample for those notes, e.g Right 3 Right 2 split using another sax for Right above the C4) the only thing I can think of is filtering. Even EQ won't work, as it's too wide a frequency range

Lone Arranger: it's the same S.Art Saxophone from the s950; you'd never have heard it if you didn't play gently and held the note for a while. I never encountered the badly played notes before because when holding a note, I always accent harder, not softer. But if you play the highest E or F (those notes use a particularly egregious sample) softly and hold for a long time, you hear the player's condensation make a horrible crackling sound. Again, I've heard this from an actual sax players in person when they don't warm up the sax first and condensation develops on a long, gentle sustained note.

And knowing what it is makes it kinda gross :p

Mark

[BogdanH]

hi Gary,

... I cannot hear what you are talking about using the S-950's sax voices...
Bogdan, how about posting that sax sound here so we can all listen to it.

You're right.. is hard to explain with words only.
I just published a video on Youtube -I hope that I managed to explain what I mean 

Greetings,
Bogdan

[Amwilburn]

hi Gary,You're right.. is hard to explain with words only.
I just published a video on Youtube -I hope that I managed to explain what I mean 

Greetings,
Bogdan

Sigh. Based on your video, you seem to think I meant that saxes are supposed to sound like that.

As I've already said, that's not "just how sax is". It is a problem when a sax player doesn't suck in the condensation and/or warm the sax before playing and you get that awful crackling sound. But no, sax isn't *supposed* to sound like that, unless played unprofessionally. But they recorded it that way. It *is* a real thing, but they should have basically taken a break and recorded the sax after the condensation was removed. Either they didn't want to pay the studio session time to wait to clear the conensation (which Yamaha should have been able to afford) or the engineer recording it didn't notice the spitty take.

Also, the same issue showing up across  7 or 8 semi tones *could* mean they sampled it only once, yes. But it would also show up if they'd sampled each note (which I also doubt they did.  As you know, typically it's every 3rd-5th note, but I've seen some get away with once an octave, so once in 7 semitones is certianly possible)

And yes, they reuse samples for other voices. They used to even tell you in the info description: "Same piano as Grand Piano1, but brighter; same etc but mellower". About 10-20 years ago, the info button would often even mention how many dynamic sample layers! Unfortunately they no longer  include the info button

I get that you're disappointed, you think it's a distortion caused by incorrect VU levels during recording. I'm explaining that's not the case *here*, it can happen with sax playing *live*, *especially* with gentler blows... but when playing for a sample, they should *never* have left the spitty recording in, unless they specifically called it "spitty sax".  No, they should *not* have used a spitty take for a pro grade sample, you're not wrong there.

Here's another youtube video explaining how the condensation builds up in a sax (they even have 'spit absorbers' for sax players to help with this issue).
51 seconds in:

https://www.youtube.com/watch?v=Z7HsGtwABkw

Are you hearing things? Absolutely not; in fact, the fact that you're probably the first person to even notice the spitty takes means you have *excellent* hearing. Does my explanation fix the issue? No.
*Is* it an issue? *YES* it is. Sax should be breathy at low breath intensities (which is how they recorded it exactly 1 octave lower*), but it shouldn't be spitty. That's unprofessional. Really *can* happen, but really shouldn't.

*Again, when recording samples in a studio, the engineers usually request lowest to highest note sequence. As spitty sax sound usually doesn't even show up right away, it's when the condensation becomes enough to condense into actual drops, you *couldn't* get a spitty take on the first notes you recorded if you tried.


It basically means the musicians and the engineers didn't communicate properly. Imagine if they sampled a guitar player pitch bending several strings at once , but the player using a Floyd Rose bridge didn't nail the relative pitch and it didn't actually end up on a standard note. It would be awful, and they should definitely not use that sample. Unfortunately, that's what they did here.

I'll email you a registration to bypass the spitty low sample takes by switching to Alto (interestingly, neither the S.Art alto includes any spitty takes, nor the older single or double dynamic layer sample saxes include any spitty notes). Not really a solution more of a workaround.

