Vienna Symphonic Library Forum
Forum Statistics

180,768 users have contributed to 42,140 threads and 254,362 posts.

In the past 24 hours, we have 3 new thread(s), 15 new post(s) and 57 new user(s).

  • SYNCHRON STRINGS vs. Old VSL strings Demos (what we were talking about...)

    last edited
    last edited

    I've been invited recently in the forum, to add music to words, to better substantiate feedback and opinions... I was thinking it's a good way of introducing my self, beside giving objective reference to nice people interested in the topic. Well, here we are.

    Here some samples of what I did in the last 5 years (after being beta tester and demo maker for some developers as Garritan, KirkHunter, BelaDmedia, Sampletekk, SampleModeling and Audiomodeling): looking for orchestral strings VSL was recently always my first choice.

    The Dimension strings enable creating little sections, perfect for baroque and classical music historical performance emulation:

    Mozart - K622

    Haendel - WaterMusic (go to 0:58 to save time)

    The blend of VSL Orchestral and Chamber Strings (Chamber still being in my opinion a milestone and a personal favorite one) produce expressive romantic sound (despite people describing VSL as "too cold"):

    Morricone - Mission (go to 1:03 to save time)

    All the samples are produced with VE Pro and MIR Pro for mix and ambience.

    Here my test of Synchron strings now, no more MIR Pro, just a very little MIRacle insert, over the naked "Stereo Mix" patches.

    Grieg - Holberg suite (wet)

    Since "wetness" is a pretty arbitrary and personal choice, if you want to play with it, please find here the totally dry output, you may download and process with your favorite finish reverb:

    Grieg - Holberg Suite (Dry)

    Comments, suggestions and questions are welcome.


  • My feedback is that it's nicely done work. ^-^


  • Wow!   Very nice - the Grieg performance is quite expressive and natural =)

    I preferred the dry version for the Grieg piece.  But the wet version was also great, I agree it's a matter of taste at that point.

    Which articulation did you use around 0:30 at the top chord ending the phrase?  It is placed beautifully.

    Compared to "Old" VSL strings, the mix sound is night and day.  The first examples you shared still sound natural, but there is a spatial definition that the synchron mix has, which I really like a lot.  I was never good enough to mix the dry VSL libraries with that spatial definition, I think ultimately it has to do with using spaced microphone pairs which was the opposite of the dry VSL approach.  So I'm excited about Synchron series for this reason.

    A few parts of the Grieg piece did feel unrealistic to me - some of the shorts were just too precise, the very fast runs didn't blur right, and the very ending chords had a weird sheen to them.  Did you feel the same way too, and just ran out of time?  About the runs in particular, I have a few questions - (a) which articulation did you use and (b) did you use mono or poly legato, and what were your release/blur settings? 

    For me, so far, I found that poly legato with some bit of overlap in the MIDI notes, using the slur legato articulation, would sound OK for runs.  I noticed that overlapped notes in poly legato mode have a different kind of blur result than the legato blur / P. Release feature, which I think works better for agile runs.  But I haven't spent enough time with Synchron yet, maybe there are better ways to do more convincing string runs.


  • last edited
    last edited

    @djw said:

    My feedback is that it's nicely done work. ^-^

    Thank you 😊


  • last edited
    last edited

    @suon said:

    Wow!   Very nice - the Grieg performance is quite expressive and natural =)

    I preferred the dry version for the Grieg piece.  But the wet version was also great, I agree it's a matter of taste at that point.

    Which articulation did you use around 0:30 at the top chord ending the phrase?  It is placed beautifully.

    Compared to "Old" VSL strings, the mix sound is night and day.  The first examples you shared still sound natural, but there is a spatial definition that the synchron mix has, which I really like a lot.  I was never good enough to mix the dry VSL libraries with that spatial definition, I think ultimately it has to do with using spaced microphone pairs which was the opposite of the dry VSL approach.  So I'm excited about Synchron series for this reason.

    A few parts of the Grieg piece did feel unrealistic to me - some of the shorts were just too precise, the very fast runs didn't blur right, and the very ending chords had a weird sheen to them.  Did you feel the same way too, and just ran out of time?  About the runs in particular, I have a few questions - (a) which articulation did you use and (b) did you use mono or poly legato, and what were your release/blur settings? 

    For me, so far, I found that poly legato with some bit of overlap in the MIDI notes, using the slur legato articulation, would sound OK for runs.  I noticed that overlapped notes in poly legato mode have a different kind of blur result than the legato blur / P. Release feature, which I think works better for agile runs.  But I haven't spent enough time with Synchron yet, maybe there are better ways to do more convincing string runs.

    Thank you for the detailed feedback: actually I agree with all your point:

    The articulation ending the phrase at 0:30 is the Long Lyric Vib, (following the S-Short pattern). I find this patch very expressive and effective, for low and mid dynamic range. It sounds a bit less effective in fff, as i will tell you later.

    The spacial definition of MIR Pro is pretty good, but due to the fact I get sometime perfect results and sometime disputable and weird results, it's obvious that a bit of research and sound engineering experience is requested to really master it: it's not enough placing the instruments on stage and press play 😊 In particulare the Demos were designed to show the fuild articulation of the solos, so the actual version does the job, but as a stand alone music they are not ready yet: they don't blend in the mix at all (because solos were coming from anechoic recordings and I didn't find the right early reflections to mix them with VSL samples sound), and it was absolutely my fault and sound engineering weakness . So Synchron is welcome, but I will still struggle with external sound sources to blend in the mix... 😛

    Yes I agree with your comments about precision and ending chords, they come from different reasons by the way:

    - the excessive precision is coming from what you say: I run out of time 😊 and didn't edit the MIDI track yet. Easy to fix with usual technics of randomization and humanization, very effective the VI Pro patterns as well.

    - the final chords were really a pain: I didn't spend enough time to experiment yet, but with a fair amount of trials and manipulations (testing all the long and dynamic patches), still the result is far away from the expected one (lack of power or brilliance: if I get power it sounds harsh, if I get brilliance it sounds too soft etc.) It was a lybrary issue I didn't find a way to fix yet. Probably blending different patches will be the solution.

    - I used mono legato in monophonic lines and poly legato in simple divisi, and didnt' play with blur yet (Blur was still at 0). I didn't find yet the best blur value, but the piece wasn't relying too much on legatos, that's why. By the way your comment about the fast run is correct, not yet edited to perfect realism, I was still testing what the lybrary does "as it is", mixing a bit of legato and a bit of spiccato to create articulation: it sounds nice, but as a "soloist" not as an ensamble should play, due to excessive precision.

    - Articulations choice of Synchron are not so large: I find using the general purpose and very consistent patches a real pleasure and a lot of time saving for the "first quick mock-up". Long LyV for sustain, S-Short, short and Sfz for rythmic patterns, Legato and legato slurs LyV as well to blend sustain.

    In conclusion I found a kind of polarizing 80/20 effect: the speed to reach an acceptable quality was incredible. In a very quick time it's possible to reach a very convincing sound, due to very consistent dynamic and control (I consider it up to now the real great step forward and power of the lybrary) and so coherent patches. 20% of time 80% of result...

    But I suspect the finishing of it will be very painful (80% of time for missing 20% of result? no more more than 80% LOL). The simplicity and wet character of the lybrary will rise new challenges for creating nuances and little fine tunings. But it's a pretty sophisticated tool... a learning curve is requested and Synchron player is still missing. I'm looking forward to the next steps.


  • Bravo, beautifully programmed. The use of Dimension strings for a small baroque ensemble is especially successful, IMO.

    Do you ever find the 'bleed-through" of adjacent instruments in the Dim.Strings recordings to be problematic? I mean the slight phasing of the sound quality. How do you work around that? 

    The Synchron strings example shows the potential success and also the danger with the more careful &  extensive sampling employed. The performance has entered the "uncanny valley", familiar from modern computer animation, where the image is extremely lifelike but still clearly machine-driven. Perhaps more editing and performance nuance would clear that up in your demo. 

    I am familiar with this issue from years of playing the physically-modelled Yamaha VL instruments in live performance. Some sounds were extremely realistic and also had the full 128 level dynamic resolution of MIDI (unlike samples, which often have only 2,3, or 4 dynamic levels.) If the perfromance was too "linear", the machine character of the instrument would come through. I sense some of that in the Grieg demo. 