Btw I just tried the Jazz sax on the PA5x. Similar problem on the highest notes, but somehow it's worse: you can tell they tried to apply some sort of AI algorithm to hide the spittiness, so what ends up happening is there's a weird... almost a click... when the alogirthm is applied and then when it comes off. And it gets even weirder because they only have 1 sample for the key off, so there's a secondary choke sound that's way too attacky for a gently blown note (it actually sounds brilliant at higher velocities, but sounds even worse than your example for lower velocities). If you can go to a store and try it! It's present in all notes D4 and above.
Tenor Jazz sax 1, they also left the spit take in (but it's not as loud)... but it will drive you nuts because where Yamaha recorded the entire length up to the spitty take, and the looped after? The Korg loops right before. So you actually hear the spitty sound on and off... so you can hear the length of the loop is about 1/5 of a second, and like it's constantly 'puffing' in the background 5 times a second. *not* a sound a sax can produce, no. To be fair, the S.Art Yamaha sax, also has the same repeating puffing distortion from the loop at F3.
It will also drive you bonkers... I can't unhear it now.

Interestingly, on the Yamaha, at  B3? The lowest sample layer isn't even there, I'm only detecting 2 (neither of which has a spitty take at low velocities).

Mark

[travlin-easy]

For me, this is a nothing burger. I could barely hear the noise you were referring to using high quality headphones. An audience would never hear it!

Good luck,

Gary

[overover]

I recommend anyone who finds such sample errors in Preset Voices to contact Yamaha Support directly (with a detailed description of the discrepancies found and links to the corresponding audio files or videos).

Yamaha may then fix some of these errors in future updates. (The multi samples of the Preset Voices can be changed at any time via a firmware update.)


Best regards,
Chris

[Amwilburn]

I recommend anyone who finds such sample errors in Preset Voices to contact Yamaha Support directly (with a detailed description of the discrepancies found and links to the corresponding audio files or videos).

Yamaha may then fix some of these errors in future updates. (The multi samples of the Preset Voices can be changed at any time via a firmware update.)


Best regards,
Chris

Great idea Chris! Simpler than the Registration workaround I suggested.

Incidentally, as someone who's been a retailer for over 35 years, I can tell you this: sometimes there are things that are intentional that don't sound good to everyone. I once had a customer privately re-sell a TYros 2 because he said there was a 'crackling scratching' sound so he thought it was was defective. I knew immeditaely what he was talking about, and played the Acoustic Rock style, he said "yes, that's the crackling scratching sound!" So I went into style editor, played the offended D0 note that's intentially a looping sample of a vinyl record, then deleted the note from the style, and voila, no crackling vinyl sound.

He looked flabbergasted. Then purchased a used CVP, which I warned him contained the same sample

Mark

[BogdanH]

Sigh. Based on your video, you seem to think I meant that saxes are supposed to sound like that.
...

-not at all! I don't know what makes you thinking that.
My comment in video (related to your posts) says "..some said that's how saxophone can sound at certain conditions.." -which is true and you explained conditions at which that happens. But I never said or insinuated that it's supposed to sound like that.
The only thing where our opinions differ is, you think that this effect happened at recording of the sample, while I believe it happened in post-processing. But we agree that it shouldn't be there.
Btw. As you suggested, I tried using Cutoff filter. As we know, it's a normal low-pass filter with variable frequency setting. I have set it to -40, recorded the sample and checked on PC. And the result I could see is, at this setting it starts to cut off all frequencies at about 4kHz and above -no wonder it drastically changes the overall sound of the voice.

.. An audience would never hear it!

hmmm.. I didn't search for voice flaws, I heard it while I was practicing... and if I heard it, so can others. Never underestimate audience 

@Chris
You're basically right. But we all know, when talking about voice/sound quality, we're always in muddy water. If we complain about "this doesn't work as described in manual" then the thing is pretty clear. But if I say "this doesn't sound as it should", then.. well, you know what I mean 

To conclude... I'm not expecting the best voices in 1200€ keyboard. But having Yamaha brand in mind, I expect them to sound as intended. Are such issues a deal breaker? Depends.. some don't hear it, for some it's irrelevant, some might find it as a voice feature, etc. -so pick your choice.

Thank you all again for sharing your opinions,
Bogdan