    Also, I'm not sure I like the baked-in reverb of Synchron on such busy music. The sound of the room seems to pulsate with the musical phrasing in a peculiar way, as if the room is coming and going with each dynamic nuance in the phrase. Maybe a layer of overall reverb from MIR would smooth it out? (I've been using Synchron drums, but only  the close mic, and adding MIR to blend it with the other VSL instruments.)

     

     

    thanks


  • last edited
    last edited

    @brunodegazio said:

    Bravo, beautifully programmed. The use of Dimension strings for a small baroque ensemble is especially successful, IMO.

    Do you ever find the 'bleed-through" of adjacent instruments in the Dim.Strings recordings to be problematic? I mean the slight phasing of the sound quality. How do you work around that? 

    The Synchron strings example shows the potential success and also the danger with the more careful &  extensive sampling employed. The performance has entered the "uncanny valley", familiar from modern computer animation, where the image is extremely lifelike but still clearly machine-driven. Perhaps more editing and performance nuance would clear that up in your demo. 

    I am familiar with this issue from years of playing the physically-modelled Yamaha VL instruments in live performance. Some sounds were extremely realistic and also had the full 128 level dynamic resolution of MIDI (unlike samples, which often have only 2,3, or 4 dynamic levels.) If the perfromance was too "linear", the machine character of the instrument would come through. I sense some of that in the Grieg demo. 

    Also, I'm not sure I like the baked-in reverb of Synchron on such busy music. The sound of the room seems to pulsate with the musical phrasing in a peculiar way, as if the room is coming and going with each dynamic nuance in the phrase. Maybe a layer of overall reverb from MIR would smooth it out? (I've been using Synchron drums, but only  the close mic, and adding MIR to blend it with the other VSL instruments.)

     

     

    thanks

    Grazie Bruno. 😊

    Yes I also had some trouble with the "bleed" of adiacent instruments in Dimension strings: it obviously prevents single players to be used in very exposed or solo application, but also makes the sections always sound "bigger". For instance the 3 players section still sounds something undefined as 4 or 5 but with higher definition and more expressive vibrato, then it's ok. To work around that I was (a bit painfully to be honest) doing several trials with MIR pro positioning and usually it's possible to find a position and a mix where it's not a problem.

    I like your comparison to Computer Animation... yes we are in the same field of "hyper-realistic rendering" still keeping a bit of fake feeling from several ingredients (mostly excess of perfection and lack of noise and random sound modulation, including the lost little performance and sonic interactions of instruments, the first maybe possible to fix with a lot of very expert time editing, but the second almost impossible to replicate: early reflections and sympathetic vibrations are impossible to reproduce with sampled instruments of course). But... in the case of Grieg it's mostly my fault: I was focusing on the sound/articulation and I admit I didn't yet spend time refining the micro-timing of the sequence, as I do with final productions.

    About the problem with "baked-in reverb" pulsations I even opened a specific post before 😊. It's a double sided feature the user has to carefully manage. I find my usual work-flow not compatible with Synchron, and I started thinking about and experimenting. My actual opinion is that to keep the quality of Synchron ambience you MUST let the reverb ring, and every serious manipulation of volume is compromising realism a lot. Of course it's possible to add umbrella reverb, (as I did in wet vs. dry examples) and also use the close MIC and try to re-position the Synchron in MIR. Well... of course it sounds as a contradiction, being the Synchron reverb one of the main added values of the library, but another incredible technical advantage of the Synchron String is the sonic and dynamic coherence of the patches and articulations: it's an unvaluable time saver, I found myself working less than 1 half of the time for articulation choice and blending compared to the same activity with Dimension. So why not? The only problem I found up to now is that the residual reverb and mono signal in close samples is creating artifacts in MIR, but probably again is a matter of research and set-up.

    Thank you again for the nice conversation, and have fun.


  • last edited
    last edited

    @fatis12_24918 said:

    About the problem with "baked-in reverb" pulsations I even opened a specific post before 😊. It's a double sided feature the user has to carefully manage. I find my usual work-flow not compatible with Synchron, and I started thinking about and experimenting. My actual opinion is that to keep the quality of Synchron ambience you MUST let the reverb ring, and every serious manipulation of volume is compromising realism a lot.

    If you or someone else has Synchron Percussion, is this a problem there at all?


  • last edited
    last edited

    @fatis12_24918 said:

    About the problem with "baked-in reverb" pulsations ...

    If you or someone else has Synchron Percussion, is this a problem there at all?

    Bruno in his post mention Synchron percussions. But I'm sorry I don't own them, just Synchron Strings. (My personal post was about strings indeed, sorry for being unclear), I have no experience with Synchron percussions, so maybe Bruno can comment about your question.


  • last edited
    last edited

    I can't really answer your question because wIth Synchron Percussion I've used only the Close mic, adding MIR for reverb. That way it mixed nicely with my other VSL instruments, and I had no problems with the baked-in reverb pulsating. 

     

    @fatis12_24918 said:

    About the problem with "baked-in reverb" pulsations ...

    If you or someone else has Synchron Percussion, is this a problem there at all?

    Bruno in his post mention Synchron percussions. But I'm sorry I don't own them, just Synchron Strings. (My personal post was about strings indeed, sorry for being unclear), I have no experience with Synchron percussions, so maybe Bruno can comment about your question.


  • last edited
    last edited

    @fatis12_24918 said:

    About the problem with "baked-in reverb" pulsations ....

    If you or someone else has Synchron Percussion, is this a problem there at all?

    Probably you know it better than me, but anyway I'm expecting Percussions being really less problematic using the wet samples, because a part for some dynamic repetitions (e.g. rolls) it's unlikely you use Expression on them. The available crescendos and diminuendos should totally cover the needs and keep the natural reverb almost always ringing. (that's the "think Synchron" phylosophy in my understanding).

    Anyway nice to know that it's possible to succesfully mix them in MIR using close samples.

    (Probably easier with Percussions, being an almost monophonic source, than a string section, which missing a bit of depth and dimension if you use the close mono mic instead of the stereo ambience image...)


  • I've never experienced Dimension Strings' bleed through as a problem at all.  It makes the players sound more like a section.  The only time it would be problematic is if you really wanted to use a player as a leading soloist, which is not recommended for so many reasons, the bleed through actually not being the most important of the reasons.  So overall DS is one of my favorite libraries.

     

    As for the demos above, they all sound very good, though I do agree that Synchron has a few choppy-sounding moments.  I don't want to point the finger at your programming or at VSL's programming, but overall that's what I'm hearing.  The shorts, however, are really some of the best out there!


  • last edited
    last edited

    @Casiquire said:

    I've never experienced Dimension Strings' bleed through as a problem at all.  It makes the players sound more like a section.  The only time it would be problematic is if you really wanted to use a player as a leading soloist, which is not recommended for so many reasons, the bleed through actually not being the most important of the reasons.  So overall DS is one of my favorite libraries.

     

    As for the demos above, they all sound very good, though I do agree that Synchron has a few choppy-sounding moments.  I don't want to point the finger at your programming or at VSL's programming, but overall that's what I'm hearing.  The shorts, however, are really some of the best out there!

    Summarizing, I agree. ðŸ˜Š


  • Bravissimo! If there is any demonstration that has convinced me of the value of the Syncron approach, it is your Holberg simulation. It is the best of its kind I have heard anywhere. You should give classes on orchestral simulation.

    I agree with most comments made by others, but I take a different perspective. The sharpness, clarity and "fattiness" of the sound here are such a welcome improvement in sampling. I've spent hours trying to imprint these qualities in my simulations, with very mixed results. It will be easier, I think, to shave off and blur the sharpness than it was to improve the sharpness of more sloppy sounds. To use a cryptic formula, it's always easier to do less with more, than to do more with less.

    A last note: I have never heard any simulation that could really stand up the comparison with a top notch orchestra. In this case, I listened to Karajan directing the Berlin Philharmonic in the Holberg Prelude, and I must say that your simulation held up very strongly. Not perfect, OK, but extremely convincing, with a sound envelope that is very dense and lifelike. Now, the challenge is to humanize and blur the contours, an easier task than trying to make blurry stuff more incisive. Once again: bravissimo!


  • last edited
    last edited

    @Yan Barcelo said:

    Bravissimo! If there is any demonstration that has convinced me of the value of the Syncron approach, it is your Holberg simulation. It is the best of its kind I have heard anywhere. You should give classes on orchestral simulation. I agree with most comments made by others, but I take a different perspective. The sharpness, clarity and "fattiness" ofthe sound hereare such a welcome improvement in sampling. I've spent hours trying to imprint these qualities in my simulations, with very mixed results. It will be easier, I think, to shave off and blur the sharpness than it was to improve the sharpness of more sloppy sounds. To use a cryptic formula, it's always easier to do less with more, than to do more with less. A last note: I have never heard any simulation that could really stand up the comparison with a top notch orchestra. In this case, I listened to Karajan directing the Berlin Philharmonic in the Holberg Prelude, and I must say that your simulation held up very strongly. Not perfect, OK, but extremely convincing, with a sound envelope that is very dense and lifelike. Now, the challenge is to humanize and blur the contours, an easier task than trying to make blurry stuff more incisive. Once again: bravissimo!
    Thank you for the kind words, and yes actually I totally agree with your point, easier to do less with more, than the opposite!

  • fatis

    Somehow I missed this - it is truly excellent! The Dimension Strings pieces sound so good they should be VSL demos - they are basically a perfect performance and mix sound.  And they have the wonderful intimate quality that the Dimension strings can give, which is totally unlike other sample libraries which usually are massive numbers of players in a block.  It is a very "musical" performance showing a great amount of sensitivity and delicate phrasing. 

    Interesting to notice how when the Synchron performance starts, at first with the very fast short notes it sounds outstanding, perfect - and then when the high legato violin line comes in it instantly identifies itself as "SAMPLED." Too bright, even shrill, the legato not natural-sounding, no sense of the individual players.  If the same line was done with humanized Dimension Strings it would sound perfect probably.  But that is not a criticism of your performance, it is an aspect of the library which is different.  Your performances are fantastic and show a real expertise and musicality -  some of the best I've heard!  


  • I just want to say, that the dimension strings sound so well, incredibly well. Respect! And what I heard from other music done with Dimension Strings that's for sure not only the library but also the way it is used. Very good.


  • last edited
    last edited

    @William said:

    fatis

    Somehow I missed this - it is truly excellent! The Dimension Strings pieces sound so good they should be VSL demos - they are basically a perfect performance and mix sound.  And they have the wonderful intimate quality that the Dimension strings can give, which is totally unlike other sample libraries which usually are massive numbers of players in a block.  It is a very "musical" performance showing a great amount of sensitivity and delicate phrasing. 

    Interesting to notice how when the Synchron performance starts, at first with the very fast short notes it sounds outstanding, perfect - and then when the high legato violin line comes in it instantly identifies itself as "SAMPLED." Too bright, even shrill, the legato not natural-sounding, no sense of the individual players.  If the same line was done with humanized Dimension Strings it would sound perfect probably.  But that is not a criticism of your performance, it is an aspect of the library which is different.  Your performances are fantastic and show a real expertise and musicality -  some of the best I've heard!  

    Thank you William. Yes I agree with your comment: I had big pleasure with some of the library features (mostly shorts and consistency) while some trouble with long notes. My main issue is I can't really use the cc11 at the extent I can do in VI, due to the wet tail being impacted. I'm also missing the power in high velocity, because in my opinion they move to "harsh" before reaching the full fff sound. For the rest I think they have pretty good applications.


  • last edited
    last edited

    @MMKA said:

    I just want to say, that the dimension strings sound so well, incredibly well. Respect! And what I heard from other music done with Dimension Strings that's for sure not only the library but also the way it is used. Very good.

    Thank you very much, and about synchron, I agree: in my humble opinion the main issue is the reverb management. The release samples and the algorithmic reverb are producing some strange effects in fast passages and long notes cross fading as well. I always find better for fast music turning the reverb off, avoid the release overlapping in legatos, and let the release ring as much as possible at the same level of the last sound to keep synchron stage natural reverb audible